Report No. 13526-TO Togo Overcoming the Crisis, Overcoming Poverty A World Bank Poverty Assessment June 25, 1996 VNOF )~)Li a t I )I1 j 11( [ I LI 1 1 ()L I ( l '1 ( )Oper, I Dit 11 Ji\ SOnl W'eSt (enltril A\rfri(i Dp[),rtnivn'i Document of the World Bank CURRENCY EQUIVALENTS (As of June 25. 1996) Currency Unit: CFAF Official Rate: US$l= 498 c CFAF I= US$.20 FISCAL YEAR January I - December 31 GLOSSARY CNCA Caisse nationale de credit agricole APS Agent de promotion sociale AVE Association villages d'entreprises FUCEC Federation des unions cooperatives d'epargne et de credit CPDE Caisses populaire pour le developpement de l'entraide socio-economique GIPATO Caisse d'epargne et de credit dii groupement interprofessionnel des artisans du Togo CECA Cooperative d'epargne et de credit artisan OTP Office togolais des phosphates CERAD Centre d'etudes de recherche-action et d'appuis pour le developpement SONACOM Societ6 nationale de commerce OPAT Office de produits agricoles Togolais ECOWAS Economic Community of West African States MDRET Ministere du developpement rural, environnement et tourisme Ministry of Rural Development, Environment and Tourism SOTOCO Societe Togolaise du coton DRDR Direction regionale du developpement rural CRAN Centre regional de recherche en alimentation et nutrition OCCGE Organisation de coordination et de cooperation pour la lutte contre les grandes endem ies CRS Catholic Relief Services DESA Direction des enquetes et statistiques agricoles DS Direction de la statistique OPAT Office des produits agricoles Togolais OCDI Organisation catholique pour un d6veloppement integre SAF Structural Adjustment Facility CRAN Centre Regional de Recherche en Alimentation et Nutrition PNLP Programme Nationale de Lutte contre la Pauvrete PEF Programme d'Emploi et de F ormation CLORED Centre de Loisirs et de Reeducation pour Enfants Defavoris6s FODES Federation des Organisations de Developpement de la Region des Savanes TABLE OF CONTENTS Page PREFACE EXECUTIVE SUMMARY .......................................................i 1. THE SETTING .......................................................i Background ......................................................1 The 1980s: Structural Adjustment Efforts ....................................................... 2 1991-93: Political Upheaval and Econornic Deterioration ................ .................5 1994-95: The Recovery Stars .......................................................6 2. A PROFILE OF POVERTY ......................................................8 Defining Poverty ......................................................8 A Poverty Line .......................................................9 Characteristics of Poor Households ..................................................... 13 Changes since 1987-89 ..................................................... 15 Expenditure Pattems and Poverty ..................................................... 17 Nutritional Indicators ..................................................... 19 Social Indicators ...................................................... 23 3. TEIE DYNAMICS OF POVERTY .............................. ....................... 26 The Concept of Vulnerability ...................................................... 26 The Causes of Vulnerability ..................................................... 27 Vulnerable Groups ..................................................... 33 4. INCREASING TEE POOR'S PRODUCTIVITY: FACTORS OF PRODUCTION ..................................................... 39 Land ..................................................... 39 Labor ..................................................... 44 Capital: Access to Credit ..................................................... 49 5. RURAL DEVELOPMENT AND FOOD SECURIrY ..................................... 54 Role of agriculture in the national economy ...................................................... 54 Policies, Instruments and Poor Fanner's Incomes ............................................. . 55 Policies and Instruments for Food self-sufficiency .............................................. 60 Are the Poor Food-Secure? ..................................................... 65 6. SOCIAL SERVICES AND SAFETY NETS .................................................... 67 Health ..................................................... 67 Education ..................................................... 73 Safety Nets from the Government ..................................................... 79 The Role of NGOs ..................................................... 82 Traditional Safety Nets ..................................................... 83 Page 7. MAIN ELEMENTS OF A STRATEGY: PRE-REQUISrrES ....................... 85 The National Program to Fight Poverty ................................................. 85 Outline of a Strategy and its Context ................... .. ........................... 89 Pre-requisite: Macro-Economic Growth Policies .............................................. 91 Pre-requisite: Improved Social Services ................................................ 93 Pre-requisite: Better Governance ................................................ 94 8. MAIN ELEMENTS OF A STRATEGY: PRIORIIES ................................. 98 Land Tenure Reform and Access to Land ................................................ 98 Managing demographic Growth ................................................. 100 Decentralization ................................................ 103 Regional Action Plans ................................................ 106 Regional Development Funds ................................................ 109 Monitoring Poverty .................................................111 The Virtuous Circle ................................................ 113 ANNEXES Annex 1: Methodological notes on the Calculation of a Poverty Line Annex 2: Value Added for Main Food Crops, 1983-90 Annex 3: Methodological Notes on Food Security Calculations Preface This is the first Poverty Assessment for Togo (a Poverty Profile was issued in 1990). Its main purpose is to foster a dialogue that lays the foundations for a broad-based collaboration in the effort to reduce poverty. While the main targets are government officials and World Bank staff, it is hoped that the dialogue will involve a much larger audience, so as to harness all available resources for the fight against poverty. The approach adopted reflects this preoccupation. A rapid poverty appraisal was carried out with two Togolese teams to obtain a first-hand view of living conditions in poor rural and urban communities, and to learn from the poor about their own definition of poverty, their problems and their ideas of what would help them most. Representatives of NGOs (local and international) and donors were consulted to learn from their experience and to ensure that the interventions proposed would be complementary to theirs. A series of workshops held in January and February 1996 with the participation of government services, NGOs and associations active at the regional level was immensely useful to discuss analytical results, test hypotheses and, most of all, to develop a strategy. Indeed, the strategy presented in this document is nothing more than a compilation of the suggestions put forth by a vast array of Togolese. The premise for a fruitful dialogue is a shared understanding of the issues at hand, hence the attention paid to identifying different aspects of poverty and their causes. Because poverty is a multidimensional phenomenon, the analysis is based on a wide variety of sources, including both quantitative and qualitative data. Also, poverty is not simply a state, but an outcome that is produced and continually reproduced through the interplay of economic, social, physical and cultural factors. An attempt has therefore been made to capture its dynamic nature through an analysis of vulnerability. Neither the multi-dimensional nor the dynamic aspects of poverty, however, can be fully understood without considering the point of view of the poor themselves. Hence the importance of the fieldwork, and of the assistance received from those who work closely to the poor (extension agents, missionaries, social development agents, NGO workers) in furthering our understanding. In order to start a constructive dialogue on poverty reduction, the document adopts a style that leaves ample room to the descriptive and the didactic. This choice should be seen as an effort to reach as wide an audience as possible, including officials and colleagues who may have had until now little chance to study poverty, or to concentrate on aspects of poverty which go beyond their specialty sector. It is expected that future updates of the Poverty Assessment will take such background information for granted and adopt a narrower focus. This report was prepared by Maurizia Tovo (Task Manager, AF4PH). Rudo Niemeijer (Consultant) provided much valuable input, especially for Chapters 2 and 5, led one of the two local research teams, and coordinated all statistical work. Adje Sewavi (Direction de la Statistique) performed calculations for the poverty indicators, with the guidance of Didier Blaizeau (INSEE), and Emile van Rouveroy van Niewaal prepared an input paper on legal issues. Field work was carried out by Kangnivi Teko-Agbo, Koffi Hounkpe, Kwam Kisaoni Koko, Wilson Akovi, P. Antoine Solitoke, Tikpi Atchadam, Totetiebe Dametougle, Koffi Mondou, Akou Mana Klou, N. Yvonne Lawson, Ako Lassey and Elena Galliano. Comments and assistance were also received from: Antonella Bassani (AF4CO), Leila Zlaoui (AF4CO), Monique Garrity (AF4TO), Roger Key (AF4PH), Mark Blackden (AFTHR, Peer Reviewer), Olga Jonas (FRMRO, Peer Reviewer), Heidi Hennrich-Hanson (EA3CO, Peer Reviewer), Giuseppe Fantozzi (Cons.) and the Togo Country Team. The document was translated by Bernadette Etienne and processed by Fabrice Bonnaire. Executive Summary EXECUTIVE SUMMARY The Setting 1. During the 1980s, Togo undertook a wide-ranging economic adjustment effort aimed at correcting macro-economic imbalances. After an initial deterioration, real GDP grew during 1984-89 by 3.4 percent a year on average, but a continued high population growth rate (3.1 percent) prevented significant improvements in the overall welfare of the population. Producer prices for cash crops increased, but those for food crops fell. At the same time, per capita expenditures in health and education decreased as a result of decreases in overall budgetary resources. The resulting decline in the quality of health and education services, and the general deterioration of the economy, contributed to lower access to these services: both primary school enrollment rates and health consultation rates went down, while essential material (textbooks, drugs) was sorely missing. 2. During 1991-93 the country went through a period of political upheaval and economic deterioration. In November 1992, trade unions and the opposition parties called a general strike, leading to the paralysis of the modem sector. At the same time, the worsening security situation caused a sharp decline in domestic services and transit trade. The output of the secondary and tertiary sectors dropped sharply, foreign aid disbursements were interrupted, and real GDP declined by 6 percent in 1992 and 15 percent in 1993. People were hard hit by the crisis: GNP per capita fell by 23 percent, schools closed and medical services fell below acceptable standards. 3. Since the end of the general strike (August 1993), there have been significant positive developments on the political and economic front: presidential and legislative elections were held, and implementation of the economic restructuring program resumed. As a result of the improved political situation and the CFAF devaluation (in January 1994), the economy experienced a strong recovery, with real GDP growing by 14 percent in 1994 and 8 percent in 1995. Nevertheless, Togo's political environment and its economic situation remain fragile. The transition to democracy is proving troublesome, the budgetary situation remains extremely difficult, and foreign aid is way below what it used to be. In spite of the rise in GNP per capita to US$306 in 1996, the average Togolese is worse off than in the early 1980s. Classes have resumed, but under very difficult conditions, and access to health services remains problematic, especially in view of the increase in the price of medicines brought by the devaluation. Defining Poverty 4. Poverty is a multi-dimensional phenomenon which cannot easily be reduced to a few quantitative indicators. For example, Chambers' model of "deprivation trap" distinguishes five clusters of disadvantage: physical weakness, isolation, poverty (in the narrower sense of consumption levels), vulnerability and powerlessness. Hence, an assessment of poverty should take into account lack of access to basic social services Executive Summary ii (health and education) and infrastructure, the inability to secure a minimum consumption basket, as well as the social and political mechanisms which contribute to vulnerability and powerlessness. The analysis presented here starts by identifying the poor through the use of poverty lines derived from household consumption levels, but goes on to discuss other aspects of poverty as well. While an effort is made to quantify indicators to the greatest possible extent, qualitative information plays an important role in filling some of the gaps left open by quantitative data and in contributing to an understanding of the socio-cultural aspects of poverty. 5. The most recent household budget survey conducted in Togo dates back to 1987- 89 (Enquete budget-consommation or EBC). On the basis of such data, poverty lines were calculated as the minimum per capita yearly income necessary to cover expenditures for an average daily food intake of 2,050 calories plus some minimal non-food expenditures (clothing, housing, cooking fuel, etc.). Since there are important differences with respect to the cost of living and the expenditure patterns between the five regions and between communities with different degrees of urbanization, poverty lines were calculated separately for main urban centers, secondary urban centers and rural areas in each region. The resulting poverty lines vary greatly, from a minimum of 29,100 CFAF a year in rural Kara to a maximum of 55,900 in Lome; the national weighted average is 35,600 CFAF. The costs of covering only food expenditures is taken as the line to define core poverty; it goes from a minimum of 21,000 in the Plateaux to a maximum of 33,600 CFAF in Lome, with a national weighted average of 25,700 CFAF. 6. Poverty is very unequally distributed across the country. It is more widespread in rural areas, reaching 69 percent of the rural population in the Savanes and 57 percent in the Kara region, followed by the Maritime region with 45 percent. The lowest poverty headcount (i.e., the percentage of the population being poor) is in Atakpame (12 percent) and in secondary urban centers of the Plateaux (13 percent). Both the depth and the severity of poverty are greatest in the rural areas as well, and particularly in the Savanes and Kara, but also Lome has a high poverty gap index, indicating the existence of pockets of deep poverty in the city. 7. Since 1987-89, the country's economy has undergone considerable change (continued economic deterioration, the 1992-93 crisis, the 1994 CFAF devaluation). In the absence of recent household budget data, new poverty lines were estimated on the basis of changes in cost of living indices since 1989 as of December 1994, giving a national poverty line of 48,465 CFAF, and broken-down estimates going from a minimum of 27,889 for small towns in Kara to a maximum of 79,423 in Lome (a 1995 UNDP-financed survey gives a national poverty line of 58,000, which may reflect the steep increase in the cost of living during 1995). At the same time, farmers' incomes, estimated on the basis of agricultural production and producer prices, have not kept up, and in the Central and Kara regions they actually appear to be lower than in 1989. This is in line with what field work and a recent UNDP-financed survey suggest: since 1989, poverty has increased. Executtve Summary iii 8. Nutritional indicators provide a picture consistent with the one outlined above. Surveys in 1988 (the latest available) show a pattern of considerable regional variation, with malnutrition increasing from south to north: stunting among children 0-3 yr. old goes from 24 percent in Maritime to 42 percent in the Savanes, and the national average is 30 percent. As expected, malnutrition is more prevalent in rural areas (33 percent vs 21 percent in urban areas). Even when the average daily food intake is enough to cover basic caloric requirements, the instability caused by seasonal fluctuations is a matter of concern, for it means that a much larger number of people is below the poverty line during part of the year. Worse, it has lasting consequences, as it is a typical part of the poverty trap: due to seasonal scarcity, farmers have lower capacity for heavy labor at the time when the need for heavy labor is greatest. 9. Nutrition levels are also affected by fertility rates. The total fertility rate is 7.1 and has remained practically unchanged for the past two decades. This means that, especially in rural areas where fertility rates are higher, almost all women of reproductive age are either pregnant or lactating, and birth intervals tend to be short. Pregnant and lactating women are more likely to suffer from anemia than others, with poor women being at a higher risk because of their limited access to proper diets (42 percent of women of reproductive age are anemic). Because anemia diminishes productivity, pregnant and lactating women enter a vicious circle whereby poor diets result in lower productivity, which makes it impossible to improve diets. Children are caught in this vicious circle as well, as they depend on their mothers for nutrition either from breastfeeding or from cooked food. 10. Access to sufficient food is only one of the primary needs that should be considered when describing poverty. Other aspects, such as access to education and health services, are key indicators, because they measure people's opportunity to improve their living conditions by using the services provided in their country. Starting in the mid-1980s, rapid population growth and economic decline combined to produce a steady deterioration in all educational indicators: primary net enrollment rates went from 71 percent in 1980 to 60 percent in 1994, and student repetition and drop-out rates are among the highest in Sub-Saharan Africa (46 and 7.6 percent, respectively). There are significant and persistent gender inequalities and regional differences in school enrollment, which reflect poverty patterns. The Savanes stands out with the lowest enrollment rates (30 percent) and the greatest gender gap (girls' enrollment rates are half those of boys). Also the health sector has been affected by the economic decline, and access to health services reached an all time low during the 1992-93 crisis. Morbidity and mortality rates are in large part the result of an unhealthy environment (stagnating water, inappropriate waste disposal, improper house ventilation) and insufficient access to potable water (estimated at 59 percent). The major diseases are malaria, diarrhea and acute respiratory diseases; AIDS is rapidly spreading and sero-prevalence is estimated at 5 percent of the population. Executive Summary iv Vulnerability 11. Vulnerability is an important aspect of poverty, because it reflects its dynamic nature. It refers not to lack or want, but to defenselessness, insecurity and exposure to risk. Poor people are usually among the most vulnerable, but not all vulnerable persons are poor. 12. To understand what makes people in Togo vulnerable, it is useful to conceive of vulnerability as a function of assets: the more assets people have, the less vulnerable they are. In this context, assets should be considered in broad terms, as a wide range of tangible and intangible stores of value or claims to assistance which can be mobilized in a crisis. Thus assets include stores (e.g., jewelry, money, granaries), concrete productive investments (e.g., farming and fishing equipment, animals, tools, land), human investments (education and health), collective assets (irrigation systems, wells), and claims on others for assistance (kinship networks, friendships, savings clubs, patrons, credit). An awareness of the diverse nature of assets, and of their hierarchy, is essential for meaningful policy action. It helps to identify the most vulnerable individuals and communities, thus facilitating effective targeting. It also makes it possible to plan interventions before vulnerability turns into absolute poverty, therefore using a preventive rather than reactive approach, and it provides guidance on how to organize assistance in a way that builds on the assets of the recipients, so as to minimize the need for future interventions. 13. Vulnerability can be determined by factors operating at different levels. To facilitate the analysis, these factors have been grouped into three categories according to their level (individual/household, community and national) and to their nature (natural, social and economic). They are summarized below. Micro Level Meso Level Macro Level (Individual & Household) (Community) (National) Natural * poor soils, lowfertility * poor soils, lowfertility * natural disasters (droughts); Factors (degradation and erosion) * natural disasters, (floods, * population pressure; droughts and parasites); * limited natural resources * difficult access to safe water basis Social * poor health (malaria, * insufficient access to health * insufficient & inefficient Factors malnutrition, handicaps, old services (excessive fundingfor primary education; age); cost/distance); * timidfamilyplanning * high dependency ratio (high * inadequate water supply and programs; fertility rates 'rmany sanitation; * insufflcientfundingfor dependents, and perfunctory * limited access to education primay health care services, contributions from male (excessive cost/distance); (drug shortages, poor quality "breadwinners" o*few * insufficient access tofamily care); providers); planning services; * discriminatoryfamily law * intrahousehold inequality, * cultural values & tradition (*women's inferiority); (children treated as property of (about gender differences, the * geographic inequality in their father's family, women role of children, attribution of resource allocation for social considered inferior to men); prestige and authority, services. * low education, (especially of agricultural and health Executve Summary v Micro Level Meso Level Macro Level (Individual & Household) (Community) (National) women). practices). Economic * low returns to labor (low * isolation/remoteness (difficult * poorly integrated marketfor Factors yields, low or unreliable marketing, lower farm gate foodproducts (localized food producer prices, low wages, low prices, reduced access to social shortages, high seasonal margins of profit in informal services); variations in prices, low activities); * lack ofemployment producer prices, high consumer * unemployment and under- opportunities (under- and prices); employment; unemployment, compressed * poor rural infrastructure * inter-household inequality, wages); (higher marketing costs, difficult (control over land, rights & * land shortages in parts of the delivery of social and extension duties related to social standing) Savanes, Kara and Maritime services); * unequal access to productive regions; * low development offinancial resources by gender (women * inefficient agricultural intermediation systems; denied equal access to production systems (land use, * unclear land tenure system household resources, such as agricultural practices and access (low incentives to sustainable land, means of transport and to agricultural inputs). agricultural practices, conflicts, working equipment); inefficient use of natural * lack of access to credit. resources). 14. Because individuals are generally part of a household and live in a community, their personal level of vulnerability will be determined by the interaction among the categories of factors identified above. This is why attempts to identify vulnerable groups for targeting purposes are unlikely to produce mutually exclusive categories. Typical vulnerable groups include women (especially pregnant and lactating) and children (especially below five years of age). Here the attention is given to two groups which have been selected not on the basis of their greater vulnerability or their larger size, but because they are specific to Togo and they are generally not included among the targets of interventions to prevent or reduce poverty. * Displaced families. A fundamental feature of displacement is that it causes a significant and sudden unraveling of existing patterns of social organization, with a profound impact on the structures of economic and cultural life. Production systems are dismantled, informal solidarity networks are rendered non-functional, and traditional authority and rights may be questioned. Togolese families have been displaced on four types of occasion: (a) forced eviction for the creation of national parks, during the 1970s and 1980s, when an estimated 60,000 people lost their land, which was declared state property, and their homes, which were often destroyed by the military, without receiving any form of compensation; (b) government-sponsored resettlement, such as in the FED-Agbassa project and in the Nangbeto dam; (c) displacement linked to the exploitation of phosphate mines in the Vo district, where eleven villages were forced to resettle elsewhere while their lands were taken over by OTP; (d) violent civil strife in 1992-93, during which about 173,000 northemers living in the south were forced to return to their land of origin. * Girls on their own. As a matter of survival, most children in Sub-Saharan Africa must contribute to their own upkeep by working, but there are cases in which they end Executive Summy vi up struggling for their survival alone. Street children are one such case, but the majority are girls from poor rural families sent to work in urban households as domestic servants. They are estimated to be as many as 50,000 in Lome alone, mostly from south-east Maritime, where demographic pressure is greatest. While in some cases their life in the city may be an improvement over the abject poverty in their family of origin, in other cases they end up over-worked, under-paid (if paid at all), and at the mercy of their employer. Increasing the Poor's Productivity: Factors of Production 15. Land. As almost 80 percent of the poor are farmers, improving the poor's productivity (and therefore their income) depends on their access to potentially productive land. First and foremost, productive land has to be available for exploitation, hence the need to address issues of rural population density and land scarcity. Second, the poor must be given the right to exploit it and enjoy the product of their labor, hence the need to address land tenure issues. 16. One way to assess whether there is enough land available for the poor to cultivate is to estimate the rural density threshold, which is the maximum number that the land can support without fertility loss (urban areas are excluded for obvious reasons). While at present there are overcrowded areas where environmental degradation is severe (e.g., north east Kara, Savanes, Maritime) and areas considered underpopulated (e.g., districts of Bassar in Kara, Nyala and Tchaoudjo in Centrale), it appears that the rural density threshold is quickly being reached throughout the country and, by the year 2000, all regions but one (Kara) will have a rural population that cannot be sustainably supported by the environment under current agricultural practices. 17. The problem of land scarcity is exacerbated by a complex land tenure system in which traditional and modem law coexist and compete. Customary law is regulated by three principles: (a) land belongs not to an individual but to a collectivity, (b) the right to exploit the land is given by the collectivity to an individual and can be maintained as long as the land is used, (c) those who are not part of the landowning clan may be given a right to exploit the land (droit de culture), but their right can be revoked at any time and certain crops are not permitted. Thus, non-landowners ("foreigners" or metayers) are in a very vulnerable situation because they are at the mercy of landowners for access to land and will never be able to own land. Women as well generally have no secure rights and cannot own land. 18. Attempts to establish a modem land titling system have given very limited results, as the 1974 reform links land tenure security to exploitation rather than ownership, thus providing no incentive for titling. In any case, the registration procedure is long (twelve months at the minimum), complex (presupposes knowledge of the different bureaucratic steps involved) and expensive (a case study found the cost of the registration to be three times the value of the land to be registered), so that obtaining an official land title is out of the question for the majority of farmers. Executive Summary vii 19. Labor. In 1983, agriculture employed 72 percent of the labor force, the informal sector 21 percent and the modem sector only 7 percent. More recent data are not available, but it is reasonable to estimate that today about 70 percent of the population is active in the primary sector, while the informal sector has grown, especially considering that public sector employment has shrunk and that many of the formally employed are forced by their low salaries to have part-time informal activities. It is estimated that at least 80 percent of the Lome labor force is active in the informal sector. 20. The official minimum wage (SMIG) is set at 13,757 CFAF per month, one of the lowest in the CFAF zone and well below the poverty threshhold for a household of average composition with no other income sources. While only a small portion of formal sector employees have a monthly wage as low as the SMIG, the majority of informal sector employees earn considerably less than the minimum wage. For example, girls working as domestic servants are paid as little as 2,000 CFAF a month (plus board and lodging) and shop assistants are paid between 6,000 and 10,000 CFAF. Scattered evidence suggests that daily salaries for unskilled work start at around 200-300 CFAF, with "female" jobs being paid less than "male" jobs. The typical daily rural wage is reported to be 500 CFAF, but there are considerable regional differences. 21. While firm data on unemployment are unavailable, 1991 estimates placed the rate of unemployment at 16-28 percent. Since then, the situation has certainly worsened because of the economic crisis, especially in urban areas (unemployment in Lome is believed to reach 40 percent). Given high fertility rates, demographic pressure on land and very limited opportunities for off-farm employment, continued rural exodus is likely to cause unemployment rates in urban areas to keep rising. At the same time, it is becoming increasingly difficult for school leavers to find work. Despite the 1991 employment training program (Programme Emploi-Formation), which employed about 2,700 graduates, there are now some 21,000 unemployed graduates whose only realistic hope to find a job is to enter the informal sector. 22. Capital: Access to Financial Services. Access to credit is important for financing investment and, in turn, for boosting the productivity and income of the poor. In addition, it is essential to reduce vulnerability because it may help the poor cope with unexpected crises (death, disease, crop failure). Indeed, information gathered in the rapid poverty appraisal leaves no doubt that having access to credit is a priority for most people, especially to buy fertilizers or as start-up capital for trading activities. 23. Credit is available through three main channels. * Formal banking is used almost exclusively to finance the modem sector and is largely inaccessible to poor (and not so-poor) people because of collateral requirements, loan size and geographic location (only in regional capitals). * Intermediary financial institutions, such as savings and loans cooperatives, have been set up, often with the assistance of donors and NGOs, to satisfy the credit need left Executive Summary viii unanswered by the commercial banking sector. Approaches and structures vary considerably, but they all operate close to the grassroots, provide short-term credit, are willing to lend small amounts, build on traditional mechanisms, and aim for long- term sustainability through the mobilization of local savings. The oldest and largest savings and loans cooperative is FUCEC. Informal credit systems represent the only access to financial intermediation for the vast majority of the poor. The preferred sources are (richer) relatives and neighbors; additional sources include itinerant bankers, usurers and other people who, because of their work, dispose of a certain liquidity and come into contact with low-income groups, such as traders and bush-taxi drivers. Interest rates are often high, but for poor borrowers interest rates tend to be far less important than quick access to cash. Savings clubs (tontines) are common, but field work suggests that in the poorest areas (Savanes, Kara) they are not as common because few can afford saving anything. Rural Development and Food Security 24. During the 1 980s producer prices for cash crops (coffee, cocoa and cotton) were raised. These increases substantially augmented the real incomes of cash crop producers. T'hey also generated a considerable supply response for cotton, which can be grown throughout the country except in the Maritime region. Indeed, cotton growing came to represent the main chance to escape poverty for farmers in the impoverished northern regions. In the same years, trade of crops was liberalized, but the impact on subsistence farmers' income is likely to have been small: many of the poor are not surplus producers and, even in the presence of incentives, do not have the means to respond because of lack of time (especially women, who are the main food producers), labor, capital (credit de campagne) and technical skills. Besides, they tend to sell their crops at local markets where few traders operate in an oligopolistic environment which is fairly impermeable to national conditions. 25. In 1986, the government unveiled a New Rural Development Strategy whose objectives were, broadly speaking, in line with poverty reduction concerns: center development actions on the small rural producer, strengthen production support structures, increase and diversify output, improve resource allocations, and protect the environment with active community participation. Subsequent policy pronouncements have been in line with these objectives, and the 1995-97 Public Investment Plan underlines the Government's intention to continue its disengagement from direct intervention in the productive sector in favor of support to producers and rehabilitation of rural infrastructure. The following constraints, however, make the realization of these policy objectives problematic: * budget allocations do not reflect the importance of agriculture in the Togolese economy, and allocations for non-salary recurrent expenditures are about half as much as what would be needed for proper delivery of rural development services; Executtve Summwy ix * extension services are weak: with the exception of cotton, they have not resulted in higher productivity or better practices, as the advice provided is seldom timely, practical or relevant, and women needs in particular tend to be ignored; * agricultural research does little to address farmers' concems; * the excessive centralization of decision-making, on the one hand, demotivates and disempowers regional staff, and on the other, results in decisions taken without understanding regional conditions and therefore often inappropriate; lack of coordination among different departments and rigid hierarchical structures inside departments make things worse. 26. Farmers' concems were summarized at the national conference of 1992 as fellows: (a) weak extension services, (b) irrelevant agricultural research, (c) insufficient access to credit, (d) unfavorable terms of trade, (e) uncertain land tenure, (f) insufficient transport and storage infrastructure, and (g) problematic access to water. Recent field work confirms that farmers' perceptions have not changed since 1992. In particular, access to credit and to fertilizers is considered of paramount importance. 27. Food Security. The government's long-standing pledge to achieve food self- sufficiency has been pursued through the establishment of Togograin, a state marketing agency to buy and distribute food crops, and farmers' incentives such as tax exemption on agricultural machinery and agricultural subsidies for fertilizers. In theory, food self- sufficiency has been achieved since the late 1970s, except during drought years. But there are important regional differences and annual variations. For example, while root crop production always exceeds requirements, cereal production in most years shows a deficit. Luckily, this shortage is generally balanced by a surplus production of cassava and yams, but the Savanes, where cereals contribute almost 70 percent of all calories, is particularly vulnerable to harvest failures. 28. Regardless of whether the country as a whole, or each region separately, is able to produce enough food to feed everybody, the real test for food security is at the household and individual level. It matters little that enough food has circulated within the country in a given year if there are entire communities where food is either unavailable or unaffordable for months. In this sense, the length of the pre-harvest "hungry" period (soudure), and the eating patterns during this time, are better indicators. The length of the soudure is mostly determined by climatic and soil conditions. In the Savanes and parts of the Kara, where rainfall tends to be erratic and soils severely depleted, the soudure lasts about five months (April-August) and food insecurity is reported by farmers and NGOs to be a major problem. In the Plateaux and Centrale regions, where soils are fertile and rainfall regular, periods of widespread food scarcity are rare, while in the Maritime region there are pockets of chronic and pervasive food insecurity (e.g., Vo district) due to land pressure. In conclusion, there appears to be little doubt that, in spite of official estimates about national food self-sufficiency, food security remains an elusive target for a sizable E&ecutive Summary x part of the population and government interventions have thus far had limited success in alleviating this problem. Social Services 29. Health. Good health is important for all, but for the poor it is especially important because their income tends to depend exclusively on their physical labor, and because they lack the assets to help them cope with periods of ill health. Investments to reduce health risks among the poor and to make basic health care accessible even to those with limited financial means, therefore, are an important element in a poverty reduction strategy, especially considering that the poor's only access to modem medicine tends to be through government services. In 1978, the government adopted a primary health care strategy giving priority to peripheral services, which resulted in considerable improvements in health status and access to health. In 1990, a National Health and Social Sector Policy reiterated government commitment to health for all, but the economic and political crisis got in the way. The sector's performance continues to be hampered by the following key constraints: * inadequately articulated priorities and strategic framework, which result in ad hoc programs, serious duplication and waste of resources; * insufficient funding, especially for non-salary operating costs; . poor collaboration between ministries and with other public institutions, which aggravates the problems caused by budgetary limitations; * structural constraints at the service delivery level, with no outreach program to ensure that the most vulnerable groups benefit from available services. 30. Geographic inequality is a major problem: most human, financial and material resources are concentrated in the Maritime region (particularly in Lome), while the Savanes trail far behind the other regions. Health indicators mirror this situation, with infant mortality rates in Lome half the national average and child malnutrition in the Savanes twice the national average. Rural areas tend to be under-serviced compared to urban centers. 31. Health problems are consistently mentioned by the poor as a main concern, but public health services appear to be significantly under-utilized--only 23 percent of the population uses them. One reason is perceived high cost, including transport, consultation fees, "motivations" for health personnel and, especially, medicines. In addition, the expected low quality of the service offered acts as a disincentive (particularly unavailability of drugs). So poor people resort to self-medication, or go to traditional healers. Traditional treatments not only may cost less than modern ones, but can be paid for in kind and in installments--no upfront cash is ever required. Should public health structures provide services more in line with the means of poor customers, Executive Summw,y xi utilization rates would probably increase, thus making more efficient use of government funding. 32. Education. After a period of growth from 1975 to 1980, the education sector entered a crisis from which it has not yet recovered. The net primary enrollment rate was 61 percent in 1994-95, down from 71 percent in 1980, while student repetition rates have been increasing and educational achievement levels declining. Two main factors are at the root of this crisis: * rapid population growth, which makes it extremely difficult to offer adequate schooling opportunities to increasingly large cohorts of students and manifests itself most obviously in class overcrowding (up to 200 children in urban first grade classrooms); * a combination of economic deterioration and policy choices which mean that education expenditures have not kept up with school age population growth not only because of sheer lack of money but also because other areas have been judged more important than education in the allocation of scarce budget resources. 33. The problems created by insufficient resources are exacerbated by disparities within the sector. In 1995, the government allocated almost 50 times as much public money for a tertiary level student than for an elementary school student. In addition, the recruitment freeze of civil servants in place since 1984 has resulted in acute teacher shortages which are particularly felt in remote rural areas, where people have had no choice but to pay the teachers out of their own pocket. So the poor, who tend to live in remote areas, end up having to pay more for their education than others (e.g., in rurl Savanes 41 percent of teachers are financed locally, against 17 percent in Lome). 34. Cost is one of the main reasons why the poor do not send their children to school. Wihile school fees are low, school supplies are expensive. In addition, there is a high opportunity cost, especially for girls, whose domestic labor is often perceived by parents as much more valuable than the benefits to be gained from schooling. Yet, returns on expenditures for educating women are higher than for men, because of the positive association between education on one side and health and nutrition on the other. Efforts should therefore be concentrated on improving female educational levels. Safety Nets 35. Formal responsibility for assistance to the most vulnerable people falls mainly on the Social Affairs Directorate, presently in the Ministry of Women in Development and Social Affairs. The main instrument to achieve the Directorate's goals are Social Centers (135 in total) located in the district capitals and in some of the largest villages, and staffed with APS (social development agents). Most Social Centers, however, are under-staffed, have almost no operating funds and no transport for outreach activities--indeed, it is not difficult to see why morale among the APS is low. In an effort to complement the Executve Summary xii activities of the Social Affairs Directorate and to raise much needed funds, in 1992 the government created the National Solidarity Agency, but its limited means and unfocussed mandate make one question the reason for its continued existence. In conclusion, it is clear that without substantial increases in resources (and a surge in staff motivation), government agencies will not be able to make much difference in the welfare of vulnerable groups. 36. The assistance gap left by the government is filled in part by NGOs, including religious organizations. Their role is becoming increasingly important, especially for the delivery of integrated rural services at the grassroots level, and NGO expenditures in 1994 were over 3.9 billion CFAF (higher than govermment expenditures for rural development). In the end, however, the vast majority of the Togolese have no choice but to rely on traditional safety nets. Extended families are the most powerful safety net, but their effectiveness should not be overestimated: very poor people are likely to have very poor relatives unable to offer much help, and modernization promotes an individualistic and materialistic philosophy of life which is the opposite of the principles from which the extended family derives its strength. The assistance provided by the extended family is often complemented by community-based associations, such as tontines, which function either as systems of labor sharing or as savings clubs. Main Elements of a Strategy: Pre-Requisites 37. In 1995, the Ministry of Planning, with 11NDP support, started a process aimed at defining a National Program to Fight Poverty (PNLP). The output is a very comprehensive program which, after presenting a diagnosis of the causes and manifestations of poverty, envisages three sub-programs: (a) employment creation and revenue distribution to benefit the rural and urban poor, with components focusing on increased rural production/productivity and urban employment; (b) improvement of social services for the poor, with components on health/nutrition, housing/water/sanitation, and basic education/literacy/vocational education; (c) accompanying measures, with components on population dynamics, women's status and social mobilization. Each sub-program contains a number of specific projects for a total estimated cost of US$572 million for 1996-2000. 38. The strategy proposed here builds on the PNLP and complements it by drawing attention to a limited set of issues and by adding a regional dimension. It is based on a bottom-up, participatory process including consultations at the grassroots level and a series of workshops with regional government and NGO representatives. 39. Macro-Economic Growth Policies. Because poverty and vulnerability are widespread, a development strategy must coincide to a large extent with a poverty reduction strategy, and significant poverty reduction can only be envisaged in the presence of sustained economic growth. The government target is a 6 percent annual growth in GDP for the 1995-97 period. To this end, Togolese authorities should pursue the implementation of the comprehensive programn of stabilization already under way Executive Summary xiii aimed at clearing domestic and external payment arrears, rebuilding the tax base, and containing expenditures. The success of this program, in turn, is dependent on public enterprise reforms that would strengthen public finances and increase the scope for private activity. Measures to improve the regulatory framework (e.g., the revision of investment and labor codes) would be an important complement to encourage private sector development. 40. From a sectoral perspective, the most promising growth areas are: (a) agriculture, where increased productivity and non-traditional crops (e.g., horticulture) should be encouraged; (b) service activities, where transit trade has the main role to play and should be facilitated by improvements in physical infrastructure; and (c) small-scale manufacturing (e.g., agro-processing, textile products), where the CFAF devaluation has increased competitiveness. The reorientation of the country's scarce public resources from non-development expenditures (mainly military) to essential social services and productive activity will also be critical. 41. Improved Social Services. Broad-based economic growth is necessary to increase the income-earning opportunities of the poor, but some will be unable to take full advantage of them because of ill-health, malnutrition, illiteracy and lack of skills. Ensuring access for the poor to basic social services of reasonable quality is therefore essential. Although the preferential treatment accorded to social expenditures from the late 1980s is an indication of the government's commitment to providing social services to all its citizens, it is also true that social indicators are far from satisfactory. Sectoral analyses suggest two main areas of concern: * Financing requirements. A public expenditures review for the education and health sectors revealed important financing gaps. As donor assistance is unlikely to return to previous levels in the medium term, Togo will have to count primarily on its own resources. Additional internal revenues, however, are most unlikely to be able to finance requirements (cost recovery plays a small role and pushing it much further could hurt the poor). * Efficient use of available funding. Because substantial budget increases cannot be expected, the only way to improve the poor's access to social services is to use better whatever resources are available. A number of measures in this sense are presented in the National Program to Fight Poverty. In general, it is a matter of streamlined procedures and transparent management on the one hand, and of addressing imbalances on the other--between tertiary and primary services, between preventive and reactive approaches, between regions, between urban and rural areas, and between salaries and operating costs. 42. Better Governance. Good governance has been shown to be closely linked to long-term, sustainable economic growth. A number of donors have raised concerns about the pace of Togo's progress on governance issues, making improved governance (and respect of human rights in particular) a condition for resumed--or increased--aid. While Executive Summary xiv progress toward better governance could take place in many ways, three areas are particularly important. * Bureaucratic Accountability. Some civil servants see their jobs primarily as a way to secure a salary, rather than as a public service to which their fellow citizens have a right, and present criteria for promotion provide little incentive for high-level performance. Suggested measures include: streamlining of procedures, delegation of authority, decentralization, an information & education campaign on the "bureaucratic" rights of citizens, random checks, and the establishment of simple mechanisms to address users' complaints. * Economic Transparency and Efficiency. The 1996 public expenditure review has demonstrated the government's intention to improve public administration. Among the many encouraging developments, five appear most promising: (a) a new budgetary nomenclature has been adopted, which will result into greater transparency and a consolidated budget; (b) procurement and disbursement procedures are under exarnination in view of simplifying them; (c) the accounting procedures and internal control systems of the Treasury are being strengthened; (d) the preparation, execution and monitoring of the Public Investment Plan will be improved; (e) the public enterprise sector will be made more efficient and accountable. * Improved Functioning of the Judicial System. The modern legal system is in a state of crisis: personnel is insufficient, the operating budget is inadequate, and legal procedures are extremely slow. Although in 1978 the judicial system was unified, it remains plagued by ambiguities which allow powerful individuals to manipulate the system, while the poor and uneducated may find themselves lost--and therefore more likely to lose. Reforms to make the system more accessible are therefore needed. Those articles in the Family Code which discriminate against women (e.g., inheritance, child custody) should be arnended. Main Priorities of a Strategy 43. Land Tenure Reform and Access to Land. Land tenure issues are a priority because land is a fundamental asset for the 80 percent of the poor who live in rural areas and because failure to address them will have disastrous long-term consequences for the environment. The 1974 land tenure reform did not succeed in improving the situation primarily because it did not take into enough account customary law, and thus it was neither applicable nor applied. Efforts to improve access to land for vulnerable people (especially women and youth) should therefore take tradition as the starting point and be based on negotiations among interested groups at the community, district and regional levels, in a bottom-up fashion. New, more secure forms of access to land should be introduced, such as leases for several years, with government and NGOs helping vulnerable groups to get organized for better bargaining power. Reassuring landowners of their rights would facilitate innovative arrangements. To encourage a more equitable Executive Summary xv and efficient use of land, incentives should be established for land-rich areas to receive migrants (e.g., social infrastructure for receiving communities). 44. Managing Demographic Growth. Unchecked demographic growth will not only magnify environmental problems, but make it impossible for the government to provide acceptable levels of social services, while unemployment and underemployment will continue to grow. Togo does not yet have a population policy, so population matters are still influenced by policy decisions which lack a coherent and realistic approach. A central element of the population policy should be efforts to limit fertility rates and promote responsible parenthood. These would include: (a) pervasive information and education campaigns for the population at large, as well as more targeted campaigns for adolescents and for men, highlighting the problems linked to reduced birth spacing, and early and late motherhood; (b) training of health personnel and staff from the Ministry of Women in Development and Social Affairs in family planning; (c) support to and coordination of all public and private institutions involved in the promotion of family planning activities; (d) the immediate repeal of the 1920 contraceptive laws; (e) a campaign to discourage early marriages and the pursuit of bridewealth (la dot) as an instrument to mobilize emergency funds. 45. Fertility rates are closely related to women's status. Indeed, the most important variable determining fertility behavior is women's education. Also, if women derive their social value mainly from their maternal role, as it is the case in most of Togo, there will be few incentives for women to try to limit family size. Efforts to control population dynamics, therefore, would have to be complemented by measures to improve women's status. These include: increasing female education, reinforcing women's and girls' self- esteem, giving women equal rights to men (e.g., reforming the Family Code) and informing them of their rights. 46. Decentralization. The 1992 constitution stipulates that the country is organized in collectivites territoriales following the principles of decentralization while respecting national unity. Up to now, however, deconcentration--rather than decentralization-- continues to be the approach in place. The nomination in November 1995 of a Junior Minister (Ministre delegue') for Decentralization augurs well, but resources allocated for decentralization purposes have been meager and progress made in transferring power to regional centers has been limited. Yet, decentralization has a crucial role to play in the fight against poverty, because it is the key to improved delivery and better targeting of public services. Keeping in mind that there is room for considerable variation in the appropriate design and form of a decentralization process, the following four areas of action should be considered: * Improve the institutional framework by creating a viable bottom-up structure and by clearly defining the responsibilities, obligations and prerogatives of central, regional and local authorities. Social capital should be strengthened through the establishment of committees at the local, departmental and regional level where civil society and Executive Summaiy XVi government would work together. Each region should have autonomous deliberating and executive bodies. * Strengthen regional institutions by giving them the authority to identify projects to be financed by the central government. Village Development Committees (CVDs), Local Planning Committees and Regional Planning Committees already exist, but in many instances they are dormant or are perceived as imposed by outside political interests. Some NGOs, however, are working with CVDs and it is conceivable that such institutions could be revitalized, depoliticized and empowered. * Mobilize local finances, as decentralization by definition means that localities take over both responsibilities and the taxing power or financial resources needed to meet them. New revenue sources should be identified, and old ones strengthened. * Encourage decentralization of staff by providing detailed descriptions of (new) regional posts, by making regional rotations an integral part of civil servants' carriers, and by providing incentives to work outside Lome. Fostering the development of regional capitals is a longer-term and much more costly goal, but its benefits would go far beyond attracting competent civil servants. 47. Regional Action Plans. A regional approach would have two main advantages: it would facilitate decentralization and a bottom-up approach, and it would make it possible to tailor interventions in a way that is responsive to the priority problems of each region, thus encouraging a more selective and efficient use of public resources. Each region should therefore produce a regional poverty reduction plan. To this end, an initial consultation with grassroots communities, regional government staff and NGOs has produced a series of recommendations which, despite important differences among regions in relative importance and specific solutions proposed, can be grouped under nine themes: (a) regulating access to land and protecting vulnerable groups (e.g., farmers expropriated by OTP); (b) stopping soil degradation mainly through village land management schemes (gestion du terroir) and the use of organic fertilizer; (c) improving access to water with additional water points and small irrigation; (d) improving food security and food crop production by promoting cereal banks (in the north), better resource management, diversification and access to agricultural inputs; (e) containing the negative impacts of traditional ceremonies by working with traditional authorities; (f) making credit available to low-income groups (farmers, artisans), combining it with savings facilities; (g) facilitating transport through better road maintenance and the abolition of road blocks; (h) making school more useful for the poor by improving both its quality and its relevance; (i) improving access to modem medicine by implementing the Bamako Initiative and making generic drugs widely available. 48. Regional Development Funds. If regional action plans for poverty reduction are to become a reality in the near future, a mechanism has to be established whereby funds to carry out recommended actions are directly available at the regional level, not just to government agencies but to all institutions that have been identified as capable of Executive Summary xvii assisting in implementation. At present, even a concerted effort to decentralize decision- making and give regional government structures direct access to public finances would not represent an adequate solution, because non-governmental structures would be excluded. When resources are scarce, as in the case of Togo, the challenge is to create a system capable of using effectively--and equitably--contributions coming from as many sources as possible (communities, NGOs, professional associations, traditional authorities, private businesses) and in as many forms as possible (knowledge, creativity, technical skills, labor, capital). 49. One option deserving further exploration could be to establish five Regional Development Funds (one per region). Similarly to social funds, they would work in a participatory and bottom-up fashion, financing the construction/repair of small social infrastructure and assisting in the establishment of productive projects. Their scope of action, however, would be broader. In addition to traditional social fund activities, they would finance a wide range of activities capable of improving the welfare of the poor in the region--such as fairs to bring business to the area, marketing campaigns for local products, regional broadcast services--so as to encourage creativity in finding solutions to the problems identified, and infuse a dynamic and entrepreneurial spirit into the fight against poverty. They would also be free to use whichever intermediary they choose to deliver services and goods, as long as the guiding criteria are competence and reliability, and to do their own fund-raising (in this sense, funds provided by the central government would be considered as seed money). An important advantage of Regional Development Funds would be that they could not only be an instrument of decentralization, but facilitate geographic targeting as well, since funds could be allocated to different regions as a function of both their population and their relative poverty (e.g., using a poverty weight). 50. Monitoring Poverty. Policies and programs designed to reduce poverty require a clear understanding of the nature and extent of the problem. Once policies have been adopted and interventions are ongoing, information is critical to evaluate their effectiveness and fine-tune their design. An accurate measurement of poverty, therefore, is an investment. Occasional large scale surveys are important, but effective monitoring demands the establishment of data collection systems. Also, for cost-effective monitoring the Government should take advantage of all existing data sources, including routine administrative records, information gathered by NGOs and studies carried out by training institutes and university departments. 51. Because knowledge is essential for effective action, the following recommendations are for the short term: (a) a comprehensive household budget survey should be carried out to obtain data on household consumption and intra-household allocations, using a multi-disciplinary approach; this survey should be repeated every five years (estimated cost: US$600,000); (b) national-level nutritional indicators should be collected on an ongoing basis (estimated cost: US$90,000); (c) light and focused surveys should be carried out at least once a year to update and complement data from the household budget survey, using both qualitative and quantitative data (estimated cost: Executive Summary xviii US$70,000); (d) socio-economic analysis should be complemented by the analysis of cultural variables which are likely to play a major role in shaping vulnerability as well as determining the appropriateness and effectiveness of poverty alleviation interventions (estimated cost: US$70,000). Togo: Overcoming the Crisis, Overcoming Poverty 1. THE SETTING Background 1.1 Togo is a small country (56,785 kM2) with a population of 4.2 million and a per capita GNP of US$310 (1996)'. In 1995, agriculture accounted for 35 percent of GDP and employed about 75 percent of the labor force, mostly as smallholders using traditional cultivation practices. The agricultural sector is also a key source of foreign exchange earnings. The main export crops are cotton, coffee and cocoa (28 percent, 7 percent and 3 percent of total merchandise exports, respectively, in 1995). Secondary output which accounts for some 23 percent of GDP is dominated by phosphates production and also comprises a small but dynamic manufacturing sector composed of micro, small and medium scale enterprises. Togo's service sector, consisting mainly of commerce and transit activities contributes the remaining 42 percent of GDP. Over a third of the population lives in the southern Maritime region where the capital city is located and rural to urban migration is high, with Lome (about 640,000 inhabitants in 1994) growing at an estimated rate of 6 percent per year. The population growth rate of 3.1 percent is one of the highest in the world and results in a population density of 72 inhabitants per kM2, greater than that of most West African Countries. 1.2 Despite its small size, the country is characterized by great diversity. Agro- ecological conditions are markedly different in each of its five regions and the presence of some forty different ethnic groups underlines such differences. Three principal ethnic clusters dominate (see Figure 1.2): the Adja-Ewe in the south (41 percent), the Kabye- Tem and Losso in the center-north (23 percent) and the Para-Gourma in the north (14 percent). 1.3 Thanks to its abundant supplies of low-cost phosphate, a favorable climate for agriculture and a traditionally dynamic tertiary sector (especially commerce), Togo has high potential for sustained economic growth. The period between independence l1960) and the mid-1970s, during which the economy grew at a rate of 7 percent per year, is a clear indication of such potential. Prudent economic policies, high international prices for coffee and cocoa, the rapid expansion of phosphate mining and an open economy favoring commerce contributed to give Togo the reputation of being one of the best African economic performers. Encouraged by a sizable primary surplus, the Government embarked in the mid-i 970s on an ambitious public investment program to modernize and industrialize the economy, primarily through the creation of public enterprises. When the price of phosphate, cocoa and coffee fell, and the price of investmnent-related imports rose well beyond initial projections (e.g., oil), the Government started relying increasingly on external borrowing. The economic situation rapidly deteriorated, and public debt servicing went from 1 I percent of government revenues in 1973 to 76 percent in 1981. I GNP per capita is calculated with the Atlas method. Togo: Overcoming the Crisis, Overcoming Poverty 2 02! : Figuret1.1: Administrative Map of Togo Figure 1.2: Main Ethnic Groups of Togo The 190s Stutua Adjstmn Efots 1.4 During the 1980s, Togo undertook a wide-ranging economic adjustment effort .. .... .... .. aiming at coretingmcoeooi maacs Btwe 199ad 98,th Government implemented an IMF-supported financial stabilization program that fcs on demand management measures. The program was, however, unable to restorA macrecoraln ta 4 prc aear 1980-83 period (DseeFigure 1.3) while real per capita incomes fell by almost 7 percent a = j | E E: E E 7 E E ' 1 :, ' :!! !.y, .!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. ..... .. { tAtA 2 ( / .r t:t.i. -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. .......... DA=I 8i Mii g \ _ ] "~~~~~~~~~~~~~~~~WT _r;=D~~~~~~~~~~~~~~~~~~~~~~~~~O /IE E The 198~~0s Struturl Adutmn Effrt 1.4 During the 1980s, Togo undertook a wide-ranging economic adjustment effort aiming at coffecting macro-economic imbalances. Between 1979 and 1982, the Government implemented an IMF-supported financial stabilization prograrn that focused on demand management measures. The program was, however, unable to restore macroeconomic stability and real GDP declined by more tha 4 percent a year during the 1980-83 period (see Figure 1.3) while real per capita incomes fell by almost 7 percent a Togo: Overcoming the Crisis, Overcoming Poverty 3 year. A marked decrease in access to social services due to constraints in public expenditures further lowered people's standard of living. 1.5 To reverse the economic and social deterioration, the Figure 1.3: Real GDP Growth, 1980-93 Government implemented a more comprehensive recovery program that was supported by three IDA- financed structural adjustment credits and by IMF standby and SAF arrangements. Overall, 0 En% structural adjustment efforts aimed to encourage financial restraint and to promote economic growth in the medium to long-term in three so, 321314 ION5 I7UU 0n o 2e ways: (a) improving public sector management and investment programming systems, (b) rehabilitating viable public enterprises and closing/liquidating non-viable ones, and (c) improving the incentive system within the economy. Major accomplishments under the adjustment program are summarized in Table 1. 1. Table 1.1: Togo's Reform Agenda in the 1980s ~~~~~~~~~~~~-_=r. . . !.;t ; ->-- ;r v. - .. ... Incentive and regulatory framework in agriculture * liberalization of food crops (prices, exports and imports) a) food crops (20/o of GDP) b) export crops (3% of GDP and 25% of exports) Incentive and regulatory framework for non. abolition of state monopoly on imports agicultural activities * elimination of export and import licensing * elimination of most price controls Size and Efficiency of public enterprises (PE) * privatization and liquidation of 30 PE out of a total portfolio of 73 * adoption of institutional and regulatory framework for public enterprises a adoption of legislation for privatization Efficiency of the tax system * simplificarion of the tax system by the introduction of income tax and of a general sales tax * elimination of export duties and taxes * reduction in the number and dispersion of tariff rates Efficiency of public expenditures * introduction of a rolling three-year investment progrnam * adoption of procurement code * annual review of level and structure of budgetary allocations for priority sectors 1.6 StructuralAdjustment and Growth. These initial adjustment programs helped the country achieve a substantial liberalization of its economy and succeeded in restoring economic growth despite an unfavorable external environment characterized by low export commodity prices and an appreciation of the real exchange rate. After three years of negative growth, real GDP grew by 3.4 percent a year on average during the 1984-89 period, and private investment more than doubled, increasing from 4 to 10 percent of GDP. Per capita GNP reached US$ 428 in 1989 compared with US$ 250 in 1984. Togo: Overcoming the Crisis, Overcoming Poverty 4 However, in the absence of a clear government commitment to address population issues during the 1980s, the country's population growth rate remained at a high 3.1 percent preventing significant improvements in the overall welfare of the population. While available data do not allow an in-depth assessment of the impact of structural adjustment on the poor, some broad inferences on changes in rural incomes and access to basic social services are presented below. 1.7 Rural production and incomes. A marked increase in producer prices for export crops generated considerable income increases for cash crop farmers. Producer prices for coffee, cocoa and cotton were raised by 29, 20 and 25 percent, respectively, in real terms with the result that incomes of coffee, cocoa and cotton farmers together were estimated to have increased 97 percent in real terms between 1982 and 19862. While these benefits were mostly limited to farmers in the Plateaux and Central regions, the spread of cotton cultivation contributed to bringing much needed cash income also to traditionally poorer areas in the Savanes and Kara regions (the number of cotton growers more than tripled between 1982 and 1987, going from 43,000 to 132,000 households). Nevertheless, it has been pointed out that the reduction of fertilizer and pesticide subsidies, and the drop in international prices during 1988-89, may have partly eroded such gains. 1.8 The 1986 liberalization of food crop marketing, which gave farmers access to markets in neighboring countries, represented a potential for increased monetary income for many low-income families. However, while production of the main crops increased substantially in 1985-90 (cassava 12 percent, yam 8 percent, maize 58 percent and sorghum 30 percent), producer prices of some of these crops fell sharply over the same period (cassava -24 percent, yam -15 percent, maize -12 percent, but sorghum +50 percent). The net result was a 10 percent increase in the output value of these crops from a three-year average of CFAF61.6 billion in 1983-85 to CFAF67.6 billion in 1988-90 (see table lA.1 in annex 2). 1.9 Social Services. The impact of structural adjustment on social services delivery seems to have been uneven. In fact, regaining financial and economic balances required cuts in public expenditures. The majority of these cuts were done through a reduction in capital expenditures by eliminating many of the unprofitable projects started under the Government investment spree of the mid-70s, but spending on the social sectors decreased in real terms. Social sector expenditures as a share of the recurrent budget were at an average 23 percent before the adjustment process began (1980-82) and remained basically unchanged during the first two adjustment programs (1984-85). However social sectors were clearly accorded a preferential treatment within the overall budgetary constraints during the implementation of SAL III and SAL IV resulting in a 29 percent average share of the recurrent budget going to social expenditure in 1988-92 (see table IA.2 in annex 2). 2World Bank, Poverty profile of Togo - Country Assessment and Policy Issues, report no. 5519-CM, December 1991. Togo: Overcoming the Crisis, Overcoming Poverty S 1.10 In spite of a shift in intra-sectoral allocations in favor of primary education, net primary school enrollment rates fell from 72 percent in 1980 to an all time low of 53 percent in 1984, slightly rebounding to 63 percent in 1988. Both drop-out and repetition rates increased significantly. It is difficult to determine to which extent this drop in educational achievement should be attributed to the fiscal stabilization and to which extent it should be attributed to the general deterioration of the economy in the early 1980s. But it is certain that, with sectoral expenditures declining in the face of a rapidly growing school-age population, the quality of education suffered. 1.11 Access to health services also appeared to have become more difficult during the 1980s. While personnel levels were maintained, expenditures on materials and investment were reduced considerably during the first half of the decade. The ensuing allocation of resources designated the bulk of all expenditures to personnel, leaving about 20 percent for essential materials. As a result, what minimal equipment existed at health facilities was often found in disrepair, special programs such as maternalchild health care or extended immunization were unable to function properly, and drug supply became erratic and insufficient (particularly in rural areas). To compound the problem, the generalized drop in per capita income in the first half of the decade and the rising costs of health care are thought to have contributed to a disturbing downward trend in the rate of consultations, which fell from 83 percent in 1978 to 67 percent in 1986.3 Starting in 1988, relative allocations to non-wage expenditures recovered substantially to reach 27 percent of the recurrent budget by 1990. 1991-93: Political Upheaval and Economic Deterioration 1.12 In the face of mounting Figure 1.4: Real GDP by Sector, 1980-93 pressure for a multi-party (constant CFAF billions) system, the Govemment convened a national conference in April 1991. The conference appointed a transitional '° government which included the opposition. The inception of the 60.00 transitional government and its 40 ------------- institutions, however, was undermined by violence, to the point that in 1992 Togo's three I i I I main bilaterial donors (France, I o SO , .. Germany and the USA) suspended their aid programs. 3 The rate of consultations is defined as the number of individuals who go to a health center for consultations as a percent of the total population. This inforrnation is reported in: World Bank, op.cit. (1991). Togo: Overcoming the Crisis, Overcoming Poverty 6 Indeed, the period between 1991 and 1993 was marked by violent street riots and mass civil disturbances, which were accompanied by economic and social deterioration. 1.13 In November 1992, trade unions and the opposition parties called a general strike which lasted nine months, leading to the paralysis of the modem sector. At the same time, the worsening security situation caused a sharp decline in domestic services and transit trade, as foreign investors and local businessmen alike sought refuge abroad while the general population reduced to a minimum the time spent outside their home. As a result, output of the secondary and tertiary sectors dropped by 35 percent and 38 percent respectively between 1991 and 1993 (see Figure 1.4). By contrast, agricultural production rose by about 11 percent over the same period, largely because of favorable climatic conditions (the 1992-93 harvest was exceptionally good) and because most secondary and tertiary sector workers fled the cities and turned to agricultural activities as a survival strategy. 1.14 Given the depressed level of economic activity and the increased informalization of transactions, public finances deteriorated seriously. During the 1992-93 period, revenues fell by more than 52 percent and investment expenditures by about 60 percent; by the end of 1993, the overall fiscal deficit (excluding grants) had reached 15.2 percent of GDP. With the interruption of foreign aid disbursements, the deficit was partly fimanced through the accumulation of domestic and external payment arrears. 1.15 People were hard hit by what they generally referred to as the "events" (les evenements). GNP per capita fell from US$431 in 1990 to US$354 in 1993. The educational system was brought to a halt for six months (1992-93 school year) during which school facilities suffered serious vandalism, theft and neglect. The situation in the health sector worsened considerably, making good quality health services unattainable to all but the wealthy. If a decade of austerity during the 1980s had reduced equipment and medical supplies to minimal levels, the events of the early 1990s meant that services fell below acceptable standards; for example, vaccination coverage for the 0-1 year old group fell to less than 40 percent. 1994-95: The Recovery Starts 1.16 Since the end of the general strike (August 1993), there have been significant positive developments on the political and economic front. Presidential and legislative elections were held in August 1993 and in February 1994, respectively. President Eyadema was reelected for a five-year term and the two opposition parties (UTD and CAR) obtained a slight parliamentary majority. In June 1994 a new government was formed which, although not a coalition one, is considered by most observers as an important step toward democracy. On the economic side, the implementation of the economic restructuring program, which had come to a halt during 1991-93, has slowly resumed. In January 1994, the Government reinforced the adjustment effort by devaluing, in collaboration with the other CFA zone countries, the CFA franc vis-a-vis the French franc. Togo: Overcoming the Crisis, Overcoming Poverty 7 1.17 Developments in the Real Sector. As a result of the improved political situation and the CFAF devaluation, the economy experienced a strong recovery in 1994-95. Real GDP growth reached 14 percent in 1994 and 8 percent in 1995 reflecting a solid recovery of secondary and tertiary activities from the depressed levels of 1992-93 and the effect of the CFAF devaluation on the production of traditional export crops. Producer prices for cotton and coffee were increased by 60 and 75 percent respectively, boosting production of cash crops by 33 percent in real terms. Following a 29 percent upward price adjustment in the two months after the devaluation, monthly inflation returned to moderate levels. 1.18 Despite these positive developments, Togo's political environment and economic situation remain fragile. The country is still in the midst of a difficult transition from an authoritarian regime to a well functioning democracy. Since basically all data collection stopped in 1990, it is difficult to determine exactly who suffered most from the deterioration and who is benefiting from the recovery. The improvement of the situation in 1994-95 still leaves the average Togolese worse off than before the crisis and the general impression obtained from field work and related by observers is that, in spite of the increase in economic activity, there has been little improvement in the life of the poorest segments of the Togolese population since the all time low of 1993. Although real GDP per capita increased by 13 percent over the last two years following the crisis, its 1995 level (CFAF98,000) is still far below its pre-crisis level (CFAF1 13,000 in 1991).4 1.19 Public Finance. As a result of the crisis, Togo's revenue base had shrunk drastically. Government revenue as a share of GDP fell from 22 percent in the late 1980s to 11 percent in 1993. Several measures were introduced to broaden the tax base and improve revenue collection. Although economic and fiscal revenue performance in 1994- 95 has met and in some cases exceeded program targets, the budgetary situation remains extremely difficult. Some foreign aid has officially resumed, but it is still way below what it used to be (US$86 million in 1994 against US$261 million in 1990) and the pace and level of disbursements remain slow and largely conditional on future progress on the political front. In 1994, the shortfall in foreign aid reached 16 percent of the budget. The tight budgetary situation did not allow for any significant improvement in the quality and coverage of social services. 4 GDP per capita is estimated in constant 1987 CFAF. Togo: Overcoming the Crisis, Overcoming Poverty 8 1.20 Social Services. Classes resumed in 1994, but under very difficult conditions: educational materials are lacking, many teachers are no longer being paid and instructional leadership has been greatly weakened because of drastic cuts in inspectors' budgets. Large numbers of students were forced to abandon school in December 1994 because their fees had not been paid (see Box 1.1). 1.21 Following the alarming Box 1.1: Kodjoviakop6 Elementary School deterioration in access to health "During the troubles people run to Ghana, because we are registered during the first half of close to the border. Some went to their village. Then the 1993, a number of donors and NGOs children started coming back, some last year and more this joined their efforts with the year. But the problems are enormous: we have many social goveniment' to re-establish a problems, many people who cannot pay school fees. Out of government's to re-establish a 346 students, only half have paid their fees. I have not yet minimum of service. UNICEF's sent home any of the children, because I feel bad for them, Bamako Initiative provided much but in other schools it isn't the same and many children have needed respite, but, although the been sent home. I know their parents don't have money, and Initiative has been successfully I sacrifice myself to try to satisfy them. But the situation is very difficult. In CPI [ffrst grade] we have 102 students with , one teacher, no chalk, not enough desks [children are three or once the initial allocations of four to a desk], the cahier de classe only arnved in January medicines run out, it proved and the children cannot afford either books or notebooks. In impossible to replenish them. While the meantime the school is falling apart and there is no money by 1996 the situation has greatly to fix it." improved, in many health centers Source: Rapid Poverty Appraisal, elementay school principal in essential drugs are still difficult to Lomr, February 1995. obtain. Following devaluation, the cost of pharmaceutical supplies rose, rendering them less affordable to many people and leaving no other recourse but to turn to traditional remedies. Hospitals and health centers lack the most basic equipment and material, which were either stolen or damaged during the period of greatest turmoil and which, given the present budgetary constraints, cannot be easily replaced. Shortages and instability of trained personnel have further affected the quality of health care delivery. 2. A PROFILE OF POVERTY Defining Poverty 2.1 A widely accepted definition of extreme poverty is that the extremely poor are those who cannot secure enough nutrition to perform adequately. People who are undernourished are more vulnerable to disease, tend to develop anthropometric deficiencies, may become brain-damaged and lethargic and, in general, are less able to lead a healthy life with sufficient energy to carry out productive tasks (i.e., labor market, domestic or educational activities) in a satisfactory manner. In this sense, extreme poverty is absolute and, in principle, individual-specific, as the nutritional needs of Togo: Overcoming the Crisis, Overcoming Poverty 9 individuals change with age, sex and life conditions. The measurement of extreme poverty is best done through indicators of nutritional status, such as the incidence of nutritional illnesses (e.g., marasmus), chronic malnutrition (e.g., stunting) or low birth weights. 2.2 Beyond this rock-bottom definition, poverty becomes difficult to determine, and therefore to measure, since poverty is a multi-dimensional phenomenon which can not easily be reduced to a few quantitative indicators. Debates on poverty tend to be conducted in terms of material deprivation as measured by income, consumption or expenditure. Such measures have the advantage of enabling comparison between individuals and households, but poverty as experienced in poor rural and urban communities is not simply a matter of low incomes and expenditures. For example, Chambers' model of the "deprivation trap" distinguishes five clusters of disadvantage: physical weakness, isolation, poverty (in the narrower sense of consumption levels), vulnerability and powerlessness5. Field work recently conducted in Togo confirms that the poor tend to perceive their disadvantage along similar lines. Hence, an assessment of poverty should take into account lack of access to basic social services and infrastructure, the inability to secure a minimum consumption basket, as well as the social and political mechanisms which contribute to vulnerability and powerlessness. 2.3 The analysis presented here starts by identifying the poor through the use of a poverty line derived from household consumption levels, but goes on to discuss other aspects of poverty as well. While an effort is made to quantify indicators to the greatest possible extent, qualitative information plays an important role in filling some of the gaps left open by quantitative data and in contributing to an understanding of the socio-cultural aspects of poverty. The following chapters complete the analysis by examining the causes of poverty. Throughout the analysis, an effort is made to give voice to the poor by including their own assessment of their situation. A Poverty Line' 2.4 In order to estimate the distribution and depth of poverty, it is generally considered acceptable to use real per capita expenditure as a proxy for income. A yardstick is then needed to determine who is poor and who is not and, for this purpose, a poverty line is drawn as a cutoff point in the expenditure distribution below which individuals are considered poor. The choice of a poverty line is a matter of preference. Here the poverty line is calculated taking as the starting point the income needed to satisfy basic nutritional needs (see Box 2.1 and Annex 1). This is estimated on the basis 5 Chambers, R., Rural Development: Putting the Last First. London, New York: Longman, 1983. 6 Aitnex I provides details on the methodology used and explains the rationale behind the approach selected. Togo: Overcoming the Crisis, Overcoming Poverty 10 of the most recent household budget survey, which dates back to 1987-89.7 Given the important changes that occurred in the country since then, the poverty lines based on these survey data are only approximations of the current situation, and a separate analysis is required to evaluate the outcome of the poverty analysis in light of recent developments, which include the effect of the devaluation in early 1994. This analysis was accomplished by estimating changes in the cost of living since the 1987-89 survey as well as changes that are likely to have occurred in the income levels of some of the poorer sections of the population. Box 2.1: Calculating a Poverty Line The first step in the calculation of a poverty line was estimating a minimum caloric intake per capita, taking into consideration the specific conditions of Togo. This was necessary because the basic nutritional needs of an individual vary according to age, sex, weight, activity patterns and health status; for women, additional factors are pregnancy and lactation. Since not all necessary data were available to obtain an independent estimate of the nutritional needs of the Togolese population,' it was decided to approximate these needs by using a value close to it--that of Benin, where conditions are very sirnilar. Three different levels were used in the calculations, since energy requirements (which vary according to the amount of physical activity) are generally lower in an urban environment than in a rural environment. Using the widely accepted 1985 World Health Organization recommendations, it was assumed that, for the population of the major urban centers, the average person would require a minimum energy intake of about 1,990 calories per day to maintain reasonable health and performance standards. For the rural population and the population of secondary towns (where many also engage in agriculture), the corresponding values are 2,080 and 2,000 calories per day, respectively. The cost of purchasing this amount of calories was then calculated, making the necessary adjustments to reflect different dietary habits in different parts of the country (see Annex 1). In urban households, food expenditures of poor households amounted to approximately 60 percent of the budget, while budget percentages for food were 70 percent in secondary towns and 80 percent in rural areas. Using this factor to calculate the income needed to cover non-food expenditures gives the poverty line. 2.5 Using the data of the 1987-89 household budget survey (Enquete budget- consommation or EBC), poverty lines were calculated for different parts of the country (see Table 2.1). Because there are important differences with respect to the cost of living and the expenditure patterns between the five regions and between communities with a different degree of urbanization, the poverty lines resulting from these calculations differ considerably. In Lomd the poverty line is CFAF 55,900 or US$1759 per capita per year; in Kara (city), on the other hand, the poverty line is only CFAF 22,900 or US$72. Further north, in Dapaong, the cost of living is higher than in Kara due to price differences and the relatively high cost of a diet based on sorghum and millet. A national poverty line, which can be calculated by taking a (weighted) average of the regional poverty lines, falls at CFAF 35,600 or US$112. 7Enquete Budget Consommation (EBC) 1987-1989. See Annex I for a detailed report on the methods used to calculate a poverty line on the basis of this survey. 9 Data are lacking with regard to the health situation, certain anthropometric measurements, and the physical activity levels associated with different occupations in Togo. The anthropometric data collected in the EBC where not sufficiently precise for this type of calculation. Togo: Overcoming the Crisis, Overcoming Poverty 1I Table 2.1: Poverty Lines and Poverty Incidence, 1987-89 Cost of caloric Poverty line % with per capita % with per capita requirements (CFAF) income lower than cost income lower than (CFAF) of caloric requirements poverty line Urban* Core Poor All Poor Lome (Maritime) 33,600 55,900 3.8% 16.9% Atakpame (Plateaux) 25,200 42,000 7.5% 12.0% Sokode (Centrale) 22,700 37,800 2.7% 17.6% Kara (Kara) 17,800 29,600 3.7% 14.8% Dapaong (Savanes) 21,400 35,700 5.8% 23.9% Small Towns** Maritime 23,400 33,400 9.3% 30.5% Plateaux 21,100 30,200 6.7% 13.00/o Centrale 21,700 31,000 2.5% 17.9% Kara 16,000 22,900 11.2% 26.4% Rural Maritime 28,300 35,400 25.0% 44.9% Plateaux 21,000 26,300 4.0% 12.1% Centrale 23,700 30,000 11.8% 25.4% Kara 23,300 29,100 38.1% 56.8% Savanes 26,400 33,000 47.7% 68.7% Average for Togo 25,700 35,600 17.3% 32.3% *The incidence of poverty in urban centers is likely to be underestimated because the poorest households have no fixed abode and because household servants are "hidden" in richer households. **Sample size for small towns in the Savanes region was too small to allow for meaningful estimates. Source: staff calculations based on EBC. 2.6 For about half of all poor persons (i.e., about 17.3 percent of the total Togolese population) poverty takes on an even more extreme form, because these persons live in households with a per capita income that is even insufficient to afford them the absolute minimum of calories to ensure proper child growth and maintenance of their health. These people are defined as core poor. The number of persons living in such circumstances is much higher among the rural population than among the urban population. Particularly in the north (Savanes and Kara), a large part of the rural population has expenditures (including own consumption) below this extreme minimum, which, since this mostly concerns farming households, means that they are unable to grow sufficient food to feed their families. This explains why, even though the food security situation of the country as a whole may not be alarming, malnutrition rates in the north are still very high. The Maritime region as well, with its dense population, has a high percentage of inhabitants with very low per capita incomes. 2.7 On average, a minimum per capita yearly income of CFAF 35,600 or US$112 is therefore necessary to cover expenditures for a daily food intake of 2,050 calories plus some minimal non-food expenditures (e.g., clothing, housing, cooking fuel, etc.). Incomes below these amounts are not enough to satisfy basic requirements, meaning that people living at this level will have to face very hard choices such as obtaining treatment 9The exchange rate used is CFAF 319 = US$I, which was the rate in 1989. Togo: Overcoming the Crisis, Overcoming Poverty 12 for a sick child or buying fuel to cook dinner. But poverty still exists above the poverty line. As the key social indicators discussed below illustrate, access to social services is low throughout the country, implying that some of the people with incomes above the poverty line could still be considered poor on the basis of their inability to afford basic social services. In addition, a considerable portion of those above the poverty line find themselves in a precarious situation whereby they are able to keep afloat under "normal" circumstances, but their lack of assets makes them vulnerable to external shocks, such as diseases or crop failures (see next chapter). The role of the poverty line, therefore, is not so much to set aside those who do not need poverty reduction interventions as to identify those who absolutely need such interventions. 2.8 Table 2.2 presents estimates of the poverty indices (the P-alpha indicators) associated with the poverty lines presented in Table 2.1 (last column). The P0 (poverty headcount) column shows that 32.3 percent of the population (or 1,268,744 persons) can be considered absolutely poor. Poverty, however, is not distributed equally over the country. The Plateaux region, with its capital Atakpame, has less poor people than the Centrale region, followed by Maritime, Kara, and Savanes, in that order. Also, the percentage of poor people in urban communities is lower than that of poor people in secondary towns, while rural areas show the highest incidence of poverty. Plateaux, with its cash crops and higher agricultural potential is an exception in this regard: poverty there is about evenly distributed over all types of communities. By contrast, the difference in poverty incidence between urban and rural communities is largest in Kara, with only 15 percent poor in Kara town and almost 60 percent poor in its rural communities. It should be further noted that the percentage of poor in Lome is about half of that of the country as a whole. Figures of this kind most vividly explain why there is a persistent trend of rural-urban migration (see para. 4.19), which only reversed during the recent periods of civil unrest because of safety considerations and the collapse of the urban economy during the strike. 2.9 For policy purposes, it is important to know not only how many people are poor, but how poor they are. The poverty gap index (P,) is used for this purpose, that is, to measure both the incidence and the depth of poverty. In other words, the poverty gap index estimates the distance between per capita expenditure and the poverty line for each individual (the so-called proportional shortfall). Unlike the headcount index (P0), it is not a figure which lends itself to an immediately intuitive interpretation, but it is valuable for comparing the relative poverty of different geographic areas, thus making it possible to target interventions toward those in the greatest need.'" For example, the poverty gap index of rural Savanes is about twice that of the country as a whole, and the poverty gap index of rural Kara region is only slightly smaller. This means that these two regions not only have the highest prevalence of poverty (poverty headcount), but also that those who 10 The poverty gap index is also the basis for other interesting calculations. By dividing PI by Po, for example, one obtains the average distance of the poor from the poverty line. In our case, the average poor person has an income which is 31 percent below the poverty line, which means that the average yearly income of the absolutely poor in 1987-89 was CFAF 24,573 (US$77). Togo: Overcoming the Crisis, Overcoming Poverty 13 are below the poverty line tend to be poorer than elsewhere--about twice as poor as the average poor. Table 2.2: Poverty Indices, 198749 P-alpha Indicators Po P, P2 Urban Lome 0.169 0.113 0.076 Atakpame 0.120 0.091 0.070 Sokode 0.176 0.098 0.057 Kara 0.148 0.108 0.079 Dapaong 0.239 0.074 0.030 Small Towns* Maritime 0.305 0.070 0.024 Plateaux 0.130 0.035 0.016 Centrale 0.179 0.024 0.005 Kara 0.264 0.074 0.029 Rural Maritime 0.449 0.109 0.036 Plateaux 0.121 0.019 0.005 Centrale 0.254 0.054 0.018 Kara 0.568 0.176 0.072 Savanes 0.687 0.203 0.080 Average for Togo 0.323 0.100 0.045 * Sample size for small towns in the Savanes region was too small to allow for meaningful estimates. Source: staff calculations based on EBC. 2.10 P2 is an index of the severity of poverty, and shows the relative inequality of the distribution of income among the poor. Similarly to PI, it is useful in targeting interventions because a high P2 indicates the existence of special sub-groups or niches of extremely poor persons who can be considered poor among the poor. A comparison of figures in Table 2.2 shows that such groups are typically found in a number of urban environments (Lom6, Atakpame and Kara) and in the rural areas of the North (Savanes and Kara). In both environments, there are cases of extreme poverty: households or individuals who even among the poor stand out as especially disadvantaged. Most likely, as suggested by recent fieldwork, these are households of disabled and/or elderly people (often widows), who are not capable of earning sufficient income in a difficult urban environment or in rural communities where there is little possibility to subsist when the efforts required by agriculture in a harsh climate are beyond their capacity. Characteristics of Poor Households 2.11 It is clear that the absolute poverty line defined above is useful for identifying the most desperately poor, but by no means includes all poor. Because under normal circumstances food security is only a localized problem (the country as a whole is self- sufficient), measuring poverty on the basis of access to food results in a relatively low Togo: Overcoming the Crisis, Overcoming Poverty 14 incidence. For many low-income people the real problem is not so much securing sufficient calories on a daily basis, but having enough resources to be able to cope with unexpected downturns. Disease is probably the most common of these, and its impact is all the more devastating when it strikes the most productive members of the household. Extreme climatic variations may cause harvest failures, thus exacerbating seasonal fluctuations to the point of extreme hardship, as reflected in the high acute malnutrition rates (30-40 percent) of toddlers found in the northern part of the country (see Table 2.8). 2.12 Table 2.3 Table 2.3: Poverty and Sex of Household Head presents an analysis of the incidence of Core Poor Poor poverty according to Male Female Male Female the sex of the Headed Headed Headed Headed household head. In all Maritine (incl. Lom) 13.1% 15.8% 16.4% 18.0% Plateaux 3.8% 6.9% 9.2% 2.7% regions but the Centrale 9.5% 5.6% 12.3% 26.9% Plateaux female- Kara 29.8% 33.1% 16.6% 21.2% headed households are Savanes 43.4% 37.0% 20.3% 27.1% more likely to be poor Togo 17.7% 15.4% 15.0% 15.1% than male-headed Source: staff calculations based on EBC. households, but in the Plateaux and Savanes male-headed households are more often extremely poor. This statistical picture is not uncommon. Although female-headed households are often very poor, this may not emerge clearly from the available statistics. One reason is that per capita incomes or expenditure do not always present the most accurate measure of relative poverty: female-headed households, because of their smaller size, do not profit from economies of scale in the same way as (larger) male-headed households do. Also, because of their combination of relatively small earning power with a small household size and high dependency ratio, female-headed households find it harder to cope with emergencies. Finally, female-headed households are not a single homogeneous category. Some of them are actually enjoying incomes from male labor migrants, or are households of female traders who earn substantial incomes. Togo: Overcoming the Crisis, Overcoming Poverty 15 2.13 Outside the main cities (Lomt, Table 2.4: Poverty and Occupation of Household Head in Lome Atakpame, Sokode, Percentage of Households Kara, and Dapaong), Occupation Of Household Head Core Poor Poor Non Poor the poor are usually Inactive 4.90/o 12.6% 82.4% subsistence farmers, Technical, Teaching & Professions 0.0% 2.3% 97.7% while most --although Administrative & Commercial Staff 3.4% 7.6% 89.0% certainly not all-- non- Sales of Agricultural Produce 0.0% 13.6% 86.4% Sales of Drinks 6.2% 10.6% 83.2% poor tend to have Non-Food Sales 1.5% 4.9% 93.6% other non-agricultural Domestic Employees & Gardeners 0.0% 8.5% 91.5% activities as primary Other Services 3.0% 22.5% 74.5% or secondary Agriculture 11.9% 1.2% 86.9% occupations. This Food Preparations 3.3% 23.2% 73.5% Tailor 0.7% 0.7% 98.7% reflects the better Laborer 0.0% 17.2% 82.8% income earning Driver 0.0% 6.4% 93.6% opportunities open to Non-Agricultural Laborer 2.3% 7.3% 90.4% the non-agricultural Armed Forces 6.4% 15.2% 78.4% occupations. In Source: staff calculations based on EBC. Lomr, poverty is most common arnong street food vendors, who are predominantly women, and other service providers (mostly working on an ad hoc basis), while extreme poverty is most prevalent among farmers, barmaids ("sale of drinks") and the armed forces (see Table 2.4). As expected, those in technical occupations, teaching, and the professions are usually earning enough to remain above the poverty line, as are most people involved in commerce. Changes since 1987-89 2.14 The poverty lines presented in the previous section are representative of the situation in Togo around 1987-1989. Since then, the country's economy has undergone considerable change, first because of the long decline linked to an over-valued currency, then because of the political crisis and the ensuing disruption of economic activities, and lately because of the CFAF devaluation. In 1995, UNDP financed a survey to collect updated information on poverty conditions," but methodological choices make such data generally unsuitable for a comparison with the EBC-based estimates presented above (see Annex 1). It was therefore preferred to obtain an indication of how the incidence of poverty has changed during this period by estimating variations in the cost of living and in household income. 2.15 Although during the period of greatest turmoil prices fluctuated considerably, mostly in response to changing levels of demand and population movements, by early 1994 (that is, just before the CFAF devaluation), prices had more or less stabilized and Apte Afrique, "Actualisation de l'approximation quantitative de la pauvrete au Togo" (UNDP Project TOG/89/006/A/06/99), Lom6, 1995. Togo: Overcoming the Crisis, Overcoming Poverty 16 most were still very near the 1989 levels or slightly below. This minor downward trend suddenly reversed when the devaluation of the CFAF made most imported goods far more expensive overnight. Official producer prices for the major cash crops were raised by about 30 percent, but producer prices for locally produced foods, after an initial period of instability, stabilized with a fairly small increase (see Table IA.3 in Annex 1). Taking 1989 as the base year (1989=100), by December 1994 the cost of living index for the country as a whole was 136, meaning that the average household budget, as determined during the 1987-1989 EBC, needed to be multiplied by 1.36 to ensure a standard of living similar to that enjoyed in 1987-89. This cost of living increase was not homogeneous: in urban areas it was greater than in rural areas, and in the south it was greater than in the north (see Table 2.5). Hence, it varies from a maximum of 142 in Lome, where the consumption of imported goods is greatest as a result of the higher standard of living, to a minimum of 111 in rural Kara, where people tend to consume what they produce locally. The poverty lines in current prices for 1994, presented in Table 2.5, were calculated to reflect the increase in the cost of living adjusting for the typical consumption bundles as determined during the 1987-89 EBC. Table 2.5: Poverty Lines adjusted for Price Increases (in CFAF)'2 1989 1994 Cost of Living Index December 1994 Poverty Lines of Household Budgets Poverty Lines Core Poor Poor (1989 is 100) Core Poor Poor Urban Lome 33,564 55,932 142 47,661 79,423 Atakpamn 25,236 42,048 141 35,583 59,288 Sokode 22,692 37,824 129 29,273 48,793 Kara 17,760 29,592 129 22,910 38,174 Dapaong 21,432 35,724 123 26,361 43,941 Small Towns* Maritime 23,376 33,396 137 32,025 45,753 Plateaux 21,120 30,168 135 28,512 40,727 Centrale 21,708 31,020 126 27,352 39,085 Kara 15,996 22,860 122 19,515 27,889 Rural Maritime 28,320 35,400 136 38,515 48,144 Plateaux 21,012 26,268 133 27,946 34,936 Centrale 23,700 29,952 125 29,625 37,440 Kara 23,280 29,100 111 25,841 32,301 Savanes 26,388 32,976 113 29,818 37,263 Average for Togo 25,699 35,636 136 34,951 48,465 *Sample size for small towns in the Savanes was too small to allow for meaningful estimates. Source: Staff calculations based of market statistics collected by the Direction de la Statistique and DESA combined with EBC expenditure data. 2.16 The situation of the poor after the devaluation is difficult to determine without a complete budget survey. Since most of Togo's population depends on farming, farm 12 The method of calculation for this table is presented in Annex 1. Togo: Overcoming the Crisis, Overcoming Poverty 17 incomes provide a proxy measure to estimate how people's standard of living was affected by price changes. Using agricultural production statistics and producer price market survey data, a comparison can be made between farmer incomes of 1989 and 1994 (see Table 2.6). In general terms, farm incomes have not kept up with the rising cost of living due to the devaluation: even in comparison with the relatively low increase in the cost of living in rural communities, farm incomes have been decreasing in relative terms. In Centrale and Kara, regions with relatively little cash crop production and distant from the major urban markets, the decrease is considerable, because farm incomes were decreasing even in absolute terms. Only in Plateaux, the rising coffee prices, especially during the last months of 1994, kept up with the increased cost of living, but the lower food crop prices still resulted in a net loss of purchasing power. With the exception of the Plateaux, average farm incomes were below the poverty line in 1989 and still were in 1994. Table 2.6: Comparison of Per Capita Farm Incomes, 1989-1994'3 1989 1994 Change (1989=100) Fanner Cash Crop Farner Cash Crop Farmer Cash Crop Cost of living Incomes* Incomes* Incomes* Incomes* Incomes Incomes Index Maritime 32,212 949 34,512 0 107 n.a. 136 Plateaux 47,490 13,365 53,499 20,109 113 130 133 Centrale 47,853 1,725 33,927 2,035 71 118 125 Kara 25,504 1,420 22,416 659 88 46 112 Savanes 35,503 1,542 37,021 1,723 104 112 113 In CFAF. Source: Staff calculations based of market statistics (Direction de la Statistique), EBC, and data available from DESA and OPAT. 2.17 The extent of the loss of purchasing power among urban low income groups (e.g., street food sellers, traders) is difficult to establish because in the urban areas the economy virtually collapsed during the political crisis. Large parts of the population left the cities, particularly Lome, in response to the crisis, and only by the end of 1994 there were initial signs of economic recovery. Data from the 1995 UNDP-financed survey indicate that in Lome average per capita expenditures went from CFAF 136,479 in 1987 to CFAF 165,328 in 1995. This means that, setting 1987 as the basis at 100, per capita expenditures raised to 121, while the price index for the same period raised to 145, thus suggesting a general impoverishment of the population. Expenditure Patterns and Poverty 2.18 Table 2.7 shows the expenditure patterns for different income levels. As one would expect, the share spent on food increases as incomes go down, going from about 50 percent for the non poor to about 80 percent for the core poor. Most of the resources available to the poor for non-food items are spent on the other two most basic necessities- 13 See Tables IA.3 and IA.4 in Annex I for detailed calculations Togo: Overcoming the Crisis, Overcoming Poverty 18 -clothing and housing. Predictably, housing expenditures are considerably higher in urban areas than in rural areas, where people can build their own huts using locally available material. Once the most basic needs are addressed, the next largest expenditure items tend to be traditional ceremonies, especially in rural areas. Although using scarce resources for social events rather than for, say, health or education may appear an irresponsible behavior, from the point of view of the poor it can be a highly rational choice, provided the expenditures do not become a source of long-term indebtedness. In fact, generous spending for traditional ceremonies (to the point of conspicuous consumption) is a way to gain prestige and to reinforce one's ties to the community, which in turn will make it easier to obtain assistance in case of need. Hence, ceremonial expenditures may be understood as investments to build social assets and decrease vulnerability (see next chapter). Table 2.7: Expenditures and Poverty Core Poor Poor Non-Poor Type of Community Type of Community Type of Community Category Rural Small Towns Urban Rural Small Towns Urban Rural Small Towns Urban Food 79.0/ 85.1% 70.4% 78.5% 76.2% 63.9% 65.9% 56.4% 45.7% Housing 3.2% 3.1% 9.9% 3.4% 3.3% 11.1% 5.9/o 10.5% 16.9% Clothing 7.60% 3.3% 9.3% 7.3% 10.5% 11.5% 9.2% 11.7% 13.3% Trasport 0.9%/ 0.8% 1.1% 1.2% 1.7% 2.2% 1.9%/. 3.5% 6.6% Ceremonial 3.0%/ 2.5% 0.6% 3.4% 1.3% 1.8% 4.2% 3.7% 2.5% Health 1.6% 1.5% 1.8% 1.7% 1.7% 2.4% 3.1% 2.3% 2.4% Hygiene 1.6% 1.2% 1.6% 1.4% 1.4% 2.0% 1.7% 1.6% 2.4% LeIsure 0.9/ 0.0°/. 0.8% 0.6% 0.1% 0.9°/ 1.3% 0.7% 1.6% Education 1.4% 0.8% 2.8% 1.4% 3.1% 2.5% 1.1% 1.4% 1.2% OtherConsumption 0.0°/ 0.0%/. 0.4% 0.0% 0.0%/. 0.1% 0.1% 0.1% 0.5% Investnent 0.8% 1.6% 1.4% 1.0/. 0.7% 1.6% 5.6% 8.2% 6.9Y. Source: Staff calculations based on EBC 1987-1989. 2.19 The loss of purchasing power which resulted from the economic crisis and the devaluation raises the question of how people have been coping with these changes14. Field work carried out in November-December 1994 confirms what logical reasoning would suggest: as a rule, the poor were less affected than others because they rely to a greater extent on own-consumption. In this sense, the impact of the crisis and the devaluation was relatively small on the expenditure patterns of the poor in general, and of the rural poor in particular. But a drop in real incomes between 1990 and 1994, and an increase in the cost of living after the devaluation, did entail a change in consumption patterns for all. While the non-poor reduced expenditures on luxury items (imported food and drinks, various services), the poor, who buy no luxury items and only the most essential non-food items, have had little choice but to reduce food expenditures. As food consumption is relatively inelastic with regard to total energy intake, the food bill was reduced by choosing cheaper calories. Already in 1987-89 the core poor bought almost three times as many calories as the non-poor for the same price (52.56 calories per CFAF compared to 18.40 calories per CFAF). With the crisis it appears that this pattern intensified: households shifted their labor power towards home production through 14 Field interviews indicate that most people see the devaluation basically as another manifestation of the economic crisis ("furst we had civil unrest, then the devaluation"). Togo: Overcoming the Crisis, Overcoming Poverty 19 migration to the rural areas, cultivation of food crops and buying of unprocessed foods (subsequently processed at home). This shift, in turn, has important implications for women's time allocation, since to a great extent the additional labor required to cope with lower purchasing power is women's--and girls'--labor. Nutritional Indicators 2.20 Nutritional indicators are particularly good measures of the general level of well-being of a population (see Box 2.2). They reflect the ability of the population to feed itself and to lead a relatively healthy life. Actually, they provide insight into the end result of many different factors that influence the quality of life: access to food, access to proper health care facilities, and, in general, the quality of the living environment. Nutritional indicators are often more representative than indicators based on income characteristics alone, because they reflect not only the income level, but also its buying power and the spending choices made by the population. Further, indicators based on incomes are often biased towards the monetary sector of the economy, which is precisely the area in which the poor tend to participate less than their richer counterparts. Nutritional indicators also have the advantage of being relatively sensitive: changing conditions almost immediately show in the level of nutritional indicators. The weights of children and adults may go up or down within a period of weeks, and changes in long-term indicators, such as faltering growth rates of small children due to the cumulative effect of under-nourishment, are often visible within a year. Box 22: Nutritional Indicatorn Most nutritional surveys employ two or three basic indicators. Height-for-age, which reflects the more chronic aspects of malnutrition, is most useful in evaluating the impact of poverty. If children grow up in a poor environment, get ill very often, and have too little to eat to make up for the energy lost, growth wil tend to slow down. This in turn is a completely functional adaptation to circumstances, because short and lean children need less food than their bigger counterparts, and to reduce growth is a way of economizing on future food intake. The process of faltering growth is called stunting and is usually measured by looking at the attained growth of young children.'s Weight-for-height is a measurement of acute malnutrition caused by inadequate intakes due to food shortage or to poor food absorption linked to disease (often accompanied by a lack of appetite), During such episodes, children stop increasing their weight and, if this condition persists over some time, existing body tissue may be affected too, a process which iscalled wasting. A third measure, which was common in older nutrition surveys, is weight-for-age. This indicator combines both aspects and is therefore more difficult to interpret: low weight may simply result from small stature, i.e., frm previous stunting, or it may indicate an acute shortage of food. In either case, there is a need for intervention, but not necessarily the same type of intervention. This makes weight-for-age a less suitable indicator for policy-oriented poverty assessments or nutrition surveys. 5The growth rate of stature itself, while a more sensitive indicator, requires measurement on more than one occasion and it is therefore less suitable for major national surveys. It is also a less stable indicator as growth rates may vary according to season, which also reduces its value for poverty assessment. Togo: Overcoming the Crisis, Overcoming Poverty 20 2.21 There have been a number of nutrition surveys in Togo. A national survey conducted by the Center for Disease Control (CDC) of Atlanta in 1976-77 found that about 20 percent of all children between 6 and 71 months were stunted, while 3 percent were wasted (see Table 2.8). Wasting was most important among the children between 12 and 23 months of age, the weaning period, while stunting increased until the age of 4, after which the number of stunted children leveled off,"6 suggesting that nutrition is adequate after the age of 4, but not sufficient to permit catch-up growth. A second national survey of children between 0 and 36 months conducted by the Unit of Demographic Research of the University in 1988 found 29.6 percent of the sample to be stunted, and 5.3 percent to be wasted (see Table 2.8). In the same year, a survey conducted by CRAN/OCCGE of children between 6 and 60 months found 33.6 percent stunting and 4.3 percent wasting. The three surveys used slightly different definitions of wasting and stunting, and somewhat different age groups,"7 so the figures are not immediately comparable. Also, it should be noted that the 1988 surveys followed a year with a relatively poor harvest, which may help explain the higher levels recorded in that year. Table 2.8: Stunting and Wasting Rates from Different Surveys, 1976-77 and 1988 CDC 1976-77 UJRD/DHS 1988 CRAN/OCCGE 1988 (6,120 children 6-71 months) (1,396 children 0-36 months) (2,863 children 6-60 months) wasting Stunting Wasting Stunting Wasting Stunting Maritime 2.4 18.4 1.8 24.0 4.3 34.0 Plateaux 2.0 16.5 0.9 26.2 3.0 31.0 Centrale 2.2 20.5 3.5 32.0 2.3 28.8 Kara 2.1 26.5 3.9 34.3 4.8 36.6 Savanes 7.7 30.8 23.7 41.8 8.5 36.8 Urban 1.1 11.7 3.7 21.3 Rural 2.9 21.1 5.9 33.0 TOGO 2.7 19.8 5.3 29.6 4.3 33.6 Sources: CDC, URD/DHS, CRAN/OCCGE. 2.22 Catholic Relief Services (CRS), which ran until September 1995 a food distribution program including monthly weighing of children, recorded annual averages of underweight children (i.e., below 80 percent of weight-for-age standards; see Table 2A.2 in Annex 2). Since the children participating in the program cannot be considered a representative sample, the figures collected cannot be used to estimate the extent of malnutrition in Togo, but a comparison over the years shows that conditions have remained relatively stable with about 32 percent of children under 80 percent of the World Bank, "Poverty Profile of Togo, Country Assessment and Policy Issues", December 1991. 17 The 1976-77 survey defines a child with a Height for Age below 90 percent of the reference population as stunted, while a child with a Weight for Height below 80 percent was considered wasted. The 1988 surveys use a threshold of -2 standard deviations to define stunting and wasting. Furthermore, the reference population differ: the older survey uses the so-called Harvard Standards, while the 1988 surveys used the NCHS/CDC/WHO reference population. Togo: Overcoming the Crisis, Overcoming Poverty 21 weight-for-age norm during the period between 1984 (when systematic data collection started) and 1992 (when systematic data collection had to be interrupted because of insecurity problems). A slight downward trend is shown by the data for the initial years of this period, when numbers fell from 39 percent in 1984 to 32 percent in 1987, suggesting that the nutritional situation did improve during that interval. This conclusion, however is contradicted by the survey results presented in the previous paragraph, which, if anything, indicates a small increase of malnutrition instead of a decrease.'8 After 1987, the nutritional situation appears to have remained stable. 2.23 There is considerable regional variation in nutrition, following a pattern which is found in each survey: malnutrition rates are highest in the two northern regions, followed by Maritime (see Table 2.8). In the first two surveys, Plateaux shows the lowest rate of malnutrition; in the CRAN/OCCGE survey, Centrale shows a slightly lower rate than Plateaux. Obviously, the malnutrition data confirm the general pattern of poverty and incomes presented earlier: the incidence of poverty and malnutrition rises moving from south to north, with a slightly higher concentration in Maritime region. The data collected at the CRS clinics confirms this pattern, showing high rates of underweight children in the north--in Kara and, particularly, in Savanes, while the Maritime figure is consistently higher than that of Plateaux and Centrale (see Table 2A.2 in Annex 2). The survey data for 1976-77 and 1988 show an important difference between urban (defined as the district capitals) and rural children, with urban rates about one third lower than rural rates. 2.24 Nutritional levels are also affected by fertility rates. The total fertility rate is 7.1 and has remained practically unchanged for the past two decades, although there are indications of a slight tendency towards a lower fertility in the last decade. This means that, especially in rural Table 2.9: Effect of Birth Intervals areas where fertility rates are higher than the onNutrition average, almost all women of reproductive age Children 0-36 months are either pregnant or lactating, and birth Wasting Stunting intervals tend to be short. The impact on Interval nutrition is felt in a number of ways. Pregnant first born child 5.5% 35.2% and lactating women are more likely to suffer less than 2 years 6.7% 28.1% from anemia (caused by iron deficiency) than 2 - 3 years 5.4% 30.1% others, with poor women being at a higher risk 4 years and over 3.7% 22.1% because of their limited access to proper diets. Source: URD/DHS, 1988. According to UNICEF, 42 percent of women of reproductive age are anemic and anemia accounts for 8 percent of maternal mortality."9 Because anemia diminishes productivity, pregnant and lactating women enter a vicious circle whereby poor diets result in lower '8The high rates recorded at the beginning of the 1984-87 interval can also result from changes in the participating population. For instance, it is likely that in the initial years participation was higher among needy families, with a relatively high percentage of malnourished children. Such families are often the first to be referred to a supplementary feeding program of this kind. 19UNICEF, La Situation de la Femme et de l'Enfant au Togo, 1994. Togo: Overcoming the Crisis, Overcoming Poverty 22 productivity, which makes it impossible to improve diets. Children are caught in this vicious circle as well, as they depend on their mothers for nutrition either from breastfeeding or from cooked food. The more young children in the household, the less likely their mother will be to have the time and energy to ensure that they are properly fed, especially if birth intervals are short (see Table 2.9). 2.25 Even when the average food intake is enough to cover the basic requirement of 2,050 calories per capita per day, the instability caused by seasonal fluctuations is a matter of concern. It means that a much larger number of people is below the poverty line during part of the year. And, worse, it has lasting consequences, as it is a typical part of the poverty trap: due to seasonal scarcity, farmers have a lower capacity for heavy labor at the time when the need for heavy labor is highest. This may not cause them to reduce labor output during a season of food shortage; rather they may use the energy stores in the body and lose weight. While such weight losses may not have serious consequences if kept within a certain acceptable range, there are limits to what can be managed without lasting effects. They certainly explain why harvesting time is a period of great festivities with many ceremonies and significantly higher consumption of food. This increased consumption serves not only to bring consumption up to normal levels, but also to restore necessary reserves, which will be needed in the next agricultural season. The human body is probably a better storage facility for such energy than the granaries, which often have losses of up to 25 percent.20 2.26 The seasonal variations of food supply have important consequences for policy development. Any interventions that aim at alleviating poverty through improvements of agricultural techniques or through income-generating activities have to take into account the balance of seasonal food shortages and labor requirements. The agricultural system depends on activities of both men and women, and unless adequate food supply for both is guaranteed, there is no room for expanding labor efforts, however great the deriving benefits could be. Further, the system as it exists now seems to be in a precarious balance. Even relatively small fluctuations may upset the system and cause considerable problems. What may seem a minor intervention may suddenly turn into a major disaster when combined with the negative effects of, say, too little rain (see Chapter 5). 20World Bank, Agricultural Sector Review, 1993. Togo: Overcoming the Crisis, Overcoming Poverty 23 Social Indicators 2.27 To understand the nature Table 2.10: Togo Social Indicators of poverty, it is also important to look at indicators other than per Birth rate 45.4 per 1,000 (1993) capita incomes, expenditures, and Mortality rate 13.2 per 1,000 (1993) nutritional' status. Access to Life expectancy 49 male (1989) sufficient food is only one of the 51 female (1989) primary needs that should be Fertility rate 7.08 Primary school enrollment considered when describing rate 60.4 % (net, 1994) poverty. Other aspects, such as Adult literacy rate 52% mrale access to health services and 31% female access to education, are key Medical doctors I per 11,270 (1991) indicators, because they measure Hospital beds 1 per 614 (1992) Maternal mortality 476 per 100,000 live births people's opportunity to improve Infant mortality 97 per 1,000 (1989) their living conditions by using Under 5 mortality 140 per 1,000 the services provided in their Low birth weight 20% (1994) country. Immunization 67.1% BCG (1994) 7 1.1I% Polio3 (1994) Access to potable water 59% (1988-91) 2.28 Compared to other Sub- Access to sanitation 22% Saharan African countries, Togo's Source: UNDP, UNICEF, Statistiques Scolaircs, Statistiques Sanitaires. primary school enrollment used to be quite high (71 percent net enrollment rate in 1980). Starting in the mid-1980s, however, rapid population growth and economic decline combined to produce a steady deteriorationi in all educational indicators. The political and economic crisis of the early 1990s, and in particular the 1992-93 strike, further weakened the sector. As a result, primary net enrollment rates have now gone down to 60.4 percent, and student repetition and drop-out rates are among the highest in Sub-Saharan Africa (46 and 7.6 percent, respectively). Severe overcrowding in urban classrooms contributes to keep instructional quality low, while lack of pedagogical material and teacher shortages hamper schools throughout the country, but are particularly felt in rural areas. Enrollment rates diminish going from south to north, mirroring other poverty-related indicators (see Table 2.11). Hence the lowest primary school enrollment rates are found in the Savanes region (30 percent), which is also the region with the greatest teacher shortages--and the highest incidence of poverty. As low as enrollment rates may be, it should be kept in mind that they are likely to provide an over-optimistic picture because they represent maximum attendance and do not show variation from month to month (if children are enrolled at the beginning of the school year, they will be counted as attending school for the whole year, even if they are withdrawn after a few days). Togo: Overcoming the Crisis, Overcoming Poverty 24 2.29 There are significant and persistent gender differences in school enrollment. Initial enrollment rates for girls are, on average, only slightly below those for boys (girls are 42 percent of first grade students), but girls drop out of the system at astonishingly high rates: by the first year of secondary schooling they represent only 29 percent of the total, and by the last year of secondary schooling this figure has dropped to 12 percent. In addition, the national average hides regional differences (see Table 2.11). Once again the Savanes stands out: less than half as many girls as boys are sent to primary school, which means that, given the low regional Table 2.11: Net school enrollment rates by region and by sex enrollment rates, only 2 out of ten girls go to 100 S.2 85.1 school. Low levels of _ 8 schooling for girls can 64.4 61.6 569 66.7 be explained by a 60- 56_ 50.3 47.5 combination of 40 - economic and cultural 22n factors (see also 20 ÷iiL zzL Chapter 6). Girls are 0 + __ needed to perform Mauitime Plateaux Centrale Kara Savanes domestic chores (especially care of CGtrls oEoys younger siblings and fetching water) or to help their mothers carry produce to the market, and these activities take priority over education. Hence, water scarcity in the Savanes, and the resulting time- consuming chore of fetching it, may help explain gender differences in the region.21 Indeed, it is difficult for parents to see much value in educating girls. From the parents' point of view, educating a girl is a net loss: while she is in school, they forego her economic contribution to the household and have to pay for school-related expenses; after she leaves school, the advantages deriving from her education will accrue to her husband's family, who in most cases will not even pay a higher bride price for an educated bride. 2.30 Education and health are closely related. Female education in particular has been linked to improved household health conditions, including lower infant and matemal mortality, as better educated women are more likely to promote proper hygiene in the family, to adopt basic preventive measures and to seek or apply appropriate treatment when disease strikes.' In addition, increased female education has an impact on health 21 The relationship between water fetching and girls' schooling is well illustrated by the situation found a village in the Oti district during the Rapid Poverty Appraisal. Water is so far from the village that girls of elementary school age are not expected to go fetch it, and girls' enrollment rates at the primary levels ame much higher than the regional average. The relationship between water fetching and schooling is readily recognized by the teacher as well as most parents. 22 King, E., Educating Girls and Women: Investing in Development. Washington, DC: The World Bank, 1990. Togo: Overcoming the Crisis, Overcoming Poverty 25 because it is linked to lower fertility (in Togo the total fertility rate is an alarmingly high 7.08). 2.31 The main health-related problems in Togo have been identified as high infant and maternal mortality (86 per thousand and 476 per 100,000, respectively) and a high incidence of transmissible and parasitic diseases. According to the last available data (1991), the major diseases are malaria (37 percent of visits to health facilities), diarrhea (11 percent) and acute respiratory diseases (9 percent).23 Malaria, anemia, diarrhea and acute respiratory infections are the four main causes of non-traumatic death, in that order. It appears from these epidemiological data that morbidity and mortality rates are in large part the result of an unhealthy environment (stagnating water, inappropriate waste disposal, improper house ventilation) and insufficient access to potable water (estimated at 59 percent). 2.32 AIDS is rapidly spreading, especially among high-risk groups such as prostitutes (46 percent sero-prevalent in Lome in 1992) and long-distance drivers. As of June 1994, 4.114 cases had been reported; sero-prevalence is estimated at 5 percent of the population (i.e., some 197,670 persons are estimated to be HIV positive). With the spread of AIDS, many households will become extremely vulnerable, as AIDS tends to strike adults in their most productive years. Already 12,000 orphans had been reported in 1994, and their number is bound to grow. 23 Houngues, K. D., and Adzozv, M.R., op. cit. Togo: Overcoming the Crisis, Overcoming Poverty 26 3. THE DYNAMICS OF POVERTY The Concept of Vulnerability 3.1 Vulnerability is an important aspect of Box 31: Poverty as Vulnerability (I) poverty, because it reflects its dynamic nature. It refers not to lack or want, but to 'I'm old and I can't work, and therefore I'm defenselessness, insecurity and exposure to risk. poor. Even my land is old and tired, so whatever Poor people are usually among the most little I manage to work does not give me enough vulnerable, but not all vulnerable persons are harvest for me and my children. Now that the During the strike period, for example, soudure comes, I will not have enough to last me poourity was low and vulnerability was through the next harvest, and I might have to sell security was low and vulnerability was a goat. But I only have one goat, and I don't widespread, affecting people who would not be want to sell it. Besides, one goat would not be considered poor on the basis of their household enough to buy enough millet So I will have no consumption. Also, the concept of vulnerability choioe but to borrow. That's the way it Is- in this is more meaningful to the average Togolese villae we oll suffer, because we all have to than that of poverty, as illustrated by the poors' borrow. We borrow from the rich people from Mandouirr'. own definition of poverty (see Box 3.1 and 3.2). Poverty, in the sense of low income or lack of Source: PWtclpoy Poverty Apprisal, old fannr in food, can be reduced by borrowing or by selling oti Distict (Svarc) Novembw 1994. an asset, but this makes a household more _ vulnerable. Thus, many Togolese may prefer Box 3.2: Poverty as Vulnerability (II) short-term poverty, such as an untreated illness or insufficient food intake, to an increased "We are all famers in this village. When two vulnerability derived from selling an asset or farmers cultivate together the same plots and, at 14 the end of the harvest season, one has made a borrowing from neighbors.2 profit that allows him to get a lot of things and the other hasn't eaned a thing, they will say that 3.2 To understand what makes people in the second faimer is poor. But noxt year it could Togo vulnerable, it is useful to conceive of be the reverse. he fact is that we are all poor in vulnerability as a function of assets: the more this village." assets people have, the less vulnerable they are. Source Partcipzov Poverty Appaisal. fam in In this context, assets should be considered in Lac.Dirict (Maritime), December 1994. broad terms, as a wide range of tangible and I intangible stores of value or claims to assistance which can be mobilized in a crisis.25 Thus assets include stores (e.g., jewelry, money, granaries), concrete productive investments (e.g., farming and fishing equipment, 24 This chapter uses a conceptual frmnework similar to the one presented in the Benin Poverty Assessment. For a discussion on vulnerability see: Chambers, R., "Vulnerability: How the Poor Cope", Institute of Development Studies Bulletin, vol.20 no.2, April 1989. 25 This definition is based on Swift J., "Why are rural people vulnerable to famine?", IDS Bulletin, vol 20 no 2, 1989. Along similar lines see also: Sen, A., Poverty and Famine: An Essay on Entitlement and Deprivation. Oxford: Clarendon Press, 1981. Note, however, that Sen sees vulnerability-nd therefore Togo: Overcoming the Crisis, Overcoming Poverty 27 animals, tools, land), human investments (education and health), collective assets (irrigation systems, wells), and claims on others for assistance (kinship networks, friendships, savings clubs, patrons, credit). An awareness of the diverse nature of assets, and of their hierarchy, is essential for meaningful policy action. It helps to identify the most vulnerable individuals and communities, thus facilitating effective targeting. It makes it possible to plan interventions before vulnerability turns into absolute poverty, therefore using a preventive rather than reactive approach. It provides guidance on how to organize assistance in a way that builds on the assets of the recipients, so as to minimize the need for future interventions. The Causes of Vulnerability 3.3 Vulnerability can be determined by factors operating at different levels. To facilitate the analysis, these factors have been grouped into four categories according to their level: individual, household, community and national. Each category is discussed and illustrated below. 3.4 Individual-level factors determine the vulnerability of a person regardless of the characteristics of the household or the community to which the person belongs. Human assets such as health and education, for example, affect the degree of vulnerability of the whole household in as much as the household is a social and economic unit, but more than anything else they determine individual vulnerability. In a country such as Togo, where there is practically no form of modem social security, illness or handicap have devastating effects on an individual's self-reliance and are often the cause of absolute destitution. Thus, it is not surprising that village children asked to draw poverty often depicted handicapped, sick or old people (see Box 3.3). Illiteracy makes it more difficult for individuals not only to find jobs or to improve their productivity (e.g., by applying fertilizers correctly thanks to the directions on the bag) but also to take advantage of other potential assets such as government services or access to credit. With an illiteracy rate of 69 percent, women are the primary victims of this aspect of vulnerability. 3.5 Much individual vulnerability, however, is the result of socio-cultural norms which assign different values to ascribed characteristics like age and gender. Such differences often find expression in intra-household inequality, meaning that within the household, resources and assets are unequally distributed according to the perceived desirability of certain characteristics. Women and children are more vulnerable because tradition gives them less decisional power and less control over assets than men, while at the samne time their opportunities to engage in productive activities, and therefore to build their own assets, are more limited. Traditional beliefs, in turn, may find expression in customary law, which in this way sanctions male privilege and the vulnerability of women and children. relative poverty-as a direct function of a household's ownership of tangible resources and endowments, but does not include claims on assistance from others in his analysis. Togo: Overcoming the Crisis, Overcoming Poverty 28 Bow 3.3: Poverty through the Eyes of a Child t..-I.l; 1%. 5hjis man [top rght] has no good egs to work, so he has to look at his old ftber work hard d The fither is tired after working hard with the daba to dig out some 4?m. (oval to rest and stretch. Their house is in poor shape, full of holes and cracks, but ity cat rt one has no good leps and the other is too old and tired". Soure Rapid Povoty Appis, clanay school child in Bassar Disct, Kla Rgn VDacmbr m 3.6 Customary law considers children as property of their family, and gives them no individual rights. The widespread acceptance of highly exploitative labor practices (see para. 3.19-3.21) and the occurrence of genital mutilation performed on girls (see Box 3.4) are among the most extreme examples of the vulnerability of children. Women's traditionally subordinate position constrains their access to factors of production: they cannot own land,26 the plots they receive in usufruct are generally those left over by men (for being too far, too small or too poor), they are seldom contacted by extension agents, and they have only residual access to tools and means of transport owned by the household. Girls' schooling is limited by the patrilineal and patrilocal tradition, which often causes parents to see female education as a waste of money since it is like investing in somebody else's (her husband's) family (see para 6.22). Women cannot inherit, but the levirate tradition makes it possible for their brothers-in-law to inherit them along with the rest of their deceased husband's estate (including children).27 Although the new constitution adopted in 1992 recognizes all citizens as equal before the law, traditions are difficult to change and the 1980 Family Code implicitely recognizes gender inequality by allowing polygamy and regulating--thus legitimizing--the brideprice. 26 A few exceptions have been reported in the Ana and Losso ethnic groups in: UNICEF, Lafemme et I'Enfant au Togo, Lome, 1988. Also, women may inherit land from their father if they have no brothers and no patemal uncles. 27For further discussion on women's condition see: Groupe de Reflextion et d'Action Femme DNmocratie et Developpement, "Situation de la Femme au Togo", mimeograph, Lome, March 1994. Togo: Overcoming the Crisis, Overcoming Poverty 29 Box 3.4: Harmful Traditional Practices on Children A number of traditional practices have harmful health effects on children. They include: femal genital mutilation (FGM), deep intestinal cleansing for purification purposes, scarifications forced feeding of infants, scolding for treatment purposes, and food taboos (which often deny children access to the most nutritious food, such as eggs and meat). FGM (excision) is perhap the most dangerous of such practices, as it poses an immediate risk to health and sometimes lifi as a result of hemorrage, shock and serious infection (such as tetanus and scepticemia), and can entail serious long-term consequences (such as urinary and pelvic infections, a variety of urinary afflictions, fistulas, cists, excessive scarring and childbirth complications). FGM can also hav significant psychological repercussions depending on the age at which the operation i performed. It is practiced by the authoctonous groups of the northern and central regions, but, due to migration, it can be found throughout the country. The estimated incidence is 40-5 percent. Source: Comite Inter-Africain sur les Pratiques Traditionelles Ayant Effets sur la Sant6 des Femmes et des Enfants (CI-AFrTogo); and Fran Hosken, The Hosken Report, Women Intemational Network News, 1989. 3.7 Household-level factors relate to household structure and composition, and to its socio-economic characteristics. Other things being equal, the higher the ratio of healthy (productive) adults to children, the more likely the household to be able to keep out of poverty. Loss of a productive adult--whether due to disease, death, divorce or neglect-- drastically reduces a household's capacity to overcome external shocks and is one of the main causes of destitution28. More generally, limited access to means of production at the Box 3.5: The Young Retirees When the arrete no. 607/MTFP took effect on April 1 1985, all Togolese civil servants and para-statal employees (agents permanents et assimiles) with thirty or more years of service were automatically retired. Hence the creation of the term "young retiree" to designate those forced to retire before reaching the official retirement age of 55. This decision, sudden and unexpected, was a very painful surprise for the first cohort of young retirees (1,175 individuals, 40% of whom were teachers). For those participating in the National Social Security Fund (Caisse Nationale de Securite Sociale or CNSS), about half of the total, pensions were at the most equivalent to 40% of their regular salary, because in spite of 30 years of service, the CNSS had only existed since 1969 and therefore benefits were paid on the basis of only 16 years of contributions. Thus, 20% of the pensions turned out to be lower than the minimum wage (CFAF 12,402 net per month). Not suprisingly, standards of living were severely affected. Paying for school expenditures became a problem for most,and more than half of the households had to withdraw at kast one child A survey six years later found that many premature deaths and broken families had been attrbted to the brutal way in which the early retirement had been carried out, while practically all young redrees declared themselves significantly impoverished both materially and psychologically. Source: Comitt DSA, Caractdrestiques et Comportements de Groupes Vulndrables en Situation d'4swtenmt Sructurel, part I, May 1992. 28High fertility rates, however, should not be seen as totally irrational behavior on behalf of the parents, for children represent not only future sources of labor but also a way to expand the family's claims on others through an expanded kinship network (including the children themselves as a form of old age security for the parents). Togo: Overcoming the Crisis, Overcoming Poverty 30 household level (including labor) results in low stores and therefore greater vulnerability. 3.8 Average-sized households disposing of less than one hectare of arable land, such as can be found in the Massif Kabye area (Kara) and in the coastal zone, may not be absolutely poor but are certainly in a very vulnerable position, because they are unable to grow enough food to feed themselves and tend to depend on employment from richer farmers. Similarly, households who do not belong to the founding clan of a village are in a vulnerable position because they cannot own land and the founding clan determines usufruct arrangements (see Box 3.7). In urban areas, uncertain employment conditions make a multitude of informal sector operators (and their families) highly vulnerable to the slightest market changes. Civil servants, traditionally considered the most privileged members of society because of their secure jobs, have seen their security decline with the financial discipline required by structural adjustment and the families of the so-called young retirees have seen their standard of living decline drastically and unexpectedly (see Box 3.5). 3.9 Community-level factors are shared by whole communities, and are primarily determined by ecological and infrastructural conditions. Massive soil degradation-- especially in overpopulated areas of the Savanes, Kara and Maritime regions--is probably the greatest ecological cause of vulnerability, as it lowers the productivity of entire villages, thus affecting not only the tangible assets of individual households, but also many of the claims that poorer households can lay on other members of the community. In the same vein, over-exploitation and contamination of groundwater around Lome are at the root of much illness, which decreases productivity, lowers the value of human capital and puts a strain on traditional assistance networks, particularly in low-income neighborhoods where people cannot afford boiling water. 3.10 Poor road conditions result in isolation (enclavement) for entire villages, which sometimes cannot be reached for several weeks during the rainy season. The breakdown of government activity during the strike period and continued budgetary difficulties have caused a significant deterioration of the transport infrastructure during the past Box 3.6: No School and No Health Center five years, with some feeder roads becoming impassable because of broken "We are all poor here, because we have no school bridges or huge potholes. In these cases, and no health center. If a woman: has a ~difficult household assets may be an effective delivery, we have to put apagne [traditional piece of cloth] between two sticks .and carry her for 7 km to buffer against the inflated prices caused the health center. You know how long t takes to by a lack of supply, but sometimes there walk like that? There is nobody who can help in case is no supply to speak of. In addition, of emergency, that's why I say th at we're all poor in isolation makes it difficult to market Kouyoria." goods and services available in the oc a oe pfa village, thus limiting people's capacity to north- Kam Regon, Dwme 1994. accumulate assets. For example, in a village in Bassar district (south-west Kara region) the collapse of a bridge caused farm gate prices for maize to fall from 300 CFAF Togo: Overcoming the Crisis, Overcoming Poverty 31 to 125 CFAF per bowl, as farmers could no longer reach the nearest market and traders coming to the village took advantage of the situation. 3.11 The availability of social services within the community greatly lowers vulnerability, for it makes it easier to optimize human investments in health and education. Distance from the nearest school or health center was found to be an important aspect of the poor's idea of poverty (see Box 3.6), and low primary school enrollment was often attributed by parents to a combination of high cost and excessive distance. 3.12 National-level factors create vulnerability which affects basically the whole population in the country, but tends to be more burdensome for the lowest socio- economic strata. These factors have to do with general geographical and political conditions, as well as with macro-economic policies and laws. For example, having more than half of its exports accounted for by four commodities (phosphates, cocoa, coffee and cotton), makes Togo's economy highly vulnerable to international market price fluctuations for these commodities. A public sector dominated by political patronage makes those without good contacts more vulnerable and often unable to "cash" on their legitimate claims for government assistance. A government budget which protects military expenditures necessarily implies a trade-off with productive expenditures and social services. The political crisis and the ensuing violence and insecurity lead to a massive withdrawal of investors and donors, thus jeopardizing economic growth and social service delivery for some years to come. The legal system may create vulnerability by its inability (or refusal) to safeguard the rights of certain groups. A case in point is the lack of a clear land tenure system, which inherently favors the most influential individuals in a given dispute. 3.13 The recognition that poverty and vulnerability can be determined by factors operating at different levels means that the fight against poverty can--and must--be fought at different levels as well. It also means that widely differing interventions can be effective in reducing or preventing poverty, but that only in a few cases interventions focusing on one or two factors (or limited to one sector) can improve significantly people's welfare if they are not complemented by interventions focusing on different factors/sectors. This explains the success of some NGOs using an integrated approach at the micro-level (community or household), but it also highlights the need to complement such efforts by addressing vulnerability factors operating at "higher" levels. 3.14 The matrix in Figure 3.1 presents an overview of the causes of poverty and vulnerability in Togo. Factors have been grouped by level (micro, meso, macro), and sub-divided according to their main sphere of influence (social, economic, natural). Distinctions between levels and spheres of influence are not always clear-cut, and the classification is not meant to be exhaustive or definitive. The main purpose of the matrix is to encourage thinking about the dynamics of poverty and, more importantly, to facilitate an assessment of the poverty reduction/prevention potential of different interventions. Togo: Overcoming the Crisis, Overcoming Poverty 32 Figure 3.1: Overview of the Causes of Poverty and Vulnerability Micro Level Meso Level Macro Level (Individual & Household) (Community) (National) Natural * poor soils, low fertility * poor soils, lowfertility * natural disasters Factors (degradation and erosion); * natural disasters, (floods, (droughts); droughts and parasites); * population pressure; * difficult access to safe * limited natural resources water; basis; Social * poor health (malaria, * insufficient access to * insufficient & inefficient Factors malnutrition, handicaps, old health services (excessive funding for primary age); cost/distance); education; * high dependency ratio * inadequate water supply * timid family planning (high fertility rates *many andsanitation; programs; dependents, and perfunctory * limited access to educat. * insufficient funding for contributions from male (excessive cost/distance); primary health care services, "breadwinners" *>few * insufficient access to (drug shortages, poor quality providers); family planning services, care); * intrahousehold inequality * cultural values & tradition * discriminatory family law (children treated as property (about gender differences, the (cwomen's inferiority); of their father's family, role of children, attribution of * geographic inequality in women considered inferior prestige and authority, resource allocation for social to men); agricultural and health services. * low education (especially practices).29 of women). Economic * low returns to labor (low * isolation/remoteness * poorly integrated market Factors yields, low or unreliable (difficult marketing, lower for food products (localized producer prices, low wages, farm gate prices, reduced food shortages, high seasonal low margins of profit in access to social services); variations in prices, low informal activities); * lack of employment producer prices, high * unemployment and under- opportunities (under- and consumer prices); employment, unemployment, compressed * poor rural infrastructure * inter-household inequality, wages); (higher marketing costs, (control over land, rights & * land shortages in parts of difficult delivery of social duties related to social the Savanes, Kara and and extension services); standing); Maritime regions;30 * low development of * unequal access to * inefficient agricultural financial intermediation productive resources by production systems (land use, systems (savings & credit gender (women denied equal agricultural practices and facilities for the poor seldom access to household access to agricultural inputs). available); resources, such as land, * unclear land tenure system means of transport and (low incentives to sustainable 29Culaural values and traditions have been listed at the community level because it is within a community (or ethnic group) that they have the strongest influence on behavior. 30 This is considered a problem at the community level, because in land-rich communities all households tend to have access to enough land. Togo: Overcoming the Crisis, Overcoming Poverty 33 Micro Level Meso Level Macro Level (Individual & Household) (Community) (National) working equipment); agricultural practices, lack of access to credit. conflicts, inefficient use of natural resources). Vulnerable Groups 3.15 Because individuals are generally part of a household and live in a community, their personal level of vulnerability will be determined by the interaction among the categories of factors identified above. This is why attempts to identify vulnerable groups for targeting purposes are unlikely to produce mutually exclusive categories. The remainder of the chapter concentrates on two vulnerable groups which have been selected not on the basis of their greater vulnerability or of their large size, but because they are specific to Togo and they are generally not included among the typical targets of interventions to prevent or reduce poverty (hence, they are more likely to be forgotten by decision-makers). This focus does not mean that the vulnerability or the importance of other, more "typical" vulnerable groups is not recognized. Certainly, the attention given by a recent UNDP report to women (especially pregnant and lactating women) and children (especially children below five years of age) is fully justified.3' What follows, therefore, should be seen not as a substitute to previous, and perhaps more conventional, lists of vulnerable groups, but as an addition. 3.16 Displaced Families. Whether families are forced to move away from their ancestors' land or whether they are simply encouraged to do so--but left free to decide, a fundamental feature of displacement is that it causes a significant and sudden unraveling of existing patterns of social organization, with a profound impact on the structures of economic and cultural life. Production systems are dismantled, informal solidarity networks are rendered nonfunctional, traditional authority and rights may be questioned (by the host population), and the abandonment of symbolic markers (e.g., sacred trees or forests) may sap at the roots of people's cultural identity and result in deep alienation. The consequences of displacement may vary considerably, depending on the manner in which the move took place (e.g., planned or unplanned), the attitude of the host population, the extent of loss and replacement of income-generating opportunities, and the resilience or vulnerability of the displaced people. But a comparison of empirical 31 The report identifies the following groups as deserving particular attention in the fight against poverty: (a) under-priviledged * women and children below fourteen years of age; (b) vulnerable * urban poor (5% of informal sector), malnourished children in the Maritime region, rural dwellers particularly hit by cuts in social expenditures, lactating or pregnant women, children below five years of age, and prematurely retired civil servants (jeunes retraites); (c) marginalized * beggars, delinquents, mental patients, and certain handicapped people. See: UNDP, Strategie de Lutte contre la Pauvrete - Avant Projet, part I, draft, March 1995. Togo: Overcoming the Crisis, Overcoming Poverty 34 findings from around the world suggests that the ultimate common factor underlying the broad spectrum of reported displacement consequences is the onset of impoverishment.32 3.17 Togolese families have been displaced primarily on four types of occasion: (a) forced eviction for the creation of national parks, (b) government-sponsored resettlement for development purposes, (c) government-induced resettlement for mining purposes, and (d) violent civil strife. Each type is discussed below. (a) Forced Eviction. In the 1970s and early 1980s, government policies to protect the country's fauna and forests resulted in the creation of several large wildlife preserves and national parks, mostly in the north.33 By all accounts, the creation of protected areas was grossly mishandled by the government.34 An estimated 60,000 Box 3.7:: The Creation of the Keran people were chased from their National PartI: homelands, at times without prior notice 'On. day t security forces 'snt the entrvlla and in any case without being provided n by buring ou houses to lt the elphat any assistance by the state. Farmers lost walk safely on our fields. The government al their land, which was declared state forces us to do sothing else than whatvwe property, and their homes, which were choose to do... Last yearI decided to come.ba often destroyed, without receiving any rm an old man and after 1S years in soutI didn't want to be teatd as a foreigneranymore. formn of compensation. Those who l Foreigners have a hard lifel Down there we were resettled closeby had to restrain their so exploited [by the host population] thatewe livestock (which otherwise could have fanned one ha. of land, we had tQ givehaway half threatened protected species) under the of the crop-tenoutoftwenysacs ofmaize- a threat of stiff fines, while their crops were t pay taxes on:te land. Never3 even aftanY 35 years, could we beCOMC.owners of thouelds often damaged by wild animals.35 In ' c w b o of th aiid i Source: Rapid Poverty Apprikal old finner in Oskik, Kara Rep-on, November199.- 32 The impoverishment process typically occurs along seven main dimensions: landlessness, homelessness, joblessness, marginalization, food insecurity, increased morbidity and mortality, and social disarticulation. For further discussion see: Michael Cernea, "Involuntary Resettlement: Social Research, Policy and Planning" and Thayer Scudder, "A Sociological Framework for the Analysis of New Land Settlements" in M. Cernea, ed., Putting People First, second edition. Oxford University Press, 1991. 33 The Keran National Park was created in 1971, enlarged in 1975-76 and expanded again in 1981-82. The Oti River valley east of Dapaong was made into a wildlife preserve in 1981; in the same year the Fosses aux Lions national park was created south of Dapaong. Thus, from 1960 to the early 1980s, the total area in the Savanes region having the status of national park, game reserve or classified forest increased from 164 km' to around 2,632 km'. This area accounts for almost a third of the region's total surface area, and for 40 percent of all protected area in Togo. (See: Ministry of Rural Development, Environment and Tourism - Direction des Parcs Nationaux, des Reserves de Faune et de Chasse. "Document de travail", 1994). 34 Sources include farmers interviewed during the Rapid Poverty Appraisal, NGOs active in the area and Amnesty International reports. 35 For example, three villages in the vicinity of the Malfakasa Park have been abandoned because elephants routinely destroyed fields. Togo: Overcoming the Crisis, Overcoming Poverty 35 addition, because the protected areas were not well defined and borders were changed over the years, they often had to resettle more than once.36 Those who moved to the central part of Togo and to the Maritime region found it difficult to integrate into host villages, especially because of problematic access to land --which sometimes resulted in prolonged disputes (see Box 3.7). (b) Government-Sponsored Resettlement. While the government showed no interest in becoming involved in resettlement issues arising from the creation of national parks, it played a very active role in the resettlement programs carried out by two donor- financed development projects: the FED-Agbassa project and the Nangbeto dam. But in spite of considerable resources and efforts spent to facilitate the relocation process, it is clear that many settlers feel themselves caught in a downward spiral, and in the case of the FED-Agbassa project many are reported leaving the area to go back home (see Box 3.6). Box 3.8: Planned Resettlement Schemes The FED-Agbassa Project The FED-Agbassa project aimed at relieving demographic pressure primarily in the Massif Kabye by moving farmers to an area located between Niamtougou and the Kara river (Bassar district). Between 1974 and 1982, 922 families were settled on 5-ha. parcels and were provided with partially delineated fields, some housing, good access to rural infrastructure and services, and free fertilizers for the first five years. Pcels were assigned to each household head according to the provisions of the nadonal land reform (1974) and to the judgment of project staff, giving settlers usufruct rights. But indigenous Lamba people retain the ultimate right to the land and can reclaim it from settlers if they need to expand their holdings, while settlers have no right to additional land regardless of the size of their household. Lack of formal agreements makes land tenure issues all the more ambiguous and increases tension between settlers and host population The Nangbeto Dam Twentywone villages were affected by the construction of the Nangbeto dam on the Mono river for a total of about 26,000 people (26% more than estimated initially), and 7,600 persons were resettled between January and March 1987 in nine villages within a distance of 70 km from their original homes. Tbo resetlment program provided the displaced populaton with new homes, water points, schools, a health center and a road network. In addition, they were given compensation for declared losses, support for agricultural production and food assistance. But seven years after the project, the area is fraught with problems. Those who could not demonstrate their land rights because they lacked a title or had only customary use-rights found themselves worse off than before; some people received their compensation three years late (e.g., in Assat); and some buildings are falling apart because they were built on land subject to flooding (e.g., the school in Assantd). Of great concern to local authorities is also the tension caused by the influx of an estimated 1,000 clandestine fishermen (pecheurs de brousse) who are quickly depleting the fish-stock in the lake formed by the dam. Source: Adapted from World Bank, Togo Couny Envirownentat Stratgy Paper, draft, March 1995. 36 For example, the creation of the Keran National Park caused a first migration in 1971 from Namoute and Komte to Ossakre and a second one in 1977 from Ossakre. Togo: Overcoming the Crisis, Overcoming Poverty 36 (c) Displacement for Mining Purposes. The phosphate mines in the Vo district are a major source of foreign exchange whose exploitation and marketing are in the hands of OTP (Office Togolais des Phosphates). Despite its positive impact on the district economy, phosphate mining has caused considerable damage to the population living in the immediate vicinity of the deposits, who have been forced to resettle elsewhere while their lands were taken over by OTP. Six villages were displaced for the Hahotoe mine and five for the Kpogame mine.3" To compensate displaced families-- essentially farmers--OTP offers housing assistance on the basis of the estimated value of the dwelling 'hat had to be abandoned: homeowners are given 10 percent of the estimated value of the house to pay for labor and the rest is given as construction material. But no provision is taken to give these farmers land to cultivate. Instead, OTP has adopted a system whereby a compensation is provided for each M2. As a result, in 1992 many expropriated farmers had an annual income of about 8,000 CFAF (about US$30) and no way to earn a living. Given the severe land scarcity in the area, it is easy to see how vulnerable these families have become (see Box 3.9). Box 3.9: Visit to the Phosphate Mines l Where there are the phosphate mines, one can see artificial hills and holes. OTP has tried to fill up some of the holes near Akoumape, and the DRDR [regional Rural Development Ministry] has planted them with eucalyptus [800 ha. as of 19941. In Toka, OTP is presently moving the whole village to the east. Note that the new OTP emigration center will not provide fields for these people who essentially make a living out of agriculture. To compensate farmers for the loss of their fields, OTP has chosen a system which consists in evaluating (lost) crops by the square meter. As a result, many expropriated farmers find themselves with an income of about 8,000 CFAF for a year. We realize that families living in these conditions cannot survive. They become welfare cases unable to feed themselves and to finance their children's schooling. As a result, still today people die of hunlger in this mining zone considered the heart of the national economy. Enrollment rates in the areas occupied by OTP are very low compared to other villages." Source: CERAD, "Prefecture de Vo - Rencontre Bilan de Togoville", mimeograph, Lome, 7 August 1992 (p.34). (d) Violent Civil Strife. When the political crisis sparked a wave of tribal backlashes against northerners living in the south, about 173,000 people were reportedly forced to leave their homes and returned to their land of origin. Among the expulses, about 47,500 belonged to the protected areas in the north, and went back to them.38 This sudden influx of people represented an enormous burden on relatives who hosted 37 The diplaced villages for the Hahote mine were: Adanlekpudzi, Kpodzi, Nekome, Ablovime, Ameganpe and Gbedzekome. The diplaced villages for the Kpogamn mine were: Adzreve, Adzrego, Sokome, Tansi and Tsame; ,hwo other villages (Akadjame and Agome) should have also been displaced, but the low quantity of phosphate in that area did not warrant such a measure. 39 Agence de Solidarite Nationale, "Programme d'assistance aux expulses victimes de troubles socio- politiques et affectds par les problemes de la faune", document de travail, 1994. Additional information was obtained from field interviews in the Savanes and Kara region, as well as from the Regional Directorate of Social Affairs in Kara (November 1994). Togo: Overcoming the Crisis, Overcoming Poverty 37 them, and on government services mandated to help them. Those going back to the protected areas found their fields devastated by the passage of wild animals and no social infrastructure, while most of the others had difficulty in gaining access to land because of the high demographic pressure. All needed considerable assistance at least for the first harvest season (tools, seeds, fertilizers). It appears now that many families have gone south again, in part because their integration in the north was problematic and in part because of the conciliatory attitude of many southern traditional chiefs. For example, an official reconciliation ceremony took place in the Kloto area (Kpalime district), where northern labor force is needed in cocoa plantations. 3.18 Girls on their own. As a matter of survival, most children in Sub-Saharan Africa must contribute to their own upkeep by working--generally by helping their families in informal sector activities in urban areas and agricultural work in rural areas. But there are cases in which children end up struggling for their survival alone. Street children can be found in urban centers, especially in market areas. According to NGOs working with them, such as Terre des Hommes and CLORED (Centre de Loisirs et de Reeducation pour Enfants Defavorises), they can be divided into two groups: those who spend the day in the street but have a place to sleep, and those who sleep in the streets and often have been completely rejected by their family. Street children survive by perforning a multitude of little jobs, often working as street hawkers and, for girls, as porters (coltineuses or portefaix, see Box 3.10). In addition, occasional prostitution is common, especially in cities, as it is a means to afford "large" cash expenditures, for example for new clothing. FAMME, an NGO concentrating on women, estimates that as many as 60 percent of girls in Lome engage in occasional prostitution. Box 3.10: The Coltineuses of Lome The coltineuses are girls who transport the merchandise of traders from bus terminals to the market or from warehouses to sale points. In addition, they carry heavy parcels from the market to the buyer's means of transport. Often they also help unload trucks. They go to the market or the bus stations early in the morning looking for clients, and they stop at about 6 or 7 pm. They work barefoot because that way they can walk faster; incomes vary greatly depending on the luck of the day, but on average daily earnings are between 400 and 600 CFAF (about US$1). Almost all of them skip lunch, to save money and maximize the time spent working--or looking for work; for dinner, they eat kom (maize porridge seasoned with hot peppers and a bit of tomato sauce). They tend to sleep in parking lots or bus stations near the main market (Grand Marche) in groups of forty or sixty, paying between 25 and 50 CFAF per night; many of tham complain that, in spite of the money paid to a guard, they feel unsafe and rapes are not uncommon. Source: Adapted from Essodina M. Abalo, 'Analyse descriptive du phenomtne des jeunes filles en situation difficile dans la ville de Lom6", mimeograph, UNICEF, March 1992. 3.19 The majority of children on their own are girls from poor rural families sent to work in urban households as domestic servants (bonnes); their number is estimated to be Togo: Overcoming the Crisis, Overcoming Poverty 38 easily over 50,000 in Lome alone.39 This phenomenon is quite common throughout Africa and is certainly not new,40 but it has been suggested that the continuing economic crisis --and especially the devaluation-- has intensified it. Girls as young as six or seven are sent to stay with richer relatives (especially if the girls are very young), acquaintances or even with total strangers. The link between the child's family and the employer tends to determine the living conditions of the girl: the closer the link, the more likely the child is to be treated like the host family's own children, that is, sent to school and properly cared for.4' Indeed, in many cases parents are led to believe that their child will have a better life in the city, and the embellished reports of those coming back may lend credibility to the claims of professional intermediaries. Often, however, the bonnes are simply unwanted children, either because their families are too poor or because they come from broken families and the traditional safety net is unable to absorb them. A survey conducted for UNICEF in 1992, for example, found that in 95 percent of the cases, the main reason for sending girls to be servants in town was the low profitability of agricultural activities and the resulting poverty. In Lome the majority of them came from the south-east Maritime region (district of Vogan in particular) where demographic pressure makes it very difficult for a family to have enough land for subsistence; girls from the north are more likely to be sent abroad, to Burkina Faso, Ghana and Nigeria.42 3.20 Servant girls are generally employed in a variety of domestic chores, but they may also be used, in part or exclusively, for lucrative activities. For example, they may be used as servants in fuiu-bars, where they wash dishes, pound fif and prepare akume spending all day in a smoky and hot environment. Regardless of their chores, they tend to work long hours (12-18 hours a day), sometimes receiving assignments beyond their physical capabilities, such as walking all day in the streets with a heavy load of merchandise to peddle and no food in their stomach. They frequently have no annual leave and no weekly day off. Their salary is around 3,000-4,000 CFAF a month (about 39 Assuming that civil servants levels A2 and above have at least one bonne, and that richer traders also have domestic help, would produce an estimate of about 50,000 (Directorate of Employment and Labor, personal communication, February 1995). On the other hand, a 1992 study found that one out of three families in Lome had a domestic aid, putting the figure at about 25,000 (Essodina M. Abalo, "Analyse descriptive du phenomene des jeunes filles en situation particulierement difficile dans la ville de Lome (Togo)", mimeograph, UNICEF, Lome, March 1992). 4OFor example, a 1977 survey found that 10 percent of migrant girls aged less than 10 years old did not live with their parents. This phenomenon may help explain Lome's negative masculinity rate, a peculiarity among African cities. See: L. N. Assogba, "Statut de la femme et migration urbaine dans le Golfe du Benin: de la decision a l'insertion" in Cahiers Qudbequois de Demographie, 21(1):121- 49, 1992. 41 A survey conducted in Benin, a country which shares the same phenomenon, found that the enrollment rate for children living in households headed by a close relative (parent, aunt, uncle) was 86.2, against 25 for children living in households headed by non-relatives. See: INSAE, "Suivi des caracteristiques et comportements des menages et des groupes vulnerables en situation d'ajustement structurel 1990- 93". UNDP: Cotonou, November 1993. 42 E. Abalo, 1992, op.cit. Togo: Overcoming the Crisis, Overcoming Poverty 39 US$6), depending on the age, the experience and the inclination of the employers.43 Often, however, girls are just given board and lodging, making them affordable even to average urban earners. 3.21 While by being on their own some of these girls may be able to escape the abject poverty or severe repression to which they would be subjected at home, in the majority of the cases they are likely to enter adulthood in a very vulnerable state. Separation from their family and village of origin will not only leave them with deep emotional scars, but will also deprive them of the traditional wisdom that can be learned from elders and of the support of a strong solidarity network, while the treatment received in their employers' house--where insults, threats and corporal punishment are generally not out of the ordinary--will lower their self-esteem. In addition, the bonnes will not be given the opportunity to learn basic skills required for survival in a rural setting; yet most of them will go back to their village to be married and will have to provide for themselves with agricultural work. Thus, these girls may be robbed both of their present, by being heavily exploited, and of their future, by being denied the chance to learn the survival skills they will need when they grow up. 4. INCREASING THE POOR'S PRODUCTIVITY: FACTORS OF PRODUCTION 4.1 Raising the incomes of the poor is necessary for poverty reduction. This can be done in two ways. One is to change a country's income distribution so that it favors those in the lowest income brackets: the pie remains the same (no economic growth), but the poor have a bigger slice. The other way relies on economic growth: the pie becomes bigger and therefore all the slices become bigger, including the poor's slice. In the case of Togo, it is clear that these two approaches have to be combined because the pie is too small and the poor are too many and too poor for either redistribution or growth to suffice alone--the whole economy has to grow and the poor's relative share in the economy has to grow as well. The key to achieving this goal is to increase the poor's productivity. This is why this chapter concentrates on the factors of production: land, labor and capital. Land 4.2 As the vast majority of the poor are farmers, improving the poor's productivity depends on their access to potentially productive land. First and foremost, productive land has to be available for exploitation; second, the poor must be given the right to exploit it and enjoy the product of their labor. Access to land, therefore, has to be explored in two steps. First, it has to be determined whether potentially productive land 43 Personal communication of E.K. Vignikin (PhD), Economist, University of Benin, Lome, Department of Demography. Togo: Overcoming the Crisis, Overcoming Poverty 40 is available for exploitation, hence the need to address issues of rural population density and land scarcity. Second, it has to be determined to which extent the poor are able to gain access to available land, hence the need to address land tenure issues. 4.3 Population Density and Land Scarcity. Togo is one of the most densely populated countries in Africa. The Department of Statistics estimated for 1990 an average population density of 61 persons per kmn2, up from 48 persons per km2 in 1981. In comparison, the average population density for all West African countries in 1990 was only 32 persons per km2, and for Africa it was 21 persons per km2. The national average hides very marked differences between regions and, within regions, between districts (prefectures) and between zones within districts. For example, the Maritime region has a disproportionately high density, largely (but not exclusively) accounted for by Lome, while within the Kara region there are sparsely inhabited areas, such as the Bassar district with 23 inhabitants per km2, but density in the mountains north-east of Kara city (the Massif Kabye) reaches 300 pers/kMn2. 4.4 The existence of areas with low population density creates the impression of a vast empty space open to agricultural exploitation, hence representing some sort of safety valve for overcrowded areas. The FED-Agbassa resettlement project is an example of such thinking (see Box 3.6). What is seldom taken into consideration, however, is that some of the uncultivated land is not suitable for cultivation and current agricultural practices require long fallow periods if the land is to maintain its fertility. For example, in 1981 only 17% of the land in the Dapaong district was cultivated, but it corresponded to 70-80% of the cultivable land leaving very little chance for long fallows;4 since then population pressure has increased. In addition, uncultivated areas are not necessarily unused: they serve for grazing, fuelwood collection and hunting, so that putting these areas to more intensive forms of use may increase agricultural productivity but interfere with other economic activities. 4.5 One way to assess whether there is enough land available for the poor to cultivate is to estimate the rural density threshold, which is the maximum number of rural inhabitants that the land can support without fertility loss (urban areas are excluded for obvious reasons). If the actual rural population density is higher than the rural density threshold, it can be assumed that the poor will be unlikely to dispose of enough land to cultivate. Given Togo's agro-ecological conditions, this threshold was tentatively calculated by FAO at 65 to 85 rural inhabitants per km2 of farmland, corresponding to 0.4 to 0.5 ha. of cultivated land per person and a rate of 2 ha. of fallows for each ha. under cultivation."5 On the basis on this estimate, it appears that the rural density threshold is 44 ADESDIDA, Dossier Togo - Le Developpement de la Region des Savanes, Dapaong, June 1981. 45The carrying capacity of land, and therefore the rural density threshold, depends on a number of factors, including soil quality, rainfall, plant cover, the use of fertilizers, cropping patterns and cultivation practices. These may vary greatly within a small geographic area, making nationwide estimates extremely problematic and often misleading. The estimates provided, therefore, should be interpreted with caution. Togo: Overcoming the Crisis, Overcoming Poverty 41 quickly being reached throughout the country and by the year 2000, all regions but one will have a rural population that cannot be sustainably supported by the environment under current agricultural practices."6 The one exception is the Kara region, thanks to the low population density in the Bassar district (see Table 4.1), while the most dramnatic situation is in the Savanes and Maritime regions. Table 4.1: Rural Population Density and Pressure on Land (% over/below rural density threshold) Region Rural inhabitants/kml of available cultivable land 1970 1981 1989 2000 Maritime III 142 168 214 (+185Yo) Plateaux 39 55 69 97 (+29Yo) Central 20 32 46 79 (+5°/) Kara 44 51 56 64 (-15Y) Savane 69 93 115 155 (+107%) TOTAL 49 65 80 108 (+44%) Source: Staff estimates and FAO, 1991. 4.6 In some densely populated areas fallows have virtually disappeared, as land scarcity and/or low yields leave farmers with little choice but to use all the land they have at their disposal in an attempt to grow enough food for their family. This is the most common way for the vicious circle of poverty and environmental degradation to start. Because of overexploitation, soils lose their fertility and as a result yields decrease. Low yields, in turn, make it impossible for farmers to buy fertilizers which could help the land regenerate. In this manner, yields in poor and impoverished areas tend to get into a downward spiral, until the better topsoils have washed away. Particularly in the North (e.g., the T6ne district), a continuation of this process may lead to desertification due to the slow regeneration of plant cover --a trend which is very difficult to revert. Land scarcity can be a powerful incentive to the adoption of more intensive agricultural techniques, but it is unlikely that significant intensification will take place at a sufficiently rapid pace to affect estimates for the year 2000. Togo: Overcoming the Crisis, Overcoming Poverty 42 4.7 Land Tenure. The problem of land scarcity is exacerbated by a complex land tenure system in which traditional and modem law coexist and compete. Customary land tenure is regulated by three principles: (a) land belongs not to an individual but to a collectivity--either a clan or an extended family; (b) the right to exploit the land (usufruct right) is given by the collectivity to an individual who is part of that collectivity (generally a male head of household) and can be maintained as long as the land is Box 4.1: Land Tenure in Kantindi used; (c) those who are not part of tne land-owning clan or family may be given -All land belongs to the clans that first settled the area. permission to exploit the land (droit de Other clans are clans non-proprietaires, which include 37 culture), often in rental or sharecropping percent of households. They are allowed to use the land arrangements, but their right can be of the clan proprietaire, but in return they cannot refuse revoked at anytimeandcertaincropsare invitations made by households of the clan proprietaire revoked at any time and certain crops are for work parties. Until about 1960, newcomers to the not permitted (e.g., perennial or cash area could borrow land from the clan proprietaire. Land crops). has become increasingly scarce since 1960, so this is no longer possible. Resident households belonging to clans 4.8 The application of these three non-propridtaires without permanent rights to land find it prin shas important consequences in increasingly difficult to obtain an extra field. At most, principles ofproductivity and povert needy households can borrow only one small plot, and for terms of productivity and poverty a short time. The user retums this favor with a reduction. The most important is that commensurate gif, ranging from a chicken to 1/3 of the people who are considered "foreigners" harvest. Conflicts over the use of this and other land by the community (often referred to as have increased in recent years. Conflicts also arise metayers or sharecroppers), that is, those between crop growers and livestock herders, because the whaeeland-owning animnals must pass along narrow tracks between the crops wvho are not members of the land-owmng to reach food and water." group, find themselves in a very vulnerable situation: not only they will Source: Leo de Haan and Annette Kruithof, "Kantindi, N. Togo never be able to own land (which cannot - Case Study" in H. Reitsma et al. (eds.), Coping with be bought), but they will also be at the Semiaridity, NetheTland Geographical Studies (p.83). mercy of land-owners concerning their right to exploit the land and the conditions under which land exploitation is permitted (see Box 4.1). Another important consequence is that parcels are becoming increasingly small47 and spread-out," making the adoption of modem agricultural practices more difficult. Women, whether belonging to the land-owning group or not, generally have no land tenure rights; they may be given a plot to grow food for their family, but they have no secure rights and cannot own land.49 47 Parcels are becoming increasingly small because of the need to distribute land equitably among all men of the land-owning clan. Between 1972-73 and 1982-83, the number of farm units increased by 31.44% essentially as a result of the subdivision of larger farms (--, "Contraintes et caracteristiques de l'agriculture togolaise", part I, mimeograph, 1988). 48A recent survey found that in the Maritime region, where population pressure on land is greatest, a farmer may cultivate up to seven separate small plots (quoted in UNDP, 1995, p.24, op. cit.). 49Because of the exogamic tradition (i.e., they marry outside their clan), they tend to be considered 'foreigners"--by their own family, because they will leave to go live with their husband's family, and Togo. Overcoming the Crusis, Overcoming Poverty 43 4.9 With the intention of "organizing a better distribution and a wiser use of land"," the government undertook a reform of the land tenure system in 1974. The right to own land was linked to its exploitation, and all land which had not been used for more than ten consecutive years could be appropriated by the state to form a national land trust (domaine foncier national). A land census came up with 669,037 ha. of undeveloped land which the state could in theory distribute to needy families upon demand. Unfortunately, such estimate appears to be unreliable as in some districts the area declared undeveloped was larger that the total area of the district, while in other cases local authorities used the land census as a way to regulate personal grievances.5' In addition, parcels identified as undeveloped were not clearly marked, thus opening the door to endless legal disputes. 4.10 Another state effort to make land available to poor farmers was the creation, in 1978, of planned rural development zones (ZAAP, zones d'amenagement agricole planifie').52 These areas were declared by the government as needed for works of public utility and were to be developed by cooperatives comprising the owners of the ZAAP land but also metayers, in case the original land-owners did not dispose of enough labor force to guarantee a full exploitation of the ZAAP. Metayers would be initially working as agricultural laborers and, after having paid their contribution (detracted from their salary), could become members of the cooperative. This, however, only entailed usufruct rights for "foreigners", as property rights to the land remained with the original land- owners. 4.11 Attempts to establish a modem land titling system have given very limited results, as the 1974 reform links land tenure security to exploitation rather than ownership, thus providing no incentive for titling. In any case, the registration procedure is long (twelve months at the minimum), complex (presupposes knowledge of the different bureaucratic steps involved) and expensive (a case study found the cost of the registration to be three times the value of the land to be registered),53 so that obtaining an official land title is out of the question for the majority of farmers. by their husband's family, because they come from another family. See: FAO, "La femme et la propriete fonciere", in Rapport du Seminaire National sur la Reforme Agro-Fonciere et lI 'nstallation des Femmes Agriculseurs, Kara, August 1990. 50 Ordinance no 12 of 6 February 1974. 51 Kafui Adjamagbo, "L'acces a la terre en milieu rural au Togo: les grandes orientations de l'action etatique", paper presented at a conference on "Le droit et l'acces & la terre en milieu rural", Abidjan, 23-25 January 1989. 52 Decree 78-46/PR of 17 May 1978. 53 In the Wawa district, it was found that it cost 102,250 CFAF to register a l-ha. parcel bought at 30,000 CFAF. See: Confere Foli Massanvi and Charles Bonissol, "Principes et realit6s de l'intervention des operateurs non coutumiers dans le domaine foncier", mimeograph, Lome, September 1987. Togo: Overcoming the Crisis, Overcoming Poverty 44 4.12 Some twenty years after the land tenure reform, it is clear that "the laws have remained on the books" and access to land is still regulated by customary right. Such phenomenon is generally ascribed to the fact that the reform was imposed top-down by the government without taking into due account cultural traditions. In particular, some saw in it a political motivation because they interpreted it as a government effort to help land-scarce Kabye farmers obtain land (especially in the Plateaux region). The result is a complex situation with considerable negative consequences. Land disputes are frequent, and there are reported cases of people having sold their land according to modem law without fully realizing the implications of the transaction. Lack of secure land tenure rights demotivates metayers from making long-term quality investments in the land they cultivate, and customary law actually prevents them from planting trees--which would help control soil erosion. Labor 4.13 The Structure of Employment. Togo does not have a recent labor force survey and, given the major changes that occurred in the economy during the early 1990s, it is impossible to estimate with precision the present distribution of the labor force. In 1983, agriculture employed 72 percent of the labor force, the informal sector 21 percent and the modem sector only 7 percent. It is reasonable to assume that the agricultural sector still occupies about 72 percent of the labor force, although during the strike period this percentage went up considerably. Farmers are predominantly small-holders growing food crops using traditional methods. The typical agricultural productive unit is composed of a household head, generally a man, his spouse(s) and their children; other members of the extended family may also be part of the productive unit, and shuffling children between members of the same extended family is not an unusual way to keep a certain flexibility in labor supply. Cash crops have gained in importance during the 1980s, especially cotton which can be grown in most of the country, to the point that in 1987 it was estimated that half of all farm families produced cotton. 4.14 Modern sector employment, both public and Table 4.2: Changes in Employment in the private, is concentrated in the Modern Private Sector, 1982-89 Maritime region. Public sector employment, which represents Sector 1982-1989 1985- 1989 three fourth of modem Agriculture +20.1% -19.1% employment, is heavily Mining +40.7% +15.8% concentrated in the region because Manufacturing -38.3% -13.2% of Lome. As for the modem Public Works -44.1% -36.1% private sector, 88 percent of the Water, Electricity +41.4% +4.6% labor force and 85 percent of the CTransportankg +2.3% +.48% enterprises are in the Maritime Services -8.5% +1.6% region. Employment in the Source: CNSS. government sector has declined since 1983, with the number of civil servants remaining almost unchanged and the Togo: Overcoming the Crisis, Overcoming Poverty 45 number of permanent government employees declining by 25 percent (see Table A4.1 in Annex 2).54 At the same time, although modem private sector employers augmented, the number of employees per enterprise diminished, indicating a movement toward smaller, more flexible production units but also a decline in employment opportunities. The sectors more affected by this decline are public works, manufacturing and transport (see Table 4.2). It is certain that the economic crisis of the 1990s has greatly affected the modem private sector, as many foreign and local investors left the country, thus further diminishing employment opportunities. 4.15 Although official estimates put informal non-agricultural employment at about 21 percent, there is reason to believe that it plays a far greater role in the economy. In the rural areas, in addition to an estimated 16 percent of artisans (on the basis of the 1989 household budget survey), most women have a secondary activity in the informal service sector, such as petty trade or production of choukoutou and chakpalo (millet-based fermented drinks). In the urban areas, the majority of the population eams a living thanks to informal activities. In Lome, where formal employment is likely to be at its highest because of the concentration of government jobs and the modem private sector, only one fourth of the population is formally employed (see Table 4.3). What is more, many of those formally employed, especially low-level civil servants, have part-time informal activities on the side to help them make ends meet--a phenomenon that has been exacerbated by the recent economic crisis. Hence, it can be inferred that at least 80 percent of the Lome work force is active in the informal sector. Similarly to what happens in rural areas, the majority of informal sector operators in urban areas are women (72 percent in Lome, according to Table 4.3), and they are mostly involved in trade. Table 4.3: Estimated Number of Workers by Sector and Gender, Lom6, 1988' Male Female TOTAL Formal Sector 29,572 6,824 36,396 (25%) Informal Sector 26,705 68,817 95,522 (65%) X retail trade 3,847 62,668 66,515 * artisans 15,408 6,149 21,557 * taxis 7,450 - 7,450 Domestic Servants 7,981 3,240^ 11,221 (8%) Agriculture 2,829 664 3,493 (2%) TOTAL 67,087 79,545 146,632 (100%) Source: 1987-89, Household Budget Survey. Excluding apprentices. This is likely to be an underestimate, as the majority of domestic servants are girls who would be regarded as apprentices and therefore excluded from the count. 4.16 Wages. The official minimum wage (SMIG, salaire minimum interprofessionel guaranti) is set at 13,757 CFAF (about US$27) per month, one of the lowest in the CFAF zone and well below the poverty line for households of average composition and with no other sources of income. Relatively few people in the formal sector, however, are paid so 54INF, "Togo - Recent Economic Developments", SM/91/151, August 1991. Togo: Overcoming the Crisis, Overcoming Poverty 46 little, as wages are generally set as a result of Collective Bargaining Agreements for different occupational or skill categories, so that each category has its own salary grid." For example, the lowest paid civil servant, a doorman (planton), has a salary of 18,260 CFAF. Salary grids are based on degrees and, as such, reinforce the emphasis on degree- based education rather than competency-based training linked to market demand. They also constitute a disincentive to high-level performance, since, no matter how competent employees can show themselves to be, they will not be able to receive salaries higher than allowed by their formal education. 4.17 While only a small portion of formal sector employees have a monthly wage as low as the SMIG, the majority of informal sector employees earn considerably less than the minimum wage. For example, girls working as domestic servants are paid as little as 2,000 CFAF a month (plus board and lodging) and shop assistants are paid between 6,000 and 10,000 CFAF. Despite a legal minimum wage below the poverty line, it is believed that many enterprises choose to remain in the informal sector to avoid social security taxes and regulations which lead to high labor costs. Efforts to keep labor costs low and an ample supply of labor result in two widespread practices. One is the use of apprentices, who are either unpaid or paid very little. The other is the integration of formal and informal business, whereby modem enterprises sub-contract to informal enterprises or use undeclared daily laborers for unskilled tasks. 4.18 Scattered evidence suggests that daily salaries for unskilled work start at around 200-300 CFAF, with typically female jobs being paid less than typically male jobs. Street hawkers (mostly girls) in Lome are paid about 200 CFAF/day; kitchen helpers in urban restaurants (almost exclusively women) are paid 250 CFAF/day to poundfoufou; in the Plateaux region, women are paid 100 CFAF to transport a 50 kg. bag of rice for 6 km (and come back), resulting in 300 CFAF for a full day. The typical rural wage is reported to be 500 CFAF, but there are considerable regional differences. In the Savanes, agricultural laborers are paid 200-300 CFAF/day to pick cotton, while in the Kara, the Konkomba (known for being a wealthy group) pay Ghanaian cotton pickers up to 1,000 CFAF/day. Laborers working for the Chinese sugar cane plant near Atakpame are paid 500 CFAF/day. 4.19 The Dynamics of Employment For the past twenty years the evolution of employment paralleled economic performance, growing substantially during the 1970s and then declining (although not steadily) in the 1980s. Migration patterns during this period suggest a shift from the primary to the secondary and tertiary sectors. Thus, the average annual growth rate of the urban population was 5.7 percent between 1985 and 55 On the basis of the household budget survey in Lome, it was calculated that the household of an individual earning the minimum wage and responsible for two persons (corresponding to a labor force participation rate of 33 percent) would have fallen in the lowest income decile. This suggests that few workers earn as little as the minimum wage (see: Annice Brown, M'Hamed Cherif and Peter Richards, "Analyse de la Problematique de I'Emploi au Togo", mimeograph, September 1990, p.23). Togo: Overcoming Jhe Crisis, Overcoming Poverty 47 1990, while for the rural population it was only 1.8 percent.56 The main reasons behind the rural exodus are the persistent low level of revenues for rural labor, and the increasing land shortages (especially in the South) and environmental degradation. There are also social causes, such as the refusal of educated people to do farm work and young people's desire to break free from traditional parental authority. Besides, when the future is certain to be a daily struggle to satisfy even the most basic needs, it is easy to see how rural dwellers may feel attracted by the opportunities offered by a city. 4.20 The rural exodus was temporarily reversed during the period of greatest insecurity (1991-93), when people fled the cities and went to the countryside in search of peace--and of a way to make a living. For a couple of years, the employment structure changed drastically, with the secondary and tertiary sectors declining--especially the modem parts of these sectors--and agricultural employment growing, as city dwellers converted themselves into subsistence farmers. The evolution of GDP by sector reflects this pattern: between 1991 and 1993 the agriculture share of GDP rose from 33 percent to 46 percent, while the secondary and tertiary sectors declined by 36 and 32 percentage points, respectively. With the end of the strike and increased security in urban areas, and in the south in general, it appears that most people have gone back to their homes. 4.21 Given the demographic pressure on land, continued high fertility rates and very limited opportunities for off-farm employment, it is likely that the rural exodus will continue. Unless there is a concerted effort to develop profitable farm-related activities in rural areas so as to satisfy employment demands, rural people will have no choice but to turn to the other sectors. Clearly, the formal sector will not be capable of absorbing significant numbers of job seekers: the need for fiscal discipline makes a growth in public sector employment practically impossible, while the private sector has proven incapable of expanding in the past ten years and in any case will take time to recover from the crisis. Already before the crisis it was estimated that at best 4 percent of those entering the labor market each year would find wage-employment in the modem sector (i.e., 2,000 out of 50,000)." Hence, the informal sector will continue to be the only hope for expanding employment opportunities and for containing the cost of unemployment. 4.22 While firm data on unemployment are unavailable, 1991 estimates placed the rate of unemployment in the range of 16-28 percent58. Since then, the situation has certainly worsened because of the economic crisis. For example, by extrapolating from studies conducted in the 1970s, it has been tentatively estimated that the unemployment rate in Lome is now likely to be close to 40 percent.59 Although Lome is believed to have the greatest proportion of job seekers in the country, the employment picture outside Lome is 56 EuroStat, Statistisches Bundesamt, Country Profile: Togo, 1991, p.32. 57 World Bank, Republic of Togo - Skill Formation in the Informal Sector, Report no. 10447-TO, 1992. 5S IMF , op.cit. 59 Department of Employment and Labor, personal communication, February 1995. Note that the 1989 household budget survey suggested a rate close to 10 percent, i.e., at least 20,000 individuals. Togo: Overcoming the Crisis, Overcoming Poverty 48 probably not much rosier. To begin with, unemployment is, in a sense, a luxury: in a country without a formal safety net for the unemployed, the poor cannot afford to be unemployed, because they have to find a way to provide for themselves. Under- employment is also a measure of the imbalance between labor demand and supply, and takes the place of unemployment. Both in urban and rural areas, under-employment is very widespread. During the dry season, most male farners are basically idle for at least two or three months. In urban areas, many informal sector workers, such as porters and dockers, only work for a few hours a day and spend the rest of the time looking (or waiting) for work. 4.23 The freeze on new hiring for government jobs in force since 1983 has severely limited the employment opportunities of high school and university graduates, who previously had found easy access to a government job. On the other hand, as mentioned before, the private modem sector cannot possibly absorb large numbers of graduates and, in any case, it tends to prefer employees with informal training because they are willing to work for lower wages. The result is that there are now some 21,000 unemployed graduates (jeunes dipl6mes sans emploi), whose only realistic hope to find a job is to enter the informal sector, possibly becoming self-employed (see Box 4.2). Box 4.2: I Have to Sell Peanuts "Whenamy father was made ajeune retraite, back in 1985, I was finishing my university studies. It was a big blow for the whole family, but I thought that with my university degree I could have helped the family. Nothing was coming along, so I started working with my mother, who is a tailor. I'm lucky she taught me how to sew, because that way I can earn some money, and sew my own clothes. Then my parents moved' to theAillage,4where life is cheaper, but I wanted to remain in Lome because here's the only place where I can fmd ajob using my qualifications. I stay in the house that my father had started to build, with two brothers in school to support. I stay there because I couldn't possibly afford to pay a rent, but the house isn't finished: there'.s only one room with the roof, and no floors. So I'm losing all my customers, because they say my house is dirty and their dresses will get dirty. I can't blame them, really: I can't even work myself in such conditions, I don't even have a clean table on which to lay the cloth to cut. Sometimes I'm lucky and I find work with a donor or an NGO. They say I'm good, they thank me, but they have no long-tenn job to offer me. When I work for them, I'm able to pay off some of my debts, or to buy books for my brothers, or to do some more work on the house. Now I have to sell peanuts on the side of the street, making 600 CFAF [slightly more than US$I] a day if things go well. At first I was ashamed to do it, and hoped nobody would recognize me. And you know what? By staying out in the street all day I saw many of the people who were at the university with me. They also have to do these little jobs to survive." Source: Rapid Poverty Appraisal, unemployed female graduate in Lome, February 1995. 4.24 In an effort to help unemployed graduates, the government launched in 1991 an employment-training program (PEF, Programme Emploi-Formation). Through the PEF, a total of about 2,700 graduates were hired by the government for one year--with the option of renewing for an additional year--at half the salary of regular employees, because their salary was considered a "training scholarship". Almost all of PEF recruits were assigned to ministries (only about 50 went to the private sector), with the majority (52%) going to teach, primarily in elementary schools. As of 1995, PEF participants are still in their job, as it is felt that their services are needed because they help off-set the personnel shortages caused by the government hiring freeze. Their salaries were initially Togo: Overcoming the Crisis, Overcoming Poverty 49 paid almost entirely by bilateral donors (90 percent by Germany and 5 percent by Canada), but since 1993 they have been paid by the government as part of the investment budget. 4.25 The basic parameters controlling the use of labor are set out in the Labor Code of 1974, supplemented by the Labor Code of l952.60 The 1974 Labor Code is presently under revision. One of the planned changes is the removal of the monopoly enjoyed by the Employment Directorate (previously Employment and Labor Service) for job placement services (Art. 161), which was established to ensure equity in employment opportunities, but has become largely irrelevant as employers routinely by-pass it; during 1994, for instance, only about ten job-seekers were placed by this service.6" Another intended revision is permitting employers to dismiss workers for "economic reason" (Art. 42), a change essential for flexibility in enterprise restructuring. In any case, official regulations are only followed by a minority of employers, and therefore leave the majority of the workforce unaffected. Labor relationships are commonly governed by tradition and change according to location, especially for agricultural labor. Vocational training is the domain of informal apprenticeship arrangements typical of different guilds, often under draconian conditions. Apprentices may have to work as house servants as well, pay for their training and for their "liberation" ceremony. At the end, there is no guarantee of paid employment. Capital: Access to Credit 4.26 Access to credit is important for financing investment and, in turn, for boosting the productivity and income of the poor. In addition, it is essential to reduce vulnerability because it may help the poor cope with unexpected crises (death, disease, crop failure). Credit, however, has a price (the interest rate) and involves transaction costs and the risk of delinquency. These costs and risks are at the core of the problems the poor face in obtaining credit, and they are the main reason why commercial credit is seldom the answer to a poor person's needs. Transaction costs can be very high because they generally involve traveling to another community. For the majority of the poor, who live in the countryside and do not own a means of transport, such trips have both an opportunity and a financial cost, as several hours may be spent by the roadside waiting for a car or motorcycle willing to take them to their destination, so that a visit to a financial institution may easily take a whole day; also, several trips may be necessary, as the prospective borrower may not be able to satisfy rapidly administrative requirements. Besides, there are considerable psychological costs. Poor people tend to have little schooling and little assertiveness (powerlessness is a dimension of poverty), so approaching a formal financial institution can be very intimidating and entail a significant risk of losing face--a risk which many are simply not prepared to take. From the point of 60 The 1974 code states in Article I that the code replaces an earlier labor code adopted on December 15, 1952, but adds that all provisions of the earlier code that do not contravene the 1974 code remain in effect. 61 Director of Employment, personal communication, February 1995. Togo: Overcoming the Crisis, Overcoming Poverty 50 view of a financial institution, the risk is to lend money to individuals with insufficient collateral guarantees, which is almost inevitably the case with the poor. 4.27 Regardless of the costs and risks involved in borrowing money, information gathered with the rapid poverty appraisal leaves no doubt that having access to "regular" credit and to emergency cash are a priority for most people in Togo. As vulnerability tends to be highest among the poor, it is not surprising that the poor are particularly insistent in voicing their demand for this type of service. When asked about outside interventions that could help them get out of poverty, the most common answer was access to credit for buying fertilizers to use on food crops. In urban areas, credit was requested primarily as start-up capital to set up modest trade activities (petit commerce) or to be able to buy raw material without being exploited by whole-sale traders. 4.28 Formal Banking. The formal banking sector is used almost exclusively to finance the modem sector of the economy. When the National Agricultural Credit Bank (CNCA, Caisse Nationale de Credit Agricole) stopped its credit activities in 1988, seasonal credit for farmers became almost impossible to obtain and the already limited confidence of the rural population in the formal financial sector was further eroded. The eight banks and two financial institutions presently operating in the country are largely inaccessible to poor (and not so-poor) people, because of their location (21 out of 42 branch offices are in Lome and none are in secondary centers), their collateral requirements (they do not accept group guarantees), and the size of their loans (too large for the needs of most prospective borrowers).62 Apart from being unable to provide the type of financial services needed by the poor, the formal financial sector is still suffering from the loss of enterprise profitability and public sector arrears accumulated during the 1992-93 crisis and, despite the economic recovery started in 1994, new credits are considered with extreme caution (they fell from 109 to 103 billion CFAF between 1993 and 1994).63 4.29 Intermediary Financial Institutions. A number of savings and loans cooperatives have been set up, often with the assistance of donors or NGOs, to satisfy the credit need left unanswered by the commercial banking sector.' Although they employ 62 For example, the lowest loan given by the Societe Nationale d'Investissement (SNI) is I million CFAF. 63 World Bank, "Aide-M6moire de la Mission Relative a la Reforme du Secteur Financier", mimeograph, Lom6, March 1995. The impact of the 1992-93 crisis is aptly described as follows: "Institutions that were facing difficulties before the crisis have seen their position become untenable. Institutions that were in good standing before the events have also suffered from a severe deterioration of their flnancial position." (World Bank, "Togo: Financial Sector Review", mimeograph, January 1994, p.12). 64 The networks of savings and loans cooperatives include: FUCEC (Federation des Unions Cooperatives d'Epargne et de Credit), CPDE (Caisses Populaires pour le Developpement de l'Entraide Socio- Economique), and AVE (Association Villages d'Entreprises) active in both rural and urban areas; GIPATO (Caisse d'Epargne et de Credit du Groupement Interprofessionnel des Artisans du Togo) and CECA (Cooperative d 'Epargne et de Credit Artisan) active in urban centers. Togo: Overcoming the Crisis, Overcoming Poverty 51 different approaches and they vary considerably in terms of size and organization (many, but not all, are networks), they share some key characteristics: they operate relatively close to the grassroots, they provide short-term credit, they are willing to lend small amounts, they build on traditional mechanisms, and they aim for long-term sustainability through the mobilization of local savings. 4.30 The Federation of Cooperative Unions of Savings and Loans (FUCEC, Federation des Unions Cooperatives d 'Epargne et de Credit), is the oldest and by far the largest network of savings and loans cooperatives. It started in 1969 in the coffee and cocoa producing areas (south-west), but it is now mainly concentrated in the Maritime region (66 percent of the total credit) and almost absent in the Savanes region,65 it counts 153 branches (called COOPEC) and 42,000 members. Although FUCEC is undoubtedly filling an important gap, as demonstrated by its considerable growth in the past year,' the poor are unlikely to be able to benefit much from it because loans cannot exceed twice the size of the applicant's savings, and savings must have been deposited for at least six months before obtaining a loan. 4.31 Many of the existing savings and loans cooperatives contain their risk by limiting credit on the basis of the member's savings. This practice accounts for very high reimbursement rates (e.g., 89 percent for FUCEC, 100 percent for CECAVE)6' and encourages savings, but, as noted above, it also excludes those who are too poor to have any significant savings. As a result, some recent initiatives have adopted a group guarantee approach which, at least in principle, should make it possible even for the lowest socio-economic strata of society to have access to credit. 4.32 Informal Credit Systems. The only credit accessible to the vast majority of the Togolese, including the poor, is through informal channels. The preferred source are (richer) relatives or neighbors, who may be called upon in emergencies and who are more likely to be understanding in case of delinquency. Hence the common--and well justified--complaints of civil servants, who see themselves being perceived as banks by their extended family. The terms of the loan may vary considerably. Field work suggests that small loans are generally due with little or no interest, while larger loans, whether in cash or kind (e.g., a sac of millet), are generally due after the harvest with a 30-50 percent interest (see Box 4.3). 65 IRAM, "Etude sur le Perspectives et Modalites de Creation, dans la Sous Region UMOA, de Nouvelles Institutions Bancaires de Type Cooperatif et Mutualiste ('Banques Populaires')", mimeograph, March 1991. 66 FUCEC's "Solidarity Bank" in Lome has doubled its members in one year (8,000 in 1995) and its savings volume has reached 1 billion CFAF. Much of this growth is due to traditional clients of the formal banking system (e.g., traders) switching to FUCEC (see UNDP, 1995, op.cit.) 67 Seydou Bawa, "Les Competences des ONG en Matiere d'Epargne te de CrEdit au Togo", paper prepared for UNDP (mimeograph), Lome, 1991. Togo: Overcoming the Crisis, Overcoming Poverty 52 4.33 When credit is not B 4: : tial available from relatives and -My husbind doos not Vow enog fo6dV l!i neighbors, as it is often the case fields are too lile and too tired. L=ryAt yeawen In in poor communities (when April, so I had no choice but to borrow money tb everybody is poor, nobody can I could find it. I bomrowed about 10,000 CAawtraM I bought a bag of sorghum and' condiments for the u.e1 o afford to lend to others), it can b* minimbg of n ae t ~ mkmu be found with itinerant bankers that I just give thm back th money I owed teom t hue. (also called "yes-yes")6", Oths tat me 2,000 mnd they Metod me y usurers and other people who, bie. It depends onte people. I had to go6 tofsarby because of their work, dispose l villagej to find money and to buy food, b hee has anything to spare. I also belng to a tonidlthem ten oaf us, of a certain liquidity and come and we contribute 250 CFAF on muket days [once a wi, but into contact with low-income sometimes I cannot find money for the tomhw, so I cFoldn*t get groups, such as traders and anything from them whon I needed to buy food."- bush-taxi drivers. Credit from Sotucec Rapid Poverty Appaisal fume in north-east Km region, these sources tends to have ornl9. :: 7- i 5 0X:a; - November 1994. high interest rates, because of the emergency situations in which money is often needed and the virtual monopoly of the lender. For example, usurers are reported to charge interest rates reaching 360 percent, and the household budget survey found that, in large urban centers, interest rates on credit from private individuals were almost always higher than bank rates (i.e., over 20 percent).69 The household budget survey also found that a considerable proportion of respondents had no idea of the interest rate they paid on their loans, confirming field work findings that interest rates are far less important than quick access to cash. 4.34 Savings clubs or tontines are widespread throughout West Africa, and Togo is no exception. The most common type of tontine is the tontine mutuelle (also called so- dzodzo) whereby members--generally between 5 and 10--contribute an agreed amount of money at regular intervals, and each member in turn collects the whole amount. This means that members who collect the "prize" during the first rounds will benefit from a no-interest credit, while those collecting last will in fact simply receive their savings back.70 The sum contributed can vary considerably, from just 100 CFAF (US$.25) a month to over 5,000 CFAF a week, and so does the interval at which money is collected-- most commonly it is once a week, but it can also be once a month, or on market days. Other types of tontine may be set up with a specific purpose, and they function more like 68 A 1992 ILO study estimated itinerant bankers to be at least 160, serving each between 190 and 400 individuals. The savings collected this way were put at about 3.2 billion a year. [original reference? quoted in UNDP, op. cit.] 69 Direction de la Statistique, Telegramme Enquete Budget Consommation No. 8 - Le Crddit et I'Epargne des Menages Urbains, Lome, February 1992. 70 In addition to a rotating savings and credit mechanism, many tontines mutuelles constitute a separate credit fund by tking a percentage of members' contributions. This money can be borrowed by membersat generally low interests rates. Togo: Overcoming the Crisis, Overcoming Poverty 53 insurances. For example, the kugadzodzo is a savings club to finance funeral expenses, which, given the importance attached to a prestigious funeral, can literally bankrupt a family. 4.35 Although tontines continue to play an important role in affording poor (and not- so-poor) people much needed access to credit, they appear to have been considerably weakened by the economic crisis. Recent field work suggests that in the poorest areas of the country (e.g., Savanes, Kara, isolated areas everywhere) tontines are almost absent because nobody can afford saving anything, while in the richer Plateaux region only about one third of the rural population belongs to a savings club. Research carried out in neighboring Benin shows that, among vulnerable groups, tontine membership went down as much as 60 percent as a result of the devaluation, suggesting that the relatively low levels of membership found in Togo are likely to be the result of a combination of continued economic decline and the devaluation.7" 71 Igue, J., "L'evolution des conditions ds vie des groupes vutnerables au Benin apres la devaluation du franc CFA", mimeograph, Cotonou, March 1996. Togo: Overcoming the Crisis, Overcoming Poverty 54 5. RURAL DEVELOPMENT AND FOOD SECURITY 5.1 The vast majority of Togo's population, its labor force, and its poor are located in the agricultural sector. Eighty percent of the poor reside in the countryside, where the incidence of poverty is 2.1 times higher than in urban areas and secondary towns. Changes in the agricultural sector, therefore, have immediate effects on rural poverty, and, through rural-urban migration and resource flows, on urban poverty as well. In addition, problems in the agricultural sector have an impact on the tertiary sector because they determine rural demand for services and products, and because a large part of the tertiary sector centers around the handling of agricultural produce--transport, marketing, processing and export. 5.2 Rural development is essential for growth, raising living standards and reducing poverty. At the basis of rural development and poverty reduction is food security, a goal which is made all the more challenging by Togo's rapid population growth and simultaneous environmental degradation. This chapter reviews the impact of agricultural policies and interventions on rural development in general and on poverty in particular. Three questions guide the review: (a) to what extent did poor farmers' incomes increase as a result of policies and interventions? (b) to what extent did the poor's access to food increase? and (c) to what extent did policies and interventions result in agricultural growth? Role of agriculture in the national economy 5.3 Agriculture--including livestock, fishing and forestry--is the pillar of Togo's economy: it accounts for more than one third of the GDP, employs roughly 70 percent of the labor force, and provides about one third of the total exports. After a weak performance in the 1970s with an annual growth rate below 2 percent (compared to 3.8 percent for the economy as a whole), agriculture grew at an average annual rate of 5 percent between 1980 and 1990 (compared to 1.4 percent for the economy as a whole). Over the last decade the growth has been highly erratic, reflecting both the vagaries of the weather and the political instability of the 1990s. Farm output slumped during the 1991-92 agricultural year, but resumed vigorously in 1993 (19 percent growth), as city dwellers turned to subsistence agriculture in response to persisting insecurity in urban areas and a significant decrease in formal economic activities (see Table 5.1). Thus, the 1993 growth was almost entirely due to foodcrop production, while cash crops actually declined by 7 percent because of problems in distributing fertilizers and insecticides'. With a better security situation and the end of the national strike, people went back to the cities, and in 1994 foodcrop production declined considerably (by 16 percent). On the other hand, distribution of inputs improved, and cash crops experienced such a growth that they reached an all time high. I Although food crop production increased by 19 percent, the cultivated area increased by 30 percent, indicating that also the productivity of food crops as well was hurt by distribution problems for fertilizers and insecticides (data from DESA). Note, however, that 1993 statistics may not be fully reliable because of the eight months strike of government services. Togo: Overcoming the Crisis, Overcoming Poverty 55 Table 5.1: Contribution of Primary Sector to the National Economy, 1985-95 1985 1986 1987 1988 1989 1990 1991 1992 19933 1994 1995 GDP by Sector (Constant billions of CFAF) Prim. Sec. 124.2 126.0 126.3 138.0 145.6 149.9 148.2 150.9 173.6 162.8 172.3 Foodcrops 82.4 80.0 76.8 87.4 94.8 93.8 89.0 91.4 112.3 94.0 98.6 Cash crop' 9.6 11.5 12.5 12.6 12.5 13.4 14.5 11.9 11.1 14.8 16.9 Other2 33.2 34.8 36.3 38.1 39.0 42.7 43.8 47.4 51.6 52.7 55.0 % rate of change Prim. Sec. 7.6 1.5 .2 9.3 5.5 3.0 -1.2 1.9 15.0 -6.2 5.8 Foodcrops - -3.0 -.39 13.7 8.5 -1.0 -5.1 2.6 23.0 -16.4 4.9 Cash crop 37.7 18.8 9.2 1.1 -1.4 7.0 8.5 -17.7 -7.2 34.2 14.0 Other2 - 4.8 4.6 4.7 2.4 9.6 2.6 8.2 8.7 2.1 4.4 % sbare of GDP Prim. Sec 33.7 34.8 33.5 33.6 32.2 33.7 32.6 36.1 45.7 38.0 36.0 Foodcrops 21.0 22.1 20.4 21.9 20.9 21.1 19.5 22.6 28.1 21.4 20.4 Cash crop 3.4 4.3 4.4 3.5 2.7 3.0 3.1 2.5 3.5 5.0 4.8 Other2 9.2 8.4 8.7 8.1 8.7 9.6 10.0 11.0 14.1 11.5 10.9 r Includes: cotton, coffee, cocoa. 2 Includes: forestry, fishing, livestock and hunting. 3 1993 statistics should be interpreted with caution because of disruptions in data gathering and processing that year. Source: MPAT, Direction de la Statistique GCnerale; BCEAO-LOME; and staff estimates and projections. 5.4 Togo's agriculture can be subdivided into a food production subsector, which represents about 70 percent of the agricultural GDP, and an export subsector, which represents about 10 percent of the agricultural GDP; livestock, forestry and fishing cover the remaining 20 percent. Permanent crops--mostly cocoa, coffee and oil palm--cover only about 85,000 hectares and are concentrated in the western Plateaux, while seasonal crops cover annually about 1 million hectares. Maize, sorghum, millet and beans are the main food crops, followed by cassava, groundnuts and yam. Cotton is the main cash crop (83,000 hectares per year) and occupies 10 percent of the cultivated land. The national meat and fishery production covers only 60 and 30 percent of demand, respectively. Policies, Instruments and Poor Farmers' Incomes 5.5 Producer Prices and Trade Liberalization during Structural Adjustment. In the early 1980s, export of cash crops--cocoa, coffee and cotton--was controlled by a state enterprise (OPAT) and resulted in heavy implicit taxation of farmers, who received about one-third of border prices (see Table 5.2). Among the earliest initiatives taken during the structural adjustment period was raising producer prices for coffee, cocoa and cotton: between the 1983-84 and the 1986-87 harvest seasons, prices for these three crops increased by 29, 20 and 25 percent Togo: Overcoming the Crisis, Overcoming Poverty 56 respectively.2 These increases substantially augmented the real incomes of cash crop producers. They also generated a considerable supply response for cotton, which can be grown throughout the country (with the exception of the Maritime region) and which requires relatively little initial capital outlay--production went from 24,437 tons in 1983-84 to 79,065 tons in 1986-87, and in 1994-95 it reached 130,000 tons. Indeed, cotton growing came to represent the main chance to escape poverty for farmers in the impoverished northern regions. Table 5.2: Evolution of Producer and World Prices for Cacao, Coffee and Cotton, 1983-94 (CFAF/kg) Cacao Coffee Cotton Prod. Price' World Price Prod. Price World Price Prod. Price2 World Price 1983/84 275 985.10 290 1164.40 75 722.90 1984/85 300 868.70 315 1187.10 90 584.20 1985/86 330 738.20 365 1014.00 105 337.80 1986/87 360 628.10 400 680.00 105 270.40 1987/88 360 596.33 400 603.60 105 435.54 1988/89 300 600.00 350 554.68 95 363.90 1989/90 225 525.64 175 466.01 95 459.45 1990/91 250 492.64 175 247.64 100 436.99 1991/92 250 438.74 185 252.93 100 383.35 1992/93 225 338.40 140 248.40 90 362.50 1993/94 300 386.00 200 290.00 110 451.00 1 Producer Price. 2 This is untreated cotton bought by SOTOCO from farmers, while world prices are for treated cotton commercialized by OPAT. The difference between world prices and producer prices is in large part due to ginning costs. Source: OPAT, Lome. 5.6 A number of measures to liberalize the export of non-cash crops were also taken, culminating in 1990 with the elimination of all taxes on non-traditional exports (i.e., all agricultural products excluding cocoa, coffee and cotton)3. That year the value of exports of non- traditional crops jumped from .80 billion CFAF to 4 billion CFAF, but neither production figures nor producer prices show a significant increase, suggesting that such dramatic growth in the export of non-traditional crops was probably due to the fact that products previously exported illegally to avoid taxes entered official statistics in 1990. 2 One of the conditions for the latest World Bank structural adjustment credit (ERAC) is the liquidation of OPAT, which could result in further increases in producer prices. 3In 1986 the export of foodcrops to non-member countries of CEDEAO was authorized, although the export of cereals and tubers remained subject to the granting of an export license (which depended on whether these "strategic" crops were actually surplus to national requirements). In 1989, tariffs were removed from exports of food staples, but to clear customs a Certificate of Origin issued by the Trade Ministry was needed. In 1990, all taxes on non-traditional exports were eliminated. Togo: Overcoming the Crisis, Overcoming Poverty 57 5.7 The direct impact of these measures on the income of the Box5.1: WhyProducerPriceStyLow rural poor is likely to have been small. One reason is that many of "Producer prices remain low for many reasons. Farmers do not the poor are not surplus producers want to go back to their village with a heavy weight on their boad, so when they get to the market they want to sell at all costs. On top and, even in the presence of of it, they're often women who have many things to do: they may incentives to increase production, be more interested in getting back home fast than in getting the baet do not have the means to respond price. What often happens is that traders intercept farmers on their because of lack of labor, capital way to the market and offer to free them from their heavy burden for a ridiculously low price. The farmer is hot and tired, and may (to buy inputs and equipment), even be grateful to be cheated that way, because she gets rid of the time (especially women, who are weight and of the worry. If there were a warehouse at the market the main food crop producers) and where farmers could leave their surplus, it would help them not to technical skills. Another reason is panic and sell their crops for nothing. But you must also realize that that the marketing channels used a wrader often has a privileged relationship with farmers, because he by the poor are relatively has the capacity to give credit So a farmer may accept a low price by the poor ae relatively to keep a good relatonshi with the trader, to be able to ask him for impermeable to changes in export a loan in case of emergency." policies. Poor farmers tend to sell their crops at local markets where Source: Rapid Poverty Appraisal, local NGO workea in a northrn village, prices are more likely to reflect Novembe 1994. local, rather than national conditions. The few traders who come to the village to buy from producers are able to keep prices low regardless of broader supply and demand dynamics, because they operate in an oligopolistic environment (see Box 5.1).4 Hence, as long as small-scale producers are not organized, efforts to liberalize agricultural trade are likely to benefit primarily traders, who will make greater profits by saving the tax money, or richer producers, who may be able to transport their products to large markets where competition among middle-men will enable them to demand a better price. Lack of information and lack of organization among farmers are probably a greater obstacle to increasing rural incomes than trade restrictions. 5.8 Sectoral Policy. In 1986, the government unveiled a New Rural Development Strategy and this has remained the touchstone for all subsequent policy pronouncements. The objectives of this strategy were, broadly speaking, in line with poverty reduction concems: center development actions on the small rural producer, strengthen production support structures, introduce measures to increase and diversify output, improve resource allocation through better project selection, and protect the environment with active community participation. These themes are reiterated in more recent statements of Togo's policy toward the agricultural sector, such as the documents prepared for a 1990 donors' roundtable and the 1991-92 working group discussions among the Togolese, organized at the behest of the National Conference.s The 1995- 4 This, incidentally, explains why farmers often express a preference for an investment in a means of transport rather than, say, oxen and plough: to be able to market their crops independently would increase their income considerably, with little additional work. ' Planning Ministry, Developpement rural: Presentation generale du secteur, document prepared for the Second Donors Conference, Rome, 1990; Rural Development Ministry, "Lettre de politique agricole", Lome, March 1991; Togo: Overcoming the Crisis, Overcoming Poverty 58 97 Public Investment Plan underlines the Government's intention to continue its disengagement from direct intervention in the productive sector in favor of support to producers and rehabilitation of rural infrastructure. 5.9 Thus far, realization of the above policy objectives has been problematic. The main constraints are listed below. * Insufficient and inappropriate budgetary allocations. Funding for non-wage recurrent expenditures is grossly insufficient. Proper delivery of rural development services is generally estimated to require a minimum of 20-25 percent of recurrent expenditures for non- wage items, but since 1988 it has been about half that much, reaching an all time low of 3 percent in 1993 (see Table 5.1). While a greater number of well-qualified extension officers would be desirable, a better balance between personnel and non-personnel expenditures is the most crucial condition for rural development services to be effective. Table 5.3: Recurrent Expenditures for Rural Development (allocations in millions CFAF) Year Material* Wages Scholarship Subsidies Total 1988 245.9 (11) 2,071.4 0.0 28.7 2,346.0 1989 264.1 (HV.) 2,043.4 1.7 2.9 2,334.0 1990 262.9 (11%) 2,102.9 4.7 24.5 2,394.9 1991 247.1 (10%/) 2,182.6 2.7 23.8 2,456.2 1992 152.7 (7%/6) 1,972.1 2.0 21.2 2,147.9 1993 43.6 (3°%) 1,342.1 17.3 0.0 1,402.9 1994 277.2 (8%o) 3,022.7 17.3 0.0 3,317.2 1995 234.7 (8%) 2,572.0 13.0 15.0 2,834.7 *As a percentage of recurrent expenditures in parentheses. Source: Mininistry of Finance and staff estimates. * Weak extension services. Extension efforts to increase agricultural production and rural incomes have seldom given the intended outcome, with the exception of cotton (but price increases played an important role as well). A number of factors contribute to the poor performance of extension services: (a) lack of transport makes it difficult for agents to reach their clients6; (b) salaries are low because they are determined by the level of schooling7 and performance bonuses are generally given on political or personal grounds, so there is little Rural Development Ministry, "Synthese des differents rapports et comptes rendus issus des reunions preparatoires aux etats generaux du MDR", Lornm, March 1992. 6This helps explaining why the Rapid Poverty Appraisal in 1994 found that DRDR agents (i.e., extension agents from the Ministry) were largely absent from the lives of farmers interviewed; only SOTOCO agents (i.e., extension agents from the cotton para-statal) were a regular presence for cotton growers. In the same vein, a survey conducted in the Savanes (Naki-est zone) in 1994 found that 70 percent of farmers did not participate in extension meetings simply because there were no such meetings (see: Koffi Mondou, "Analyse socio-6conomique du systeme d'encadrement agricole excute par les services de l'etat en milieu rural au Togo", thesis no. 94/03AE, University of Benin, Lome, August 1994.) 7About CFAF 27,000 per month, equivalent to about US$54. Togo: Overcoming the Crisis, Overcoming Poverty 59 incentive to strive for good results in the field; (c) extension topics tend to be of little interest to farmners or provide advice which the farmers cannot realistically follow; (d) women's needs in particular tend to be ignored, in part because extension agents have little relevant advice to offer and in part because women are seldom included in contact groups;' (e) the approach followed privileges state intervention, and does not encourage farmer participation and self-reliance. * Poorly adapted agricultural research. Research does little to address farmers' needs because: (a) research centers and human resources are concentrated in the south (e.g., 59 out of 84 researchers), in particular a Lome; (b) researchers tend to lack adequate training (most of them are engineers); (c) access to scientific documentation and research results elsewhere is limited; (d) the operating budget is insufficient; and (e) linkages with extension agents and producers are weak. * Inefficient management practices. A major problem is the excessive centralization of decision-making, which on the one hand demotivates and disempowers regional staff, and on the other results in decisions taken without understanding regional conditions and therefore often inappropriate. To make things worse, different departments tend to operate in isolation, thus limiting the circulation of information important for decision-making and follow-up, and encouraging duplications and contradictions within the Ministry. Rigid hierarchical structures and formality within the departments are additional obstacles to staff initiative and the free flow of information.9 5.10 The main concerns of farmers, as reported in the 1992 agricultural sector review and in the national conference (etats generaux) on agriculture and land tenure held in February of the same year, reflect the constraints identified above. The farmers' point of view can be summarized as follows: (a) extension services in agriculture and animal husbandry were considered weak, (b) agricultural research did not address the real problems of the sector, (c) rural credit was insufficient, (d) terms of trade were very unfavorable because of low producer prices for food crops and stiff competition on the international market for export crops, (e) land tenure was often uncertain, (f) transport and storage infrastructure was insufficient, and (g) access to water was problematic. 5.11 Recent field work confirms that farmers' perceptions have not changed much since 1992. In particular, access to inputs is considered of paramount importance, especially access to fertilizers (see Box 5.2). Access to credit is also a major concern, in part because it is often perceived as the only realistic way to obtain fertilizer and other inputs, and in part because it makes it possible for poor farmers to purchase expensive items (e.g., a means of transport) or to weather external shocks (e.g., failed crops, illness). s Ablavi J. Atchikiti and E. Victorine Womitso, "Etude sur le role de la femme et les moyens de son integration dans les services de vulgarisation", mimeograph produced for the Agricultural Extension Project, Lomd, May 1994. 9For an evaluation of the structure and activities of MDRET see: FAO, "Preparation d'un plan de restructuration des institutions rurales" (preliminary version), Project TCP/TOG/445 1, December 1994. Togo: Overcoming the Crisis, Overcoming Poverty 60 Box 5.2: Access to Fertilizer During the Rapid Poverty Appraisal, more than two thirds of farmers mentioned better access to fertilizers as a measure to help them improve their welfare, with Savanes and Kara inhabitants placing particular emphasis on it. The priority given to fertilizer is to be understood in light of two factors. One is that soil fertility is declining and yields for staples show a Yield Trends for Staple Crops downward trend (see Table), Yield (kg/ha) with the exception of maize, which is often planted in 1972-74 1988-90 % change rotation with cotton and Cassava (Maritime) 24,164 6,656 -72.5 therefore benefits from the (Plateaux) 17,252 7,919 -54.1 fertilizer used on cotton (cereal Yam (Maritime) 9,756 10,771 10.4 yields went from close to 700 (Plateaux) 10,683 8,580 -19.7 kg per ha. in 1982-83 to about Sorghum & millet (Savanes) 676 556 -17.8 1,200 kg per ha. in 1993-94). (Centrale) 806 496 -38.7 became othereltiv thaccsmeds N.B. Crops are staples for the regions concerned. became relatively accustomed Source: DESA to using fertilizers for food crops during the 1970s and early 1980s, when fertilizers were heavily subsidized and more easily affordable to the average producer. Considering also that high-yielding seed varieties generally require the application of fertilizer, it is easy to understand why fertilizers are widely perceived as the key to increased productivity--and a better life. Unfortunately, at present fertilizer is basically available only to cotton growers, who can buy it on credit from SOTOCO at the beginning of the agricultural season (the credit is then repaid at harvest time, when SOTOCO deducts the fertilizer cost from payments for cotton). But food crops fertilizer, available from SOTOCO or through DRDR agents, can be bought on credit only by cotton growers and in any case, because of long-standing distribution problems, it is seldom available on time--or available at all. For example, in the Savanes survey mentioned earlier (Mondou, 1994), all of the farmers interviewed reported delayed arrival of fertilizers for food crops as a major problem. Starting with the 1994-95 season, a new distribution system has been put in place whereby only DRDR provides food crops fertilizer (in principle the fertilizer market is open to private traders, but government-subsidized prices have made it impossible for the private sector to develop in this area; also, the fact that farmers are not grouped together to place wholesale orders makes private fertilizer sales much less viable). According to the Directorate for Agricultural Statistics (DESA), this new system has caused a considerable drop in fertilizer, use because payment has to be upfront in cash, which is out of question for most farmers who could only reimburse after the harvest. The recent price increase from CFAF 65/kg to CFAF 100/kg is also believed to have contributed to lower fertilizer use (DESA,"Evaluation de la situation agricole et alimentaire du Togo. Campagne 1994/95", preliminary draft, Lome, November 1994). In addition, fertilizer has now to be bought at regional warehouses rather than at zonal warehouses, which means that farmers have to organize their own fertilizer transport over long distances--a feat almost impossible for the average food grower who lacks a means of transport or money to hire one. Given this situation, it should not be surprising that almost all cotton producers use some of their cotton fertilizer on other crops, and a few reported planting cotton just to be able to have access to fertilizer for their food crops. Policies and Instruments for Food Self-Sufficiency 5.12 The Togolese goverwnent has a long-standing pledge to achieve food self-sufficiency. In 1973 a state marketing agency (Togograin) was established to buy up and distribute crops in order to smooth inter-seasonal and inter-regional price variations and to guarantee a secure supply of staples. However, Togograin has never had the storage capacity or the purchasing Togo: Overcoming the Crisis, Overcoming Poverty 61 power to significantly influence food crop prices, which typically exhibit extreme seasonal variations (see Figure 5.1a & 5.1b and Annex 2). Such variations, which have remained practically unchanged for the past twenty years, have been mainly attributed to the cost of stocking and the size of losses during stocking, while differences between regions appear to be directly related to transport costs.'0 Moreover, Togograin has consistently operated at a loss, even after a reorganization and the abolishment of its monopoly in 1986." Figure 5.1a & b: Food Crops Seasonal Price Variation, 1989-94 Oil Seeds and Nuts Yam Mn- DD- B~~~~~~~~~~~~~~~~~~~~~~~~~~~ I , in Di P* In id a D AgkI ki Di Ni Sp A I A Q NYi 1 -X- a.. - - ... (ih I I 5.13 Despite the existence of Togograin, in 1977 there were steep rises in food prices which were attributed to lack of regulation of food crops exported to Gabon and Nigeria during the oil boom. In response, the government proclaimed a Green Revolution and set the goal of becoming self-sufficient in food production by 1985. To achieve self-sufficiency, farmers were granted tax exemption on agricultural machinery and productive materials. Other incentives included stepping up agricultural subsidies (e.g., for fertilizers) and expanding the agricultural credit system. Although these measures probably benefited better-off farmers, poor farmers are likely 10 World Bank, "Examen de l'lntervention Gouvernamentale dans la Commercialisation des Produits Vivriers au Togo", January 1984. For example, in 1990, it made a loss of CFAF 23 on every kg of grain stored and a total loss of CFAF 261 million. Togo: Overcoming the Crisis, Overcoming Poverty 62 have derived little benefit from them, as agricultural machinery remained beyond their reach with or without taxes, while agricultural inputs were made more affordable but still required cash outlays at a moment in the agricultural cycle when money is hard to come by. The expanded credit system opened greater opportunities to farmers as long as they could satisfy collateral requirements and had transport to reach banking facilities in larger towns; for the typical smallholder, however, credit remained largely inaccessible because of terrns of credit and distance problems. 5.14 Has the government achieved its goal of food self-sufftciency? The answer found in the majority of official documents is, broadly speaking, yes. According to official estimates, Togo has achieved food self-sufficiency since the late 1970s, except during drought years (such as 1983 and 1985) and with the exception of wheat, rice and meat/fish. It is difficult, however, to reach such conclusion with certainty, because available data suffer from a number of shortcomings and there are wide discrepancies in estimates provided by different agencies. More importantly, food self-sufficiency refers to the ability of domestic production to cover overall country requirements, but it does not necessarily mean that the whole population is food secure. Even in the presence of sufficient national production, sufficient food may not be regularly available or accessible (financially or physiologically) for some population groups--generally the poorest. Hence the importance of going beyond estimates at the national level. 5.15 On the basis of the annual per capita requirements, it is possible to calculate the food requirements of each of the five regions (see Table 5.4; calculation details are in Annex 3). Maritime, which is densely populated, shows a large deficit in cereals (maize, rice, sorghum, millet, and fonio), and a surplus production of cassava. Plateaux is a clear surplus region, with a deficit only for rice and legumes. Centrale has a surplus in yams and the northern cereals (sorghum, millet, and fonio). Kara has small surpluses only in roots, while showing a deficit for all cereals. Savanes shows a surplus in cereals and yams. As a whole, in 1993-94, Togo showed a large deficit of rice, and considerable surpluses of cassava and yams. Togo: Overcoming the Crisis, Overcoming Poverty 63 Table 5.4: Food Balance 1993-94 (in 1000 tons) Maritime Plateaux Centrale Kara Savanes Country Maize -55.8 40.0 -0.5 -14.3 34.9 4.4 Rice(paddy) -41.1 -14.1 -2.8 -4.1 4.8 -57.3 Other cereals -5.2 6.5 17.6 -7.0 14.6 26.6 (Sorghum, Millet, Fonio) Cassava 35.4 75.9 10.3 26.2 -1.3 146.6 Yams -4.6 30.6 78.3 43.8 24.5 172.6 Other Tubers 19.0 -3.0 -1.0 -1.4 -0.5 13.1 Legumes -2.5 -6.1 -1.8 -2.9 8.3 -4.9 Groundnuts 0.6 0.3 0.1 4.3 10.4 15.7 Source: staff calculations based on EBC 1987-89, DESA and Direction de la Statistique. 5.16 Annual variations are important. Thus, while the root crops production may vary in size but always exceeds requirements,'2 this is not the case for cereal production, which in most years shows a deficit (see Figure 5.2). Only in exceptionally good years the cereal production satisfies total requirements. Luckily, in most years cereal shortages are balanced by the surplus production of cassava and yams, which means that food security problems remain relatively moderate. Only during the 1988-89 agricultural year this was not the case: cereal shortages were very large in that year, and cassava production was low as well.'3 12 See Annex 2 which presents information on food crop production and requirements of yams, cassava, maize, legumes, rice, wheat, and other cereals. 13 See Annex 2. Togo: Overcoming the Crisis, Overcoming Poverty 64 Figure 5.2: Food Balance: Cereal Requirements and Cereal Production, 1983- 2000 700000 650000 600000 550000 500000 o / _- ~~~~~~~~~~~~~~~~~~~~~~~~~~~Production , 40000 350000 / / 300000 / / 250000 200000 l 1913 194 1985 1916 19S7 1911 1919 1990 1991 1992 1993 1994 1995 1996 1997 199S 1999 2000 Note: Cereals include wheat, rice, maize, sorghum, millet & fonio. Source: Staff Calculations based on EBC, DESA. 5.17 The Savanes, where cereals are contributing almost 70 per cent of all calories, is particularly vulnerable to harvest failures. In most years, the cereals harvested cover the nutritional requirements of the region. However, regular harvest failures during drought years result in food shortages, because this region does not have a sufficiently large buffer of alternative crops to provide a safety net against harvest failures in the main staple. In recent years this has happened about once every two to four years: 1983-84, 1987-88, 1988-89, and 1990-91 (see Figure 5.3). Togo: Overcoming the Crisis, Overcoming Poverty 65 Figure 5.3: Cereal Requirements and Cereal Production in the Savanes Region, 1983-2000 150000 140000 Needs 130000 Production c 120000 1 / e SI l/ *~11000 / o looooo-1 ,, 1000 ', i l /- 900000 80000 / _ 70000 , 60000 '-e ~--+-------±-- -i-i------- v± --------,------,--- I + I 1983 1984 1985 1986 19S7 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Note: Cereals include wheat, rice, maize, sorghum, millet & fonio. Source: Staff Calculations based on EBC, DESA. Are the Poor Food-Secure? 5.18 Regardless of whether the country as a whole, or each region separately, is able to produce enough food to feed everybody, the real test for food security is at the household and individual level. It matters little that enough food has circulated within the country in a given year if there are entire communities where food is either unavailable or unaffordable for months. Measures of global and average consumption, therefore, have to be supplemented with information on the distribution and seasonal variation of such consumption. Eating patterns during the pre-harvest "hungry" period (soudure), and the duration of the period itself, provide this type of information. In fact the pre-harvest season is the time of the year during which food scarcity is most likely to be experienced, as food stocks might be exhausted, money from the past harvest spent and the new harvest not yet ripe. 5.19 The length of the soudure is mostly determined by climatic and soil conditions. In the Savanes, where rainfall tends to be erratic and soils severely degraded, the soudure lasts about five months (April-August); in east Kara, where land is scarce and soils depleted, the soudure may last up to six months, while in center-west Kara, with fertile soils and good rainfall, it lasts only about two months. In the Plateaux and Centrale regions, periods of widespread food scarcity are rare in part because farmers can grow enough food for their needs and in part because cash crops make it possible to buy food in case home production is not enough. In the Maritime region, food shortages are less likely to be concentrated during a particular season, as the climate makes it possible to grow crops all year long, but there are pockets where food insecurity is pervasive and chronic due to pressure on the land, such as the Vogan district. Togo: Overcoming the Crisis, Overcoming Poverty 66 5.20 Food insecurity is reported by farmers and NGOs to be a Box 5.3: Almost al Granaries are Empty . major problem (if not the major problem) in the north, where the "Now people can still eat out of their granaries, but you wait until d4 April or May, and you'll see that almost all granarie are empty. Already in March school attendance starts going down, because you The high frequency of absolute can't expect children to walk long distances to come to school with poverty among farm households an empty stomach--they just don't have the strengt to -walk. (defined on the basis of access to Instead they go in the forest to look for baobab leaves, or somethig calories) and the high malnutrition else to eat. People will mix leaves with flour to make the flour last, but that stuff doesn't really fill you up. Still, it's better than nothing, rates found in the rural Savanes. and there are certanly families in this village who may eat once and Kara confirm farmers' claims every two days during the soudure. What happens in many families that entire communities may go for is that children are asked to cup their hands; the mother tfills the weeks without anybody getting hands, and that's all the food for the day--if they want more. theY enough to eat (see Box 5.3). have to fend for themselves. These children are not going to think People cope with food shortages in about learning to read and write, they're going to think: about two main ways. One is to add wild eaiiiig. food to their diet (baobab and Source: Rapid Poverty Appraisal, elementary school teacher in a village in manioc leaves, noix de palm iste), southei Kaa region, December 1994. the other is to reduce the amount of food eaten each meal or, more frequently, to reduce the number of meals. In the richer central regions, it is common for farmers to eat twice rather than three times a day during the pre-harvest season, or to keep to three meals but lower the quality and quantity of the food. In the northern regions, meals are often reduced to only one a day, and some times even less. 5.21 In conclusion, there appears to be little doubt that, regardless of official estimates about national food self-sufficiency, food security remains an elusive target for a sizable part of the population, especially in the north. Government interventions have thus far not been very successful in alleviating this problem. While profits from cash crops have indeed increased considerably over the past decade, little progress has been made to increase the productivity--and therefore the food security-- of subsistence farmers. Unfortunately, the rural population cannot count on an effective safety net, as there is no early warning system to act rapidly on localized food shortages and Togograin is far from being able to fulfill its intended role of granary of last resort. 14 For example, the Participatory Poverty Appraisal found that only 15 percent of the farmers interviewed in the Savanes and 10 percent in the Kara region had enough to eat all year long. In the Plateaux, almost half of the respondents reported eating enough throughout the year. While the sample is not representative and is likely to give an overly grim picture, findigsof widespread food insecurity have been reported by a number of NGOs as well. Togo: Overcoming the Crisis, Overcoming Poverty 67' 6. SOCIAL SERVICES AND SAFETY NETS 6.1 Broad-based economic growth is necessary to increase the income-earning opportunities of the poor, but some will be unable to take full advantage of them because of ill-health, malnutrition, illiteracy and lack of skills. Ensuring access for the poor to basic social services-- particularly basic health care, primary education and family planning--is therefore doubly essential, as it alleviates the immediate consequences of poverty and at the same time attacks some of its main causes. Greater investment in human capital, therefore, is both a means and an end in the development process. It is a means because it is a pre-requisite for increasing productivity. It is an end because those with a well-developed human capital base can ensure a better quality of life for themselves and their children. 6.2 But no matter how pro-poor policies are, there will be people who may not be able to benefit from them and may need special assistance. For some, such as those affected by economic recession or natural disasters, it may be a question of time; for others, such as the old and disabled, it may be for life. And even among those who do benefit from economic and social policies, some will remain acutely vulnerable to external shocks. Indeed, broadly speaking, all people falling in the category of the extremely poor are likely to be in need of special assistance, as by definition they are unable to secure adequate nutrition and access to basic social services. For all these individuals, policies leave a gap that must be filled by safety nets and targeted interventions. 6.3 This chapter focuses on the poor's access to social services and safety nets. It analyzes the effectiveness of government policies and interventions in providing such services to the population as a whole and to the poor in particular. As with the rest of the report, an effort is made to exarnine the situation from the point of view of the poor. Thus attention will be given to the ways in which vulnerable groups try to gain access to education and health services by organizing themselves or using traditional approaches. Similarly, traditional safety nets will be considered along with safety nets provided by the government. Health 6.4 Good health is important for all, but for the poor it is especially important because their income tends to depend exclusively on their physical labor, and because they lack the assets to help them cope with periods of ill health--when disease strikes, they have no savings to make up for lost income or to pay for needed treatment. The poor tend to have worse health conditions than the average. They are more exposed to disease-causing agents such as polluted water and inadequate sanitation, they often live in crowded conditions where infections spread easily, they have lower access to preventive health services and their poor nutritional status makes them more vulnerable to disease. Field work shows that disease, sometimes followed by premature death, is often the cause of extreme poverty, which explains why grass-root communities routinely mention poor health (including disability) as one of the characteristics of the poorest people. Investments to reduce health risks among the poor and to make basic health care accessible even Togo: Overcoming the Crisis, Overcoming Poverty 68' to those with limited financial means, therefore, are an important element in a poverty reduction strategy. 6.5 At independence, government policy focused on curative care and was primarily oriented towards urban dwellers, thus implicitly neglecting the needs of the majority of the population. However, persistent high mortality and morbidity, especially in rural areas, led to a shift in government policy. In 1978, in accordance with the Alma Ata declaration, the Government adopted a primary health care strategy that gave priority to peripheral services. The new policy called for expanded health infrastructure, increased staffing and more attention to broad-based preventive services such as immunization, health education, malaria control, environmental sanitation, birth spacing, and maternal and child health. In the subsequent years, health status and access to health improved considerably. Life expectancy rose from 42 to 51 years (1989), while infant mortality declined from 182 to 97 per 1,000. Between 1974 and 1986, the number of people per doctor went from 18,200 to 13,387 and the number per nurse from 4,200 to 2,454. 6.6 These encouraging results, however, hide considerable inequalities in the allocation of resources, and consequently in health indicators. Although by the late 1980s health care was more accessible to the average Togolese than it had been immediately after independence, the health situation left much to be desired. Primary health care, particularly in rural areas, continued to be neglected in favor of hospital care and the health status of the population compared unfavorably with other SSA countries with a similar GNP per capita. In 1990, a National Health and Social Sector Policy reiterated the government commitment to satisfy the medical needs of the population through a three-pronged strategy: (a) reduce mortality and morbidity due to endemic diseases by strengthening curative services, setting up a program to provide low-cost essential drugs, and improving the functioning of existing infrastructure; (b) focus on maternal and child health services and family planning; (c) make the health system more effective by improving planning, budget allocation and management, and by reorganizing the Ministry. Unfortunately, the economic and political crisis got in the way, and the sector's performance continues to be hampered by the following key constraints. * Inadequately articulated priorities and strategic framework. Because there is no mechanism for systematic and broad-based policy formulation, government policies and programs tend to be formulated on an ad hoc basis and are not backed by strong commitment. Insufficient monitoring and evaluation make strategic planning and effective implementation all the more difficult. Consequently, programs have rarely achieved their targets and have been carried out in an uncoordinated manner, leading to serious duplication and waste of resources.' * Insufficient funding. In the 1980s, inadequate budgetary resources led to under- fmnancing of non-salary operating costs as staff recruitment, even for non essential personnel, continued to grow. The 1992-93 financial crisis caused a firther reduction of 'The 1990 National Health and Social Sector Policy Statement intended to address this constraint, with the support of the IDA-financed sector adjustment operation, but the country's events of subsequent years prevented its irnplementation. Togo: Overcoming the Crisis, Overcoming Poverty 69' critical non-wage operating funds, with dramatic consequences for the sector: special programs such as maternal and child health care and the expanded program of immunizations came almost to a standstill; rural areas were left starving for essential drugs and medical supplies; health training was drastically reduced and some of the professional health schools had to be closed. In 1994 the health budget increased and allocations for non-salary expenditures went from less than 20 percent in 1991 (last year before the crisis) to over 35 percent in 1996 (see Table 6.1). Given the dire status of the sector, however, the recent increases are still insufficient to cover even the most basic needs. Table 6.1: Health Budget, 1990-96 (allocations in millions CFAF) 1990 1991 1992 1993 1994 1995 1996 Recurrent budget 6,685 6,126 6,355 4,338 8,205 8,087 11,297 Wages 4,134 3,633 3,862 2,596 4,677 4,549 5,417 Materials 1,255 1,192 1,192 390 1,982 2,032 4,099 Transfers 1,296 1,301 1,301 1,352 1,546 1,506 1,781 Investments* 600 600 1,728 942 3,017 1,273 1,843 Total health budget* 7,285 6,726 8,083 5,280 11,222 9,360 13,140 As % of general budget* 8.8% 8.2% 11.3% 7.4% 11.7% 9.3% 12.1% Excluding foreign investments. Source: Budget General (1995), Project de Budget (1996), and staff calculations (Ministry of Finance, Ministy of Planning, World Bank). 3 Poor collaboration between ministries. The problems caused by budgetary constraints have thus far been aggravated by a highly centralized public expenditures management system, resulting in annual allocations for the sector which do not reflect its priority needs. Initial steps to correct this situation have been taken during the public expenditures review process in early 1996. Also, little effort is made to join forces with other public institutions with complementary mandates, such as the Ministry of Education (for health and nutrition education) and the Water Department (for better water and sanitation services). * Structural Constraints at Service Delivery Level. The consequence of policy, management and financial weaknesses is a service delivery hindered by serious dysfunctions. No outreach program is in place to ensure that the most vulnerable groups, particularly women and children, benefit from available services. In addition, the referral system functions without operational norms or treatment protocols, thus it is neither reliable nor systematic. 6.7 The constraints mentioned above hamper the functioning of the public health sector as a whole, but have a particularly negative effect on the poor. To begin with, the poor have higher morbidity rates than the non-poor, so they will be disproportionately affected by shortcomings in health care delivery. This is even more so considering that the poor's only access to modem medicine tends to be through government services, as they cannot afford private care and in any case they generally live too far from any private provider. In addition, the consequences of inappropriate policy decisions, inadequate resources and inefficient delivery systems are Togo: Overcoming the Crisis, Overcoming Poverty 70' primarily felt at the periphery, where most of the poor live. As a rule of thumb, the farther a community is from Lome, the less likely to have access to good quality health care. Table 6.2: Health Facilities by Region, 1993 Hospitals Health Centers* Private Region Public Private e Public Private % Practices Maritime 9 2 31 125 6 23 102 Plateaux 7 3 28 171 2 30 2 Centrale 4 - 11 128 3 23 1 Kara 7 - 19 80 3 14 1 Savanes 3 1 11 47 9 10 - Total 30 6 100 551 23 100 106 Includes: Centres medico-sociales, dispensaries, health posts. Source: Ministry of Health, Population and National Solidarity. 6.8 Geographic inequality has indeed been identified by the government as a major problem2. Most human, financial and material resources for the sector, including infrastructure, are concentrated in the Maritime region, particularly in Lome (see Table 6.2). This region, which has 35 percent of the country's population, receives close to 90 percent of the country's drug supply and has about 70 percent of the country's health personnel.3 Health status mirrors these statistics with the infant mortality in Lome half of the national average. The northern regions are the worst off in terms of health services and, not surprisingly, in terms of health indicators as well. The Savanes, in particular, appears to be far behind, with half the number of medical personnel and equipment than most other regions--and twice the percentage of children under- weight for their age.4 Unfortunately, regional budget allocations appear to reinforce regional and urban/rural differences (see Table 6.3). Table 6.3: Health Budget by Region, 1995 Personnel Non-Personnel Total CFAF per capita (1,000 CFAF) (1,000 CFAF) (1,000 CFAF) Maritime* 1,184,897 1,650,004 2,834,901 1,777 Plateaux 694,816 155,742 850,558 883 Centrale 375,998 166,422 542,420 1,212 Kara 603,336 368,448 971,784 1,844 Savanes 217,844 74,885 408,474 784 * Includes allocations to the two university hospitals. Source: Budget General, Gestion 1995 and staff calculations. 2 Mawuli R. Adzodo and Kouami D. Houngues, "Analyse de la situation sanitaire au Togo au 30 avril 1995", paper prepared for the Ministry of Health, Population and National Solidarity, Lome. 3For example, in 1990 the Maritime region had 1,082 people per doctor while the rest of the country averaged 33,846 people per doctor. Statistics for nurses and midwifes are similar: a nurse for 397 people in the Maritime area vis-a-vis 3,388 in the rest of the country and a midwife for 916 people in the Maritime region vis-&-vis 16,633 in other areas. 4 Houngues and Adzodo, op. cit., p.84. Togo: Overcoming the Crisis, Overcoming Poverty 71 6.9 Another way to assess the effectiveness of health services in reaching the poor is to look at the demand side, that is, at the behaviors and opinions of the intended beneficiaries. Health problems are consistently mentioned by people living in poor communities as one of their main concerns, hence the frequent requests for health centers. It is therefore surprising that health services appear to be significantly under-utilized. It is estimated that only 23 percent of the population uses public health facilities and only for certain diseases,5 while bed occupancy in public hospitals is generally between 50 and 20 percent (with the exception of Lome, where primary health care facilities are not available and the two university hospitals operate more like giant dispensaries). Field research suggests two main reasons for the low utilization of public health services: cost and quality of service. 6.10 Primary health care is in principle relatively cheap, with initial consultations costing between 100 and 300 CFAF. But getting well has a much higher price: first there is transport, with both its direct cost and its opportunity cost (one might have to wait a whole day for a ride); then there is the all too common need to "motivate" health personnel (from small tips to outright extortion); and finally there is the cost of medicines--a problem that has been exacerbated by the devaluation. A large amount of prescribed drugs are either unnecessary or unnecessarily expensive. Some physicians over-prescribe to show their competence or to be on the safe side if they are not sure of the diagnosis. Others prescribe expensive brand names either because they are unfamiliar with generic drugs or because they are afraid that prescribing "simple" drugs like quinine could be interpreted as a lack of up-to-date professional knowledge. The pressure of pharmaceutical companies on pharmacists and physicians plays a role as well. The result is that, as an NGO representative put it, "people do not go to the health center because they are afraid of the cost of medicines" (see Box 6. 1). ,:-1 - I -:, . T#e, -o.--d.H t - . r,l J : < ' -. I, j3 ,> V I,. x USAID, "Analyse du secteur de la sante au Togo", 1990. Togo: Overcoming the Crisis, Overcoming Poverty 72' 6.11 Should prospective patients be prepared to pay, the expected low quality of the service offered acts as a disincentive. The most common complaint heard during field work is the unavailability of medical supplies, and anecdotes of people who died because of lack of even the most basic drugs were common. The Bamako Initiative has brought great improvements in the availability of generic drugs, but in many cases initial supplies have not been replenished timely and the situation is still far from ideal. Over 90 BOX6.2:WhenKoNgotsick percent of the drugs "Kofi began feeling sick to his stomach eight months ago, when he was five", imported in the country said his grandmother sitting next to him on a bed of the pediatric pavlion. '$His are expensive brand mother first took him to the gu*risseur who gave hirn a [herbl] treatmen to be names sold in urban areas, taken for a month. He was not to be paid until Kofi had recovered, and since where over 90 percent of that didn't happen and Kofi remained "vague" for a long time, the guise. was not paid. Kofi's father refused to even pay for thelherbs. We then'took the private pharmacies are Kofi to thefiticheur, but he wasn't able to make Kofi healthy either. Instead he located. The private made a hole in his belly during a ceremony that should have had an immediate. sector is supplied by effect. Kofi lost a lot of blood and was nearly dead when we decided to bring TOGOPHARMA, a him to the hospital. They operated on him right away. Because of this, my parastatal company which daughter and her husband got into a lot of debts. But they don't really mind, because they know that Kofi is getting well." until recently had the ::Ei: monopoly of drug Source: rapid Poverty Assessnent, patient's relative at hospital in maritime imports, and two private region, July 1995. wholesalers. The Government is the only supplier of drugs to the public sector, although in recent times an increasing number of public health facilities has started acquiring medical supplies through non- gouvernment sources (at a fraction of the TOGOPHARMA price). 6.12 So, what do poor people do when they get sick? First of all, they try to medicate themselves with traditional remedies or with counterfeit modern drugs sold (illegally) at the market. If that does not work, they go to the gu&risseur (traditional healer) who will use traditional methods usually based on herbs, asking about 500 CFAF for several days of treatment plus the cost of the herbs. In case of failure, a visit to the feticheur (witch doctor) is arranged at an additional cost of 700-1,000 CFAF or more. Recourse to a health center is the last resort, often when it is already too late and irreparable damage has occurred (see Box 6.2). 6.13 It is tempting to explain the preference for guerisseurs and feticheurs by the strength of cultural beliefs. But there are sound economic reasons as well. Traditional treatments not only tend to cost less than modern ones, but can be paid for in kind and in installments--no upfront cash is ever required, and no payment is due if the client is not satisfied (see Box 6.2). Should public health structures provide services more in line with the needs of poor customers, there is reason to believe that utilization rates would increase, thus making more efficient use of government funds. An interesting example of the direction in which public health services could evolve is the Catholic Hospital of Afagnan. This hospital is always filled to capacity and has a waiting list of up to six months, with patients coming from all over the country. Its popularity is due to a number of factors: health personnel is competent and client-oriented, cost is about one Togo: Overcoming the Crisis, Overcoming Poverty 73' fourth that of public hospitals, and deferred payment plans Box 6.3: Buoni kondo (including in-kind payment) "The situation is unbelievable" says Father Fiorenzo, surgeon at the are accepted (see Box 6.3). Afagnan Catholic hospital. "Although our hospital has the largest number Patient contributions cover of beds in the area, we cannot satisfy demand and every morning we have only a quarter of recurrent to turn away herds of people who remain just outside the hospital hoping expenditures, so running to be accepted the next morning". In the area around Afagnan hospital public hospitals in a similar "the herds" of Father Fiorenzo lodge in hotels made of straw and mud fashion would be a named buoni kondo, for 100 CFAF a night. It is estimated that during the year there are several hundred patients and their relatives living in buoni considerable burden on the kondo waiting to be accepted into the hospital. "To stay away from the government budget. But village for a long time has a very negative financial impact for a lot of underutilized public health patients", continues Father Fiorenzo. "Some of them begin spending the facilities have a high cost as money set aside for the operation while they are still waiting, running the well, and produce much fewer risk of not having enough money when their time comes. So we have well, and produce much fewer started a system whereby people can pay in advance and we give them a benefits. paper certifying that they have already paid for the operation--some sort of a surgery coupon. The problem is that in this way they may not have Education money for the buoni kondo and often they do not even have money to eat. When the day of the operation arrives, their physical condition is such 6.14 Evidence from cross- that the operation's success is jeopardized." country comparisons shows countran educmparisons pop ions Source: Rapid Poverty Appraisal, surgeon at Afagnan hospital, Maritime that an educated population s region, February 1995. critical to long-term growth. A country may grow for some time by mining its natural resources, but to sustain the growth process there is no substitute for educated people. The private rate of return to primary and secondary education in developing countries is at or above 18 percent, and just four years of education have been shown to increase small-farm productivity by 7 percent across 13 developing countries and by 10 percent in countries where new agricultural techniques were being introduced.6 Returns on expenditures for educating women are higher than for men, because of the positive association between education on one side, and health and nutrition on the other.7 Indeed, it has been convincingly argued that educating women is the best investment a country can make. 6.15 The government of Togo was well aware of the importance of education when it launched an education reform in 1975 with the ambitious goal of providing free and compulsory education to all children between 2 and 15. However, after a period of growth from 1975 to 1980, when the net primary school enrollment rate increased from 59.9 percent to 71 percent, enrollment rates started declining. In the mid-80s, a combination of rapidly rising school population and a 6 See amnong others: Barro, R., "Economic Growth in a Cross Section of Countries", Quarterly Journal of Economics, vol. 107, May 1991; and World Bank, "Primary Education: A World Bank Paper". Washington, DC 1990. 7See, for example, Summers, L., "Investing in all the People: Educating Women in Developing Countries", remarks prepared for a Development Economic Seminar at the World Bank, Washington DC, September 1992. Togo: Overcoming the Crisis, Overcoming Poverty 74' serious decline in Togo's financial situation plunged the education sector into a crisis that the sociopolitical circumstances of recent years have exacerbated, seriously compromising access to schooling and its quality (see Table 6.4). At present, the net primary enrollment rate is estimated to be 61 percent (1994-95 school year), student repetition rates have risen from 36 percent in 1990-91 to 46 percent in 1993-94, and drop-out rates are 7.6 percent (1990-91). Considering the whole school system from primary to tertiary level, fully 35 percent of students fail each year. While now there is no capacity for systematically sampling the level of student leaming outcomes, evaluations carried out in the mid- and late-80s already reported declining levels of performance. Between 1984 and 1990, on average it took 16.3 years of schooling to produce a primary school graduate. Table 6.4: Main Education Indicators for Togo Indicators* Males Females Gender ratio Prbnary Enrollnent rate in CPI 89.19% 66.00% 0.74 Gross enrolknent rate 112.00% 72.80% 0.65 Net enrollmnent rate 75.86% 53.10% 0.7 Completion rate 44.31% 31.90% 0.72 Repetition rate 33.05% 34.70% 1.05 Secondary Gross enrollnent rate 38.79% 12.80% 0.33 Net enrollment rate 20.74% 5.60% 0.27 Completion rate 15.13% 5.90% 0.39 Repetition rate 26.18% 26.70% 1.02 Tertiary Gross enrollnent rate 3.16% 0.60% 0.19 Adult literacy rate 56.85% 30.70% 0.54 *Because the school-age population is likely to be under-estimated, enrollment rates are likely to be over-estimated. Source: Direction Generale de la Planification de l'Education. Staristiques Scolaires, 1989-90, 6.16 One of the factors at the root of the country's educational crisis is the rapid population growth, which makes it extremely difficult to offer adequate schooling opportunities to increasingly large cohorts of students. The school-age population (children between 6 and 15 years) is expected to grow by 33 percent between 1991 and 200 1,8 a growth rate which could not be matched by budgetary increases even under the best economic circumstances. Inevitably, this will reinforce the present pattern of serious deterioration in the quality of basic education--which was, even before the mid-80s, characterized by low efficiency and lack of relevance to the needs of a developing country. Overcrowding is the most obvious manifestation of this problem, with students numbering as high as 200 in urban first grade classrooms. In such circumstances, it is not surprising that repetition and drop out rates are rising. aIt has grown from 1,003,270 in 1991 to 1,093,350 in 1994, and it should reach 1,335,350 in 2001. The same growth rate, even a little bit higher, is foreseen to take place in the primary school age population, estimated at 746,890 in 1994. Togo: Overcoming the Crisis, Overcoming Poverty 75 ' 6.17 In addition to the challenge represented by demographic trends, one has to consider the role played by policy decisions in shaping the performance of the national education system. Although allocations for the sector as a percentage of the national budget are high for African standards (29 percent against 22 percent as the average for Sub-Saharan Africa), non-wage allocations are far below what would be needed to ensure minimal instructional standards, especially in terms of pedagogical material and of infrastructure (see Table 6.5). Table 6.5: Education Budget, 1990-1996 (allocations in millions CFAF) 1990 1991 1992 1993 1994 1995 1996 Recurrent budget 21,816 22,879 25,351 17,238 27,721 28,796 31,454 Wages 15,176 15,659 17,475 12,565 19,706 20,386 20,919 Material 1,124 1,120 1,081 366 747 802 3,127 Transfers 5,516 6,100 6,795 4,307 7,268 7,608 7,408 Investrnent* 1,000 2,200 1,500 1,102 334 521 1,025 Total eduction budget' 22,816 25,079 26,851 18,340 28,055 29,317 32,479 As%of general budget' 27.6% 30.7% 37.4% 25.6% 29.3% 29.1% 29.9%/e *Excluding foreign investments. Source: Budget General (1995), Project de Budget (1996), and staff calculations (Ministry of Finance, Ministry of Planning, World Bank). 6.18 The problems created by insufficient resources are exacerbated by disparities within the sector. Eighty-three percent of the school population --and 100 percent of the poor-- are at the primary level, yet only 41 percent of the total sector budget goes to primary education. By contrast, the tertiary level has only 1 percent of the students but receives 25 percent of the budget, and the secondary level has 16 percent of the students but receives 34 percent of the budget. This means that in 1995 the government was planning to spend almost 43 times as much for a tertiary level student than for an elementary school student (722,785 CFAF against 16,859 CFAF; see Figure 6.1). In addition, because the bulk of the money allocated to primary education is for teacher salaries, the gap between levels for non-salary operating expenditures is even greater (479,239 CFAF per capita at the tertiary level against just 87 CFAF per capita at the primary level). Togo: Overcoming the Crisis, Overcoming Poverty 76' Figure 6.1: Per Capita Budget allocation by Educational Level, 1995 600,000 on 400,000. ~~~~~~~~~~~~~~~aar xpenditures 200,000- on alary expenditures 0- Tertiary Secondary Primary 6.19 Another _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ policy decision that ~~~Box 6.4: Teachers' Salaries has ad a impct o Ony 66 percent of primary and secondary.school teacher have regular civil serice: the schooling status; the others are considered auxiliaries, with. 13: percent hired on a contract opportunities of the basis through a variety of governument programs and 21. perent paid by local poor is the way in commnix ities. The salaries of auxiliary primary school teachers vary widely.1 which the recruitmnent Locally hired teachers average approximately CFAIF 7,000 monthly, 'but man1Y. are freeze of ~civil paid considerably less (see below) and may be ~paid in kinmd. Udrulfe teachers on governmnent contracts receive CFAF 20,000 monthly; :teachers with a servants which has four-year secondary certificate (brevet) receive CFAF 25,000 monthly and teahers been in place since with high school diplomas (baccalaur-iat) or above receive CFAF 30,0001mnthly 1984, has been This compares .to an average of CFAF 90,000 monthly for starting civil service applied. he hiring teachers, including benefits. freeze as resuted in I started teaching practically for frem I'm from this village, y ou see, and Iwa acute teacher lucky I could study. So I wanted the children of my village to go to' school, and I shortages, and the taught at the village school for three years for no money at all-parents gave me adoption of a number some food from time to time, when they could. During that time MY wife. worked of alternative as a domestic servant in town, for 2-3,000 CFA a month. She would leave the older arrangemets whichchildren at home, taki.ng with her only the one still on her back. As of last year,. I arrangments which was hired as a contract teacher, so now my wife can stay at home and mind the contribute to the fields. There are two other teachers, one regular and one temporary.Th deterioration of the temporary one is paid only 15,000 CFA [about US$30]. for the whole year, and he education system. has to live with his father." The most usual way to cop with eacher Source: Rapid Poverty Appraisal, elementary school teacher in a village in~ Bassa shortages is fordistrict (Kara), December 1994. communities to pay for auxiliary teachers, who are generally unqualified. In some cases, responsibility for paying the salaries of auxiliary teachers has been taken over by the Togo: Overcoming the Crisis, Overcoming Poverty 77' government, while additional teachers have been employed under various government programs.9 But regardless of the employer, the salaries of auxiliary teachers are far below the civil service pay-scale, which results in a highly fragmented teaching corps and low motivation among auxiliary teachers (see Box 6.4). 6.20 The outcome of this situation is that many elementary schools have to operate in the most primitive conditions. Teachers are lacking, classes are severely overcrowded, textbooks tend to be a rare luxury, school buildings are in serious disrepair and equipment (desks, blackboards, etc.) is sorely missing. Children from poor families are the most disadvantaged. Since parents must pay for school supplies and contribute to building maintenance and repair, the poorer communities have the most difficulty in maintaining these at an acceptable level--so the poorer the children, the worse the quality of their schooling. Not surprisingly, the poorest region in the country, Savanes, is also the one with the lowest enrollment rates (see Table 6.6). 6.21 The rural poor are particularly disadvantaged. In fact rural areas suffer disproportionately from teacher shortages because teachers in urban areas who retire or die are replaced by teachers from rural schools. Thus, the most remote communities, which are the poorest, often lack teachers and have no choice but to pay for them out of their own pockets, adding this expense to that for equipment and maintenance. As a result, the poor may end up having to pay for their education considerably more than others. For instance, in the rural Savanes region, where poverty incidence is the highest, 41 percent of all teachers are financed locally whereas in Lome only 17 percent of the teachers are financed locally. 6.22 Cost is one of the main reasons why Table 6.6: Primary Enrolment the poor do not send their children to school. R 1993-94 Keeping in mind that poor families have Enrollment Rate trouble meeting the most basic survival Gross Net expenses, and are also likely to have a large Martime 89.8 57.9 number of children (with a high dependency Plateaux 94.1 69.8 ratio), it is easy to see how some parents may Centrale 117.6 80.9 find it difficult to put together the required Kara 95.7 64.4 savanes s55.4 36.3 school fees, however low they may be, let Savanes 5.4 6. Togo 90.2 60.4 alone the money for uniforms and school Source: Ministy of Education supplies (as a teacher put it, "the school fee is only 350 CFAF, so it's not really a problem; the problem are the books: nobody can afford 9 One such program is the PEF (Programme d'Emploi et Formation), which started in 1991 with German and Canadian funding. The program was meant to provide young graduates with a one-year hands-on training, paying them half a regular salary (the salary was considered a training scholarship). However, all the 2,700 PEF participants--the majority of whom are teachers--have been retained to this day because they are considered necessary to cope with staff shortages. Since 1993 their salaries have been paid by the government Togo: Overcoming the Crisis, Overcoming Poverty 78' them")."0 In addition, there is an important opportunity cost, as most poor children are productively employed from age 6 or 7. This is particularly true for girls, whose domestic labor is often perceived by parents as much more valuable than the benefits to be gained from schooling. Indeed, lack of perceived benefits plays a very important role in explaining low enrollment rates in general, and the consistent lower enrollment of girls in particular (for the 1993-94 school year, there were 66 girls for every 100 boys in primary school; see Box 6.5). Especially in rural areas, for many families there is no perceived connection between real adult life and education. Education and teaching have little social prestige ("no husband will want an educated wife") or economic value ("school is useless: children spend time in school and then they're unemployed and haven't even learned to work the land"). Given the low quality of instruction and the limited relevance of the curricula, such attitudes are understandable. Box 6.5: Why Girls don't go to School In 1993-94, girls in primary school were 42 percent in the first grade, and 32 percent in the last grade. Overall, there wee nlyj 66 girls for every 100 boys. Intertwined material and social obstacles are typically responsible for the disooon between boys and girls' enrollment in primary education. Some of these factors are listed below. 0* jWThe perceived lower status of women and their roles--"Parents' think that girls ure goodfor nothing justgrow :t:iatand get married,f thy drop ouof school, it's because their pars encourage them less than b " * Lack oftpositive female role models, especially in rural areas--" There: aren't many girls lin school in rural areas,; beiause there's no example ofsuccessflwomen; In town, there are manyfemalecti ts".i serant '.' f Fear that educating girls might have a disrupting influence on their families and social enviromnent-"A g irL gosi to school misbehavs, does not serve her community and is less respecfid f traditlon:" (father). MWAOe ft husband who hais eth brideprce is afraidthat the girl will become intolerable because ofthe ngstuids. (teacher) * The loss resulting from forgone labor and income, which is much higher for girls than for s thefirt so de~ r girls: waK numerous, fbut as they grow up, they become increasingly useful to their mothers and theylea i scwhool to sellfuelwood' (teacher). i* Girls domestic obligations, and the ensuing lower educational achievement- "Try something: try having boys an l; girls do the same domestic chores. 7here will be the sameproblemsfor both sexes. If girls do not do at well fas boys in schoo4l it :s beca se ofallthe house work they must do" (mother). .*. Perceived low return on investment, as the benefits of a girl's education areseen asar not to her family who bears the cost, but to her futurehusband's family--"Schoolfor girls is less profitable: she leaves with her hsban4 andheis the one who will reap the benfits- (teacher). XC Soce.:. World Bank, Scolarisation et &olaritdf des Filles dans I'Enseignemer Primaire au Togo, Waington DC, Ioln a survey conducted in 1994, parents' estimates of total annual education expenditures for primary school (school fees, uniform, books,supplies and sometimes pocket money for meals) varied greatly, ranging from 3,000 to 30,000 CFAF. See: World Bank, "Scolarisation et scolarite des filles dans l'enseignement primaire", Washington DC 1995. Togo: Overcoming the Crisis, Overcoming Poverty 79' Safety Nets from the Government 6.23 Formal responsibility for assistance to the most vulnerable people falls mainly on the Social Affairs Directorate which is presently in the Ministry of Women in Development and Social Affairs. Although since its creation at independence it has changed ministry several times," its mandate has remained essentially the same: to carry out community development activities (Community Development Departrnent and Literacy Department ) and to provide assistance to under-privileged groups, such as the unemployed, the destitute, the elderly, the handicapped and abandoned children (Department for the Protection and Promotion of the Family). The intended philosophy is to avoid a welfare approach, encouraging instead productive activities and self-reliance. 6.24 The main instrument to achieve the Directorate's goals are Social Centers (135 in total)'2 located in the district capitals and in some of the largest villages (chefs lieu de canton), and staffed with APS (agents de promotion sociale or social development agents). In principle, each Social Center should have a team of APS comprising at least a man and a woman, who should act as a liaison between communities and the state, bringing the concerns of the population to the attention of government institutions and bringing the government's answer to the population. In reality, most Social Centers are under-staffed (four have no staff at all), have almost no operating funds,'3 and no transport for outreach activities. Needless to say, this situation makes it extremely challenging for APS to carry out their responsibilities. With motorcycles becoming increasingly rare since 1986,'4 APS can only reach villages close to the large communities where the Social Centers are located, which are often not the most needy (it will be recalled that isolation is one of the dimensions of vulnerability). No new hiring has taken place since 1983, so a booming population is being served by a decreasing corps of professionals (in 1992 there were 290 "regular" APS plus 92 "temporary" APS hired under the PEF prograrn). It is not difficult to see why morale arnong the APS is low. 6.25 With no operational budget, many Social Centers would be idle if it were not for the collaboration with donors and NGOs, who provide funds, transport and sometimes even salaries. Perhaps the most important (in termns of scope and duration) of such collaborations was for the implementation of the national nutrition program of the Catholic Relief Services (an American " At independence, the Social Affairs Directorate was under the Ministry of Labor and Social Affairs; in 1974 it was moved to the Ministry of Health and Social Security which in 1977 was renamed Ministry of Health, Social Affairs and Women's Affairs. In 1991 the Social Affairs Directorate was moved to the Ministry of Social Welfare and Solidarity which was renamed in 1992 the Ministry of Social Welfare, Solidarity and Human Rights. In May 1994 the Directorate was placed in the Ministry of Employment, Labor, Public Service and Social Affairs; since November 1995 it has been part of the Ministry of Women in Development and Social Affairs. 12 Social Centers are distributed as follows: Lome municipality: 11, Maritime:23, Plateaux: 27, Centrale: 25. Kara: 31 (of which 4 without staff), Savanes: 18. 13 The 1995 budget allocates a total of 9.5 million CFAF (about USS20,000) for "aid and subsidies" for the whole country. Assuming an equitable distribution among Social Centers, this means that each Center would have CFAF 70,370 (less than US$150) with which to assist the needy. 14 Interview with the Technical Advisor to the Minister, February 1995. Togo: Overcoming the Crisis, Overcoming Poverty 80' NGO), consisting mainly of monthly nutrition education sessions and distribution of food at minimal prices to mothers with small children.'5 When civil servants went on strike in 1992, however, Catholic Relief Services could no longer rely on Social Centers for training and distribution activities, and started using the services of another NGO (OCDI). The result is best illustrated by the statement of a farmer: "Social Centers have never done much, but at least before we could go there to get cheap food for our children; now they don't do that anymore, so they do nothing for us". 6.26 In an effort to complement the activities of the Social Affairs Directorate and, in particular, to raise much needed funds, in 1992 the government created the National Solidarity Agency, initially under the then Ministry of Social Welfare and now under the Ministry of Women in Development and Social Affairs. The objectives of the agency are quite ambitious: to sensitize the population and instill a spirit of solidarity as well as a self-reliant attitude, to assist vulnerable groups and of all needy people, to help train unemployed youth, to provide assistance in case of calamities, to reduce socio-economic inequality, and to collaborate with like-minded organizations."6 Unfortunately, the resources given to the Agency to carry out its mandate are meager (the 1995 and 1996 investment budget shows only a 30 million CFA allocation to assist the victims of civil disturbances)."' The Agency has very limited personnel in its headquarters in Lome, and only one social worker to deal directly with the intended beneficiaries. In principle there should be decentralized units, but thus far they have not been formed. 6.27 For the moment, it appears that the main activities of the Agency are to raise funds and to act as a referral facility. Fund raising activities involved a trip to Europe but are now concentrated locally (see Table 6A. I in Annex 2). Referrals are generally made to the Social Development Directorate, or to other ministries. In addition, the Agency has a small fund to provide direct assistance to special cases that other ministries could not help. The example provided for such cases was that of a student who did not have money to photocopy his thesis". 6.28 It would be difficult to conclude on the basis of the foregoing that the Togolese government is able to provide adequate safety nets for its less fortunate citizens. While the National Solidarity Agency had an important role to play in the immediate aftermath of the socio-political crisis, as indicated by the detailed assistance program it conceived for victims of the turmoil, its limited means and unfocussed mandate make one question the reason for its continued existence. 15 Catholic Relief Services closed its Togo operations in September 1995. 16Ministry of Health, Population and National Solidarity, Arrtet no. 12/92/MBES-SN Portant Statut, Organisation et Fonctionnement de I'Agence de Solidarite Nationale, 1 September 1992. Note that, soon after its creation, the National Solidarity Agency was placed in the Ministry of Health, Population and National Solidarity, where it remained until the November 1995 government reshaffle. 1 General Directorate of Planning and Development, Programmes des Investissements Publics Triennal, 1995- 1997, Lome. 1 Interview with Agency officials, 9 February 1995. Togo: Overcoming the Crisis, Overcoming Poverty 81' 6.29 As for the Social Affairs Directorate, its present activities are generally perceived by both the intended beneficiaries and NGOs as having little impact, especially when compared to the need. It is also clear that without substantive outside funding and a surge in staff motivation, the Directorate will not be able to carry out its responsibilities. Dependence on outside financing means that the Directorate's activities will tend to be donor-driven. This is not necessarily a bad situation, as donors are motivated by the best intentions and tend to have logistical and organizational capabilities superior to the public sector. It means, however, that the government is effectively abdicating its responsibilities toward its weakest citizens. In view of the financial straits in which the country finds itself and the presence of other actors at the grassroots (NGOs, agricultural extension officers), it may well be decided that the Government should concentrate its limited resources elsewhere. But keeping the mandate of the Directorate as it presently is, while denying it the resources to implement it, is likely to produce frustration and disillusionment. The mandate or the budget should change. The Role of NGOs 6.30 During colonial times, safety nets were provided by missionaries, and religious organizations in general. During the last two decades, however, non-religious NGOs started playing an increasingly important role as a result of the combination of several factors, including the sharp decrease in public resources (linked to economic deterioration), the preferences of donors, the growth of the NGO movement worldwide, and reduced employment opportunities in the civil service. With the 1992-93 crisis, the position of NGOs was much strengthened, as all humanitarian assistance went through them while public services remained paralyzed by the strike (see para. 6.25). This in turn, earned NGOs the confidence of people, making it easier to step into the void left by the government. 6.31 It is difficult to classify NGO activities. Some consist of service delivery (agricultural extension, education, health, access to water, etc.), others could be more narrowly likened to safety nets (e.g., giving food to destitute people during the pre-harvest season), others still may be cultural or religious in nature. NGOs differ considerably among themselves with respect to their goals, their capacity, and their means. A recent NGO census, for example, found that international NGOs spent on average CFAF 265 million in 1994, while national NGOs averaged CFAF 17 million.'9 The most common sector of intervention was rural development, often following an integrated approach at the community level. In this sense, it is possible to say that the majority of NGOs play a role in poverty reduction, although the geographic concentration of NGO activities suggests that poverty alleviation is not the only consideration guiding the choice of interventions. In fact the Savanes, which is the poorest region, resulted to be the region with the least NGO involvement.' 19 Ministry of Planning and Amenagement du Territoire and UNDP, "Les organisations non gouvernamentales au Togo", draft, June 1995. 20 One explanation could be that national NGOs (which are the majority) tend to concentrate in areas that are familiar to the founder; the Savanes, being a historically disadvantaged region and the farthest from the capital, is Togo: Overcoming the Crisis, Overcoming Poverty 82' 6.32 Regardless of the diversity among NGOs and of the doubtful nature of some of the them (the so-called "briefcase NGOs"), it is undeniable that many NGOs are making a significant effort to improve people's welfare. NGO expenditures in 1994 were upward of 3.9 billion CFAF2' (for comparison purposes, the government recurrent budget for rural development was 3.3 billion CFAF) and farmers who reported receiving outside assistance during the Rapid Poverty Appraisal mentioned NGOs (often missionaries) as the source more often than government (generally in the form of agricultural extension). Over half of the farmers interviewed, however, was not aware of any source of assistance. In some cases, the source had "dried up" because the missionary had left or the project had stopped (e.g., the FED North Togo project), but in most cases there was simply no agency or organization there to assist. Traditional Safety Nets 6.33 So, where do the Togolese turn for help? To their family, of course, and to their clan. In line with African tradition, extended families tend to be synonymous with safety nets. Which is why not having children is considered by many as an indication of poverty, while a large family represents an important social asset that limits vulnerability. This is perhaps most obvious in the case of orphans, who are generally taken in by relatives as a matter of course, and in the case of widows, who, according to the levirate tradition, tend Box 6.6: Traditional Safety Nets in North Togo to be "inherited" by their "The group is essentially the guarantee of survival for the individual in cse of oldest brother-in-law (this problems. SuTvival is only accorded to those who deserve it by behing is done in principle to give according to the rules. Lack of respect for the rules resuls in the them protection, although marginalization of the families concerned, who can no longer benefit fiom the in some cases greed may services usually provided by the community. The faults considered most well be the overwhelming severe may result in the use of drugs or poisons... Social status is essential. It depends on the family chief. And it has a worrisome aspect Resped is motivation). In this sense, acquired by showing the capacity to make the best of one's environment, be-it the family or clan plays the one's land or other activities carried out during the migration periods. This role of an insurance. capacity is measured by the expenses the family can afford when it is time for Needless to say, such ceremonies (funerals, marriages, fitiche celebrations) open to the whole insurance is all the more community. It is to maintain its social status that a family often goes into debt, together with their close relatives and friends who also feel obliged to valuable if members of the participate into the expenses. People will not hesitate to empty their granaes family are in influential at the beginning of the dry season to ensure their status, at the cost of having positions, or at least have to suffer from severe food insecurity until the next harvest. In any se, once reasonably regular access to the social status is assured, the group wili solve the problems that will face dii cash. Civil servants are -7 frequent benefactors to their Source: Pouzot, Denis. Contraintes a la Gestion de la Fertilt4 des SolW:Ciw. du Village de Naki-Est am Nord Togo, IFDC-A, Lom6, 1994, p. 49. less likely to be in this situation. Such explanation is all the more plausible if one considers that many NGOs are funded by unemployed graduates or by civil servants. 21 The NGO census mentioned above reported 3,867 million CFAF disbursed in 1994 and 6,368 million CFAF budgeted for 1995, but only 68 out of the 181 NGOs included in the census provided budget information. Hence, total NGO expenditures in Togo are certain to be higher. Togo: Overcoming the Crisis, Overcoming Poverty 83' the village through cash contributions for education and health expenditures or to buy food during the hungry season. 6.34 The effectiveness of traditional safety nets rests on a socialization process which stresses the importance of the family/clan and the idea of homogeneity (see Box 6.6). Modernization, therefore, poses a significant threat. As Western values make their way into the country, young people are tempted to embrace an individualistic and materialistic philosophy of life which is the opposite of the principles from which the extended family derives its strength. The emerging phenomenon of old people abandoned by their children is perhaps the saddest manifestation of this trend. Because urban areas are more exposed to new ideas and outside influence, it is to be expected that extended families will progressively lose their importance, both as a safety net and as a reference point, for urban dwellers. 6.35 While there is no doubt that the extended family is a powerful safety net, its effectiveness should not be overestimated. Its most troublesome limitation is obvious: its effectiveness is related to the means available to its members (and to their generosity, of course). Because people tend to marry within their own social strata, very poor people are likely to have relatives who are also very poor, and therefore will be unlikely to be in a position to provide much assistance. This is true at the community level as well, since many vulnerability factors are shared by the whole community (see Chapter 3): it is not uncommon that in a remote village nobody has enough to eat during the pre-harvest season, so nobody can offer a free meal to the neighbor's children. 6.36 The assistance provided by the extended family is often complemented by community- based associations. The most common are the tontines, which may function either as systems of labor sharing or as savings clubs. An example of the former are the tontines des champs ("field associations") whereby a group of farmers, generally of the same sex, will organize themselves to work together on the field of each member in turn. In the absence of in-depth studies, it is difficult to quantify the contribution of these arrangements to indigenous social security systems22. It appears, however, that communal labor systems play an important role, because they enable poorer families to share not only labor and possibly agricultural tools, but also food (often "superior" food, such as fish sauce) with richer farmers. The fact that exclusion from the tontine des champs may be used as a punishment for having broken clan rules, shows the value attached to such arrangements by farmers.' Besides playing a redistributive role, labor-sharing clubs are also a form of insurance against illness or other forms of incapacitation, because they guarantee continued agricultural production--and therefore survival--for their members. In this sense, their existence is particularly valuable for older people, who might not otherwise be able to cultivate their land. 22 For a quantification of labor transfers and a discussion of their implication in an African context, see von Braun, J., "Social Security in Sub-Saharan Africa: Reflections on Policy Challenges" in E. Ahmad et al., Social Security in Developing Countries. London: Oxford University Press, 1991. 23 In addition to the loss of a safety net, exclusion entails a direct cost, as the "culprit" will have no choice but to hire salaried labor or be late with field work (thus foregoing part of the benefits deriving from a good harvest). Togo: Overcoming the Crisis, Owercoming Povert 84' 6.37 Tontines as savings clubs were discussed in Chapter 4. What is important to add here is that their safety net function is somewhat secondary, and more often by default than by design. Basically, it is limited to situations in which a member is in desperate need of cash and the others decide to rescue her out of a sense of group solidarity. Just as it is the case with other forms of traditional safety nets that depend on group solidarity, the effectiveness of tontines in providing social security depends on the resources of their members. Persons with reasonably reliable sources of income are able to join more than one tontine and may count on several potential emergency funds. The poor, however, may not be able to make regular contributions, and therefore to participate. Thus, tontines can be an effective safety net only for those who can afford them, that is, who manage to have something, however little, to spare. The poorer the person, the smaller her contribution--and her safety net. The extremely poor are excluded. Togo: Overcoming the Crisis, Overcoming Poverty 8S 7. MAIN ELEMENTS OF A STRATEGY: PRE- REQUISITES 7.1 The main objective of the Poverty Assessment is to formulate a strategy for poverty reduction with specific and easily monitorable measures. However, these measures can only be implemented and produce the desired benefits if a number of more general pre-conditions are present. These pre-conditions have to do with the wider context in which the development process takes place. They encompass both process and outputs, and touch also upon the institutional and legal frameworks which shape the development process and determine its impact. 7.2 This chapter starts by summarizing the National Program to Fight Poverty. It then highlights the main elements of a proposed poverty reduction strategy. This strategy should be seen not an alternative to the National Program to Fight Poverty, but as a complement to it. It is hoped that identifying a limited set of priorities, as it has been done, will result in decisions that will optimize the allocation of the limited resources presently available in Togo and make implementation more likely. While the priorities are discussed in greater detail in the next chapter, three basic pre-requisites are discussed here: macro-economic growth policies, improved access to and quality of social services, and better governance. The National Program to Fight Poverty 7.3 In 1995, the Togolese government, assisted by LNDP and other UN agencies, started a process aimed at defining a National Program to Fight Poverty (Programme National de Lutte contre la Pauvrete or PNLP). This process requested the input of well over 100 experts from the government, the donor community and the private sector (consultants, academicians), who worked in thematic groups; drafts were also discussed at various meetings with government representatives and with the donor community. The result is a comprehensive program which, after presenting a diagnosis of the causes and manifestations of poverty, envisages three sub-programs and identifies a number of specific projects (see Figure 7.1). These projects, the majority of which are already in the Public Investment Plan, are estimated to cost US$572 million for the period 1996-2000, with US$265 million already pledged and US$300 million to be identified. A summary of the PNLP is provided below. (1) The sub-program for employment creation and revenue distribution to benefit the rural and urban poor. This sub-program presents two strategies, one for the rural poor and one for the urban poor. The strategy for the rural poor hinges on a three- pronged approach. (a) Productivity and production in the agricultural sector must be increased, and specific recomnmendations are provided for food crop production, cocoa/coffee, Togo: Overcoming the Crisis, Overcoming Poverty 86 cotton, animal husbandry, and fishing. In particular, it is recommended that food crop production be intensified to a level compatible with farmers' capacity for savings and credit, and that cultivated areas in the national land trust (domaine foncier national, see para 4.9) be expanded to encourage "internal colonization".' (b) Natural resources must be protected, especially to prevent further soil erosion. It is recommended that a national natural resources management policy be defined, accompanied by a comprehensive forestry program (especially for Lome, Aneho, Kara and Dapaong) and a strengthening of the institutional and legal context. In addition, communities should be given a more active role to play, e.g., in the management of national parks and in pilot activities for a global approach to natural resources management (gestion des terroirs villageois). (c) Rural communities must be better organized and assisted by promoting farmers associations and the creation of autonomous chambers of agriculture, by setting up a rural credit system (based on savings for food crops and using guarantee funds for cash crops), and by addressing land tenure problems. The solution to land tenure problems would be facilitated by the existence of farmers associations and the elaboration of regulatory frameworks based on traditional systems. 7.4 The strategy for the urban poor consists mainly in increasing employment opportunities for unskilled or semi-skilled individuals through labor intensive public works, while encouraging the creation of small and micro-enterprises through training and literacy programs. (2) The sub-program to improve social services for the poor. This sub-program has three components: (a) health and nutrition; (b) housing, water and sanitation; and (c) basic education, literacy and vocational education. (a) For the first component (health and nutrition), the accent is put on primary health care and the Bamako Initiative; malnutrition, and iodine deficiency in particular, are flagged as problems that need attention. (b) The strategy to improve access to water recommends that additional water points be created in rural areas, private connections be subsidized, and water at standpipes be paid. Greater reliance on the private sectors should be promoted trough the use of local enterprises (in urban areas) and networks of "repairing artisans" (in rural areas), while increasing community participation through Water Committees, community management of standpipes, and greater women involvement. The strategy for better sanitation concentrates on Lome and other urban centers, calling for the promotion of latrines and septic tanks, and for better The intensification of food crop production, in turn, requires better tcchnical assistancc from the govemment to farmers, a strengthening of functional literacy programs, support to help farmers organize, the availability of micro-credit, and efforts to diversify (e.g., counter-seasonal crops). Togo: Overcoming the Crisis, Overcoming Poverty 87 solid waste management. Also with an emphasis on urban poverty is the strategy for housing, which calls for regulatory changes to establish land registers in all urban centers and make both land purchasing and house construction simpler and cheaper. (c) The strategy for basic education starts by recommending actions to improve pre- school education. Primary education is seen not just as a bridge toward further education but also as preparation for employment. The accent is on improved quality. Proposed measures include: teacher training, better and cheaper pedagogical material, lighter administrative procedures, decentralization of routine administrative decisions, and better integration of schools in their environment. A Loi d'Orientation should sanction these measures and establish clear guidelines for the sector (e.g., teacher/students ratios). Functional literacy programs should aim at cutting in half illiteracy rates, particularly for women, while teaching practical subjects (e.g., agricultural techniques). The strategy envisaged would expand programs of proven effectiveness, open new literacy centers, and institutionalize literacy programs in enterprises. Vocational training (including apprenticeships) should be made more relevant by strengthening the information base on employment opportunities, training the trainers, and ensuring professional standards. Artisans and apprentices should be assisted. (3) The sub-program "framework and accompanying measures to stabilize the fight against poverty". This sub-program focuses on three issues: (a) population dynamics, (b) women's status and (c) social mobilization. (a) The population strategy focuses on three main objectives: limit growth, reduce morbidity and mortality rates, and redistribute population more equitably. To limit population growth, extensive information, education and communication campaigns should be accompanied by better access to contraceptives (e.g., through the integration of family planning services in matemal and child health care services), increased girls' schooling, support to concemed institutions, and the abrogation of the 1920 law prohibiting contraceptive marketing. To reduce morbidity and mortality rates, a combination of preventive measures (vaccination, sanitation, end of female genital mutilation) and primary health care practices (e.g., oral rehydratation) are recommended. To encourage a more balanced geographical distribution of the population, a national regional planning [amenagement du territoir] policy should be elaborated so as to improve living conditions in rural areas and promote the development of secondary urban centers. (b) The concem for women's disadvantaged status is addressed by a number of specific measures throughout the PNLP. Here the accent is on the need to offer women equal opportunities (e.g., better access to credit, girls' education, support to women's groups) and to reduce their burden (e.g., appropriate technology, child care centers). Togo: Overcoming the Crisis, Overcominz Poverty 88 (c) Organization and empowerment at the grassroots level is seen as crucial for the implementation of PNLP. Efforts in community development, including organization for access to credit, are therefore recommended. Figure 7.1: The National Program to Fight Poverty (PNLP) Production & productivity Rural Protection of natural resources 1. Employment & revenues Organization Urban Primary health care Malnutrition Access to Water 2. Social Srices Sanitation Housing Basic Education Literacy Vocational training Population 3. Accompanying measures _ Women status Social mobilization Source: Based on UNDP and MPAT, "Avant Projet de programne national de lutte contre la pauvrete (PNLP)", mimeograph, Lome, December 1995. 7.5 The comprehensiveness of the PNLP is at the same time its major quality and its main shortcoming. Because of the multi-dimensionality of poverty, interventions need to be envisaged in all sectors. But Togo's present financial straights and its limited implementation capacity are unlikely to make it possible to carry out effectively large numbers of interventions at the same time in all sectors. Hence the need to identify a restricted number of priorities on which to concentrate efforts. In addition, the focus on a comprehensive national strategy tends to overshadow the regional dimension. Given marked regional differences in problems and opportunities, such dimension needs to be developed. 7.6 On the basis on these concerns, the strategy proposed in the following pages builds on the PNLP and complements it. Recommended interventions are limited not because some sectors or projects are deemed unimportant for poverty reduction, but because of the need to prioritize and target interventions. On the other hand, Togo: Overcoming the Crisis, Overcoming Poverty 89 interventions are conceived in a broad manner: they can be new projects (that is, investments for which financing is to be sought, as in the PNLP), but they may also entail less costly measures, such as policy or regulatory changes (the PNLP identifies several) and re-allocations of existing resources. The point being that a lot can be done spending better rather than more. Outline of a Strategy and its Context 7.7 The Context: Pre-Requisites. In the case of low-income countries such as Togo, it may be difficult to distinguish between a development strategy and a poverty reduction strategy. Because poverty and vulnerability are widespread, a development strategy must coincide to a large extent with a poverty reduction strategy and significant poverty reduction can only be envisaged in the presence of economic growth. Hence policies aimed at ensuring macro-economic stability and consolidating the economic recovery which started in 1994 should be seen as the necessary context for the success and sustainability of interventions to attack the direct causes of poverty. Economic growth, in turn, has to be accompanied by improved social services for the poor, both in terms of accessibility and of quality/relevance. In fact, greater investment in human capital is both a means and an end in the development process. It is a means because it is a condition for increasing productivity; it is an end because those with a well-developed human capital base can ensure a better quality of life for themselves and their children. Improved governance, which has been shown to be closely linked to long-term economic growth, should be an overarching preoccupation. Three aspects of governance should receive particular attention: bureaucratic accountability, economic trasparency, and efficiency, and a well functioning judicial system. Box 7.1: Defining a Strategy: the Process is part of the Output The process used to define a strategy was based on the consideration that poverty is primarily a micro-level phenomenon. It is possible to say that a country or a region is poorer than another, but in the end poverty is experienced by each household and person in a unique way (see Chapter 3). When designting interventions to reduce poverty, it is therefore preferable to adopt an approach whereby the point of departure are the poor themselves. Putting individuals or households at the center, and trying to see poverty tbrough their eyes, makes it easier to appreciate the multi-dimensionality of poverty, thus facilitating an understanding of how different factors (and sectors) interact with each other. In additio, a bottom-up approach has practical advantages. Unless poverty alleviation interventions address the causes of poverty as they are perceived by the target population, there will be litte sense of ownership and participation within the community, making interventions more costly and sustainability doutl In line with a participatory, bottom-up approach, a rapid appraisal was first conducted in all regions to learn from people in a sample of poor communities what were their main problems and which solutions they envisaged (see Annex 2). Then, in a series of regional workshops, representatives of NCOs and regional govermment sevices were asked to explore the manifestations and causes of poverty in their region of operation, and to identify poverty reduction strategies at the household, community and regional level. This was complemented by discussions with donor and centrl government representatives, as well as academicians and experts (especially those who worked on the PNLP). The strategy outlined here reflects these inputs. Togo: Overcoming the Crisis, Overcoming Poverty 90 7.8 The Strategy. A poverty reduction strategy should hinge on four main priorities: * land tenure reform, * demographic growth management, * decentralization, • regional action plans. 7.9 Land tenure problems have to be addressed because land is a fundamental asset for the 80 percent of the poor who live in rural areas and because failing to address them will have disastrous long-term consequences for the environment. Unchecked demographic growth will not only magnify environmental problems, but make it impossible for the government to provide acceptable levels of social services while unemployment and underemployment will continue to grow. Decentralization is necessary to improve the effectiveness, speed and targeting of interventions: the best way to help the poor is to move the center of action closer to them. Region-specific plans are the complement to decentralization: because the causes and manifestations of poverty are context-specific and vary greatly across agro-ecological zones, each region should identify priorities in the fight against poverty and elaborate an action plan based on the region's particular problems and opportunities. 7.10 Central government efforts should concentrate on the first three priorities (land tenure, demographic growth and decentralization), as there is no conceivable way in which progress could be made without strong and committed government leadership and coordination. While donors and NGOs could provide financial and technical assistance, political will is the most crucial ingredient. Concerning regional action plans, the role of the central government should be to provide general policy guidance, ensure that efficient and consistent implementation processes are in place, coordinate and monitor activities in all regions, and facilitate exchange of information both between regions and with other countries (much can be learned from poverty alleviation experiences elsewhere in SSA). The plans themselves, however, should be defined with the input of all development actors at the regional level, including government representatives (services techniques), NGOs, traditional authorities and professional associations.2 The creation of five Regional Development Funds is presented as a possible option to support this approach and at the same time allow for some targeting (see next chapter). 2 Th regional workshops held in preparation of this document represent a first stop in this sense. Togo: Overcoming the Crisis, Overcoming Poverty 91 Figure 7.2: Main Elements of a Strategy Pre-requisites macro-economic growth improved social services better governance bureaucratic accountability economic transparency and efficiency \ improved functioning of the judicial Priorities land tenure reform demographic growth management decentralization regional action plans I Regional Development Funds Pre-requisite: Macro-Economic Growth Policies 7.11 The Government's development strategy is to consolidate the economic recovery of 1994 (see Chapter 1) by putting the country back onto a sustained poverty-reducing growth path, and stabilizing the institutional and political framework. The target is a 6 percent annual growth in GDP for the 1995-97 period. Most of the anticipated increase in growth is initially expected to come from agriculture, service activities, and small-scale manufacturing. In each of these sectors there is great potential for increasing the contribution of the poor, and therefore their revenues, but this potential can only be realized if service delivery is improved (see next section). Agricultural growth should come through productivity increases, in both traditional food and cash crops (primarily cotton) and non-traditional products (horticulture). The main source of growth for tertiary activities should be transit trade, provided timely rehabilitation and restructuring of the physical infrastructure in this sector. There is also scope for further increasing micro, small and medium scale manufacturing activities, especially for agro-processing, wood, and textile products, because of their increased competitiveness following the recent CFA devaluation. Growth in financial services is expected to follow over the medium term, since restoring Togo's position as a regional financial center will require comprehensive measures to rebuild banks' assets and confidence in the financial system. 7.12 Because Togo is a small country, the external environment will also be critical in determining the country's economic prospects. Four external factors play a crucial role: Togo: Overcoming the Crisis, Overcoming Poverty. 92 (a) heavy concentration of exports in a few primary commodities (primarily phosphate rock and cotton) which, given weak price prospects, will require strong measures to expand their exports, such as liberalizing the producer price and marketing systems for cotton, and increasing efficiency in the phosphate industry; (b) external debt obligations which, although in principle manageable, are high (debt is equivalent to about 34 percent of exports earnings);3 (c) reliance on concessionalfinancing to support the economic recovery, particularly with respect to the rehabilitation of the country's physical infrastructure (total foreign financing over 1995-98 is projected at about US$743 million, about US$36 millions short of projected external financing requirements); (d) increased regional integration, and in particular, the push towards the West African Economic and Monetary Union, creates new challenges and opportunities for Togo, which could exploit the increased potential for intra-regional trade, economies of scale and capital flows. 7.13 At a macro-economic level, the Togolese government is implementing a comprehensive program to restore macroeconomic stability and encourage growth by supporting a strong supply response from the private sector to the CFAF devaluation. Measures to restore macroeconomic stability include: (a) efforts to improve public finance by rebuilding the tax base, containing public expenditures and increasing their efficiency, and clearing internal and external arrears; (b) the restructuring of the public enterprise sector by speeding up the liquidation and privatization program to eliminate loss-making enterprises and shield viable enterprises from pressures to finance government deficits. Measures to support private sector development include: (a) improving the regulatory environment, especially through price liberalization and revision of investment and labor codes; (b) increasing the scope of private activity by encouraging privatizion and abolishing the remaining public sector monopolies in commercial functions. From a sectoral perspective, several constraints have been identified, some of which are already the object of government attention (see Chapter 4, 5 and 6). 7.14 Strong and sustained government commitment to continue implementing the stabilization and adjustment program and sectoral reforms will be critical for the resumption of economic growth--a necessary pre-condition for poverty reduction. The reorientation of the country's scarce public resources from non-development expenditures (mainly the military) to essential social services and productive activity will also be critical. 3 Togo's external debt strategy is to: (i) limit public external borrowing on non-concessional terms to short-term trade related credits; (ii) reschedule bilateral debt on concessional terms in 1995 to substantially reduce the debt service burden in present value terms; and (iii) with the assistance of IDA, buy back its commercial bank debt in 1966. In February 1995, Togo signed an agreement with the Paris Club creditors, which alongside a commercial debt buyback, will reduce the country's stock of debt by about one third. Togo: Overcoming the Crisis, Overcoming Poverty 93 Pre-requisite: Improved Social Services 7.15 Broad-based economic growth is necessary to increase the income-eaming opportunities of the poor, but some will be unable to take full advantage of them because of ill-health, malnutrition, illiteracy and lack of skills. Ensuring access for the poor to basic social services of reasonable quality is therefore doubly essential, as it alleviates the immediate consequences of poverty and at the same time attacks some of its main causes. Although the preferential treatment accorded to social expenditures from the late 1980s (see Chapter 1) is an indication of the government's commitment to providing social services to all its citizens, it is also true that social indicators are far from satisfactory and have worsened during the crisis. The sectoral analyses presented in Chapter 6 suggest two main problems: (a) the combination of economic deterioration and high population growth has made it impossible for expenditure levels to keep up with demand, and (b) whichever funds are available, they are not spent in the most efficient way. 7.16 Financing requirements. A public expenditures review for the education and health sectors revealed important financing gaps. For education it was estimated that just to ensure minimum instructional standards to the existing school population while maintaining current efficiency levels, would require extra financing of approximately 13.2 billion CFAF for the period 1996-98. For the health sector, extra financing requirements over the next three years are estimated at 3.1 billion CFAF. Although cost recovery in the form of user charges (for health) and school fees (including mandatory contributions for school maintenance) are widely practiced, it would be unrealistic to expect them to fill even a small part of the financing gap: cost recovery revenues should be used to supplement, not substitute for allocations from the central budget. In the past, donor money has played a major role in financing social sectors expenditures, but the dramatic decrease in foreign assistance registered in the 1990s (see para 1.17) suggests that, in the absence of major governance improvements (a condition given by some donors for resumed aid flows), Togo will have to count primarily on its own resources. In this context, the re-orientation of scarce public resources away from non-development expenditures (mainly the military) suggested earlier becomes crucial. 7.17 Efficient use of available funding. Because substantial budget increases cannot be expected over the medium term, the only way to make social services more accessible and more useful for the poor is to use better whatever resources are available. A number of measures to increase the efficiency of public expenditures and to target social sector spending towards the poor are highlighted in this document (especially in Chapter 6 and in this chapter), and are described in detail from a sectoral perspective in the National Program to Fight Poverty. In general, it is a matter of streamlined procedures and transparent management on the one hand, and of addressing imbalances on the other-- between tertiary and primary services, between preventive and reactive approaches, between regions, between urban and rural areas, and between salaries and operating costs. Sectoral ministries have already started tackling some of these issues. It is now a question of staying the course. Togo: Overcoming the Crisis, Overcoming Poverty 94 Pre-requisite: Better Governance 7.18 Governance is the exercise of political power to manage a nation's affairs. Good governance, therefore, requires the satisfaction of a wide range of conditions including political accountability, freedom of organization and association, an objective and efficient judiciary, bureaucratic accountability, freedom of information and expression, and efficiency within public institutions.4 These conditions, in turn, have been shown to be closely linked to long-tern, sustainable economic growth.5 A number of donors have raised concerns about the pace of Togo's progress on governance issues, making improved governance a condition for resumed--or increased--aid. 7.19 In the past few years, however, improvements have taken place: a National Human Rights Commission was created in 1987, multi-party elections were held in 1994 (the first ever in Togo), a number of independent periodic publications have seen the light, and in November 1995 a Ministry of Human Rights and Rehabilitation was created. In particular, improvements regarding human rights violations have resulted in Togo being taken out of the list of countries under special scrutiny by the United Nations Human Rights Commission (23 April 1996). While progress toward better governance, and therefore faster growth, could take place in many ways, three areas are singled out below for their particular importance. 7.20 Bureaucratic Accountability. One of the greatest obstacles to better governance is represented by the ideology that prevailed during the 1970s and 1980s, and which still influences public life. On the one hand, the effects of the colonial legacy and the heavy role of the state in social and economic life have encouraged private citizens to expect the government to provide an answer to every problem. On the other hand, some civil servants see their jobs primarily as a way to secure a salary, rather than as a public service to which their fellow citizens have a right. In addition, because criteria for promotion are mainly based on seniority, there is often little incentive to make an effort to carry out responsibilities effectively and efficiently, let alone to provide particular assistance to the least priviledged members of society. 4This characterization of good govemance is derived from Landell-Mills, P., and L. Serageldin, "Goveinance and the Extemal Factor", World Bank, Washington DC, April 1991, p.-3. 5The importance of good govemance for economic growth is convincingly argued by Surjit Bhalla using an econometric technique to test the direction of causation (i.e., whether govemance affects growth or vice- versa) for 90 countries for the 1973-90 period. The conclusion is that both economic and civil/political freedom have a strong impact on annual per capita growth. See: Bhalla, S., "Free Societies, Free Markets and Social Welfare, paper presented at the Nobel's symposium on democracy, Uppsala University, August 1994. For a summary of current thinking on the subject see: The Economist, 27 August 1994 (p. 15-17). In addition, civil liberties have been shown to have a strong positive correlation with aid-financed govemment investment projects. See: Isham, J., Kaufman, D. and Pritchett, L., "Govemance and retums on investment: An empirical investigation", Policy Research Working Paper no. 1550, The World Bank, November 1995. Togo: Overcoming the Crisis, Overcoming Poverty 95 7.21 Changing bureaucratic mentalities is a slow process, but a number of measures to improve public sector management could be taken at a relatively low cost and could encourage greater client orientation. While the need to keep busy a bloated bureaucracy could justify lengthy administrative procedures in the past, staff reductions and the continued need to limit personnel expenditures call for streamlined, more "user-friendly" procedures. Government agencies should be asked to prepare a plan in this sense, putting an emphasis on simplifying procedures that are likely to concem the poor. This would imply increased delegation of decision-making authority, and greater decentralization. At the same time, rights and responsibilities within the public sector hierarchy have to be clarified, and objective and transparent criteria for promotion have to be enforced. Although low salaries are cited by civil servants as a justification for requesting the public to give them "special incentives", steps to protect powerless people from such requests should be taken immediately. They could include an information and education campaign on the rights of citizens (e.g., which services must be provided free of charge), disciplinary actions against civil servants found guilty of corruption, random checks, and the establishment of simple mechanisms to address users' complaints. 7.22 Economic Transparency and Efficiency. Greater bureaucratic accountability goes hand in hand with greater economic transparency. Together, they can improve dramatically the efficiency with which public resources are used in pursuing poverty reduction goals. Because investment and current budgeting remain largely independent exercises, investments have often been executed without programming recurrent expenses for operation and maintenance. The 1996 public expenditures review has demonstrated the government's intention to address problems in this area and improve public administration. Among the many encouraging developments, the following five appear the most promising. * A new budgetary nomenclature has been adopted for both current and investment expenditures. It will improve transparency in budget allocations and integrate all central government expenditures into one consolidated budget. * The presently cumbersome procurement and disbursement procedures are under examination and various reform options are being considered. * The accounting procedures and internal control systems of the Treasury are being strengthened by, among others, computerizing Treasury transactions, eliminating certain control steps, and centralizing payment operations in the Treasury. * The preparation and execution of investment projects will be improved by adopting more rigorous procedures. 6 For example, a recent govermment report found that in 1994-95 an average of 168 days elapsed between the time the firms submitted their bids and the award of contracts. See: "Delais de passation des marches", July 1995, DGTP. Togo: Overcoming the Crisis, Overcoming Poverty 96 * The public enterprise sector will be made more efficient by revamping the public enterprise refonn and by strengthening authorities' ability to adequately assess its budgetary impact. 7.23 Improved Functioning of the Judicial System. Law is the expression of the norns and values of a society (or at least of the ruling groups in a society) and plays a crucial role in promoting good governance and participation. The proper functioning of legal institutions is a requirement for sustainable economic development in that they ensure predictability and security of property rights and transactions, and protect against the arbitrary and discretionary power of the state and its agents. In addition, law can be an important safeguard of the interests of vulnerable groups. Conversely, the legal system can also be a barrier to development and an instrument in the perpetuation of vulnerability. A case in point is Togo's Family Code (Code des personnes et de la famille) which contains articles that discriminate against women, thus sanctioning their vulnerability (see Box 7.2). Box 7.2: Women and Law The Constitution of the IV Republic affirms the equality of all citizens before the law regardless of gender. Womnen, however, tend to be discriminated both by customary law and by modem law. Customary law conceives of women basically as goods belonging to their husband's family: they cannot own or inherit land, they cannot inherit any household goods upon the death of their spouse or have custody of their children in case of divorce or widowood. Thev can, however, be "inherited" by their husband's family (assuming they are wanted) along with their children and all other household belongings. Modern law tries to guarantee greater equality, but a number of articles in the Family Code (Code des personnes et de lafamille) leave room for discrimination. For example, Art. 109 allows men to oppose the exercise of a separate profession by their wife. Art.391 stipulates that a woman can inherit her husband's goods only if the husband had previously renounced customary law either by testament or in the presence of an offtcier de 1 'etat civil. In addition, although the Family Code clearly states that in case of death, the surviving spouse will have sole authority over the children, judges tend to give the tutelle and administration of the children's possessions to a member of the deceased husband's family. Even pensions of civil and military servants can only be versed to the widow and orphans if there are written, notarized proceedings (procas verbal lggalis.6 a la mairie et homologue en justice) of the family council allowing it. Sou le: CRIFF/GF2D. "Fenines togoses aujourd'hui et demain - Livre blanc". Lomd 1995. 7.24 In 1978, Togo joined the majority of African countries and unified its judicial system. Traditional tribunals were replaced by the Tribunaux de Premiere Instance (trial courts of initial jurisdiction) which could adjudicate cases according to modem and traditional law.' But in spite of this effort, the judicial system remains plagued by ambiguities. Ordinary citizens who seek legal resolutions may turn to a variety of actors, such as the Tribunaux de Premiere Instance, traditional chiefs, prefects, the marshal's 7Ordonnance no. 78-35 of 7 September 1978, slightly modified by Law no. 31-3 of 30 March 1981. Togo: Overcoming the Crisis, Overcoming Poverty 97 office (gendarmerie) and the police.8 For example, prefects tend to have regular meetings with traditional chiefs (chefs de canton) to settle disputes that could not be settled by the Tribunaux de Premiere Instance,9 and accusations of witchcraft (which are common) are generally handled by traditional chiefs or by marshals rather than by modem tribunals. As often it is not clear who has the mandate to settle which question, powerful individuals can manipulate the system to obtain the desired outcome, but the poor and uneducated may find themselves lost--and therefore more likely to lose. 7.25 While resolving ambiguities in the administration of justice would be an essential step to protect the rights of the vulnerable, it would not be enough. As aknowledged during a two-day national conference on "Justice-Freedom-Security and Development" in 1994,10 the modem legal- system is in a state of crisis: personnel is insufficient (there are only 60 magistrates for the whole country), the operating budget is inadequate (lack of equipment, office space, etc.), and legal procedures are extremely slow (resulting in enormous back-logs). The obligation to be represented by a lawyer, the complication and formality of legal procedures, and the legal fees demanded" make access to modem law problematic. From the point of view of the ordinary citizen, modem law is the last resort. From the point of view of the poor, modem law is simply out of question: it is too costly, too complicated and too slow. 8Van Rouveroy van Nieuwaal, E.A.B.,"Pauvrete judiciaire au Togo: une question a suivre", unpublished manuscript, February 1995. 9To provide an illustration, one such meeting held in December 1994 in the Golfe district dealt with five cases of land dispute and one domestic case. 10 The conference was organized by the Ministry of Justice and took place on 8-9 December 1994 in Lome. I IAt the Tribunal de Premi&re Instance in Tsevie, the following fees were quoted: CFAF 5,000 to open a case, CFAF 15,000-20,000 for transport; CFAF 7,000-15,000 for divorce. Legal fees are detailed in Decree no. 82-50 of 15 March 1982. Togo: Overcoming the Crisis, Overcoming Poverty 98 8. MAIN ELEMENTS OF A STRATEGY: PRIORITIES 8.1 The previous chapter summarized a comprehensive National Program to Fight Poverty, outlined a strategic approach to poverty reduction, and presented a series of recommendations to create the conditions necessary not just to improve the welfare of the most disadvantaged, but to promote social and economic development for the country as a whole. This chapter focuses on the four priority areas identified as key for a strategic approach: access to land, management of population growth, decentralization, and the regionalization of poverty reduction. The last two of these priorities are very closely linked: decentralization is the condition for the elaboration and effective implementation of regional poverty reduction programs, while at the same time the implementation of regional poverty reduction programs would be an instrument to make decentralization a reality and reap its benefits to the fullest. 8.2 The last part of the chapter presents a possible option to give a concrete answer to many of the recommendations made--the establishment of five Regional Development Funds. These funds could represent the vehicle through which regional poverty reduction programs and decentralization could initially take place. They would make it possible to provide assistance in a way that is flexible and creative, going one step beyond "typical" social funds that concentrate on the provision of social infrastructure. Finally, attention is drawn to the importance of monitoring the impact of interventions on the poor, and some short-term actions to this end are suggested. Land Tenure Reform and Access to Land 8.3 As the vast majority of the poor are farmers, improving the poor's productivity depends on their access to potentially productive land. This access is limited by population density and by land tenure arrangements (see Chapter 4). Both of which, in turn, have a negative impact on the environment. Population density leads to the overexploitation of natural resources as fallows become shorter and marginal land is put under cultivation. At the same time, present land tenure arrangements tend to provide little incentive for interventions to protect the environment, because lack of secure rights demotivates "foreigners" (i.e., people not belonging to land-owning clans) from making long-term quality investments in the land they cultivate, and customary law may even forbid conservation measures such as planting trees (see Chapter 4). Women and young people are particularly disadvantaged by prevalent land tenure practices because their access to land is dependent on men and, in the case of women, they will traditionally not be allowed to own or control land. 8.4 Not surprisingly, land tenure issues and access to land were identified throughout the country as crucial to reducing vulnerability and poverty, and while their resolution will undoubtedly be difficult and time-consuming, both government and civil society representatives recognized that ignoring them will simply exacerbate problems and make Togo: Overcoming the Crisis, Overcoming Poverty 99 it much more difficult to find solutions later on'. They also agreed that the 1974 land tenure reform had not succeeded in improving the situation, primarily because it did not take into enough account customary law, and thus it was neither applicable nor applied. Hence the need for an approach which takes tradition as the starting point, and builds on it through a process of consultation, negotiation and consensus building. While in each region there are particular groups that find access to land problematic and for whom specific measures will have to be implemented (e.g., farmers expropriated by OTP in the Maritime region; see matrix where?), a number of actions are recommended nation-wide. 8.5 The fundamental objective of actions to improve access to land should be to establish regulatory frameworks at the community, district and regional levels which protect the interests of vulnerable groups, such as women and pastoralists, and are acceptable to all actors directly concerned. This objective should be pursued in a bottom- up fashion, starting with negotiations at the village level between land-owners, village authorities and other common interest groups (metayers, women, youth, pastoralists, etc.). It is likely that in many cases some outside assistance may be needed from NGOs and front line government workers (extension officers, agents de promotion social, SOTOCO staff) to assist vulnerable groups in organizing for better bargaining power in their negotiations with land-owners and other authorities. Some NGOs are already active in this sense, and their experience should be of guidance. A media campaign, especially by radio, would be helpful in sensitizing communities to the importance of addressing land tenure issues and of striving for arrangements that will result in a more equitable distribution of assets (e.g., women's access to land). Special initiatives should be undertaken to encourage debate, such as a national conference for traditional chiefs or regional conferences with traditional chiefs, land chiefs (chefs de terres) and large land- owners. 8.6 Regulatory frameworks should increase security of access to land for non land- owners, thus encouraging quality investmnents to protect soil fertility and rationalize the exploitation of natural resources. The 1996 agricultural census provides a unique opportunity in this sense, as it can be used as a starting point to clarify the land tenure situation and reassure land-owners of their rights. This, in turn, would encourage land- owners to rent their land and to consider new, more secure forms of access to land, such as bails for several years. For example, a nine-year bail for metayers in the Plateaux has already been successfully negotiated with the help of an NGO (Association Villages Entreprises). 8.7 Access to land should not only be made more secure, but also more equitable and efficient. While population density and soil degradation are a pressing problem in some areas (Savanes, east Kara, Maritime), other areas could support a larger population without damaging the environment (Bassar, Tchaoudjo). Indigenous populations, IAccess to land emerged as a key problem during the participatory poverty assessment. Land tenure problems and their solutions were discussed during the regional workshops mentioned in Box 7.1, as well as during individual interviews with donor representatives, independent consultants and NGOs. Togo: Overcoming the Crisis, Overcoming Poverty 100 however, may be unwilling to welcome "foreigners" and to give them land, especially under non-traditional arrangements such as leases. To facilitate population movements that would promote a more equitable use of land, communities in land-rich areas could be offered incentives to receive migrants, such as preferential treatment for the construction or rehabilitation of schools, health centers or other social infrastructure. Summary of recommendations to improve access to land * Negotiate access to land on the basis of customary law at the community, district and regional levels, with a bottom-up approach. Who? Common interest groups (metayers, women, youth, pastoralists, land-owners, etc.) with assistance/encouragement from development agents (government and NGOs) and media campaign. * Encourage NGOs and government (e.g., extension officers, social development agents) to assist vulnerable groups (women, youth) in organizing for better bargaining power during negotiations with authorities. Who? NGO associations, government, media, donors (by providing financial support for pilots). . Introduce new, more secure forms of access to land, such as bails for several years. Who? NGOs, services techniques, interest groups. * Use the 1996 agricultural census as the starting point to clarify the land tenure situation, e.g., to reassure land-owners of their rights and encourage them to rent land. Who? Ministry of Rural Development, with donor assistance, to feed information at regional level to services techniques and NGOs. * Establish incentives to receive migrants in land-rich areas, such as social infrastructure for receiving communities. Who? Ministry of Plan Managing Demographic Growth 8.8 Economic growth at the national level can only result in lasting improvements in the individual standards of living if the population grows at a slower rate than the national economy and, in the long run, if total population is contained within levels that are environmentally sustainable.2 In the case of Togo, GNP per capita has been on a slowly declining trend since 1972, with a sharp deterioration in the 1990s. UNDP estimates that in 1993 GNP per capita was 52 percent lower than it was twenty years earlier (in real terms). Although some improvement has occurred since 1993, it is clear that a population growth rate of 3.1 percent per year constitutes a formidable challenge to any effort to reduce poverty. As shown in previous chapters, the problem is not just to raise 2 It is assumed that for the foreseeable future Togo will remain an essentially agricultural economy and vast population movements outside the country could not be sustained by neighboring countries, many of which experience similar population problems. See also the discussions on land availability in Chapter 4 and on food security in Chapter 5. Togo: Overcoming the Crisis, Overcoming Poverty 101 GNP per capita. With the population doubling every 22 years, simply maintaining the presently insufficient access to basic social services (education, health, water and sanitation) will require increases in resource allocations to these sectors that are incompatible with the financial and human capital available in the foreseeable future, even under the most optimistic assumptions about public sector management. Finding employment for the increasingly large cohorts of young people entering the labor market will be another major problem, which could easily lead to social unrest as idle and disillusioned youth take to the streets. Strong action to change population dynamics is therefore needed.3 8.9 Togo does not yet have a population policy. An initial declaration of population policy was elaborated in 1988, but it has not been adopted, so population matters are still influenced by policy decisions which lack a coherent and realistic approach. A case in point are laws based on the 1920 French code which prohibit the marketing and distribution of contraceptives. Beyond such measures, a comprehensive population policy should be adopted without further delay, widely publicized and supported by the allocation of resources commensurate with the importance and the scope of the actions to be taken. While the budgetary constraints faced by the government are well understood, greater efficiency in public expenditures along the lines illustrated in the previous chapter could go a long way in making available some of the financial resources needed. Evidence from around the world, including Sub-Saharan Africa, shows that much can be done to influence population dynamics even under tight budgetary situations, provided the political commitment is strong.4 8.10 A central element of the population policy should be efforts to limit fertility rates and promote responsible parenthood. These would include: (a) pervasive information and education campaigns for the population at large, as well as more targeted campaigns for adolescents and for men, highlighting the problems linked to reduced birth spacing, and early and late motherhood; (b) training of health personnel and staff from the Ministry of Women in Development and Social Affairs in family planning; (c) support to and coordination of all public and private institutions involved in the promotion of family planning activities; (d) the immediate repeal of the 1920 contraceptive laws; (e) a campaign to discourage early marriages and the pursuit bridewealth (la dot) as an instrument to mobilize emergency funds. 3While the urgency of the poor's problems would appear to argue in favor of a strategy giving priority to fast-acting measures rather than interventions that need 10-20 years to produce significant results, as it is the case with population interventions, wavering would make the success of other measures in the fight against poverty more costly, and in the end short-lived. 4Kenya and Zimbabwe are examples of countries which have been able to lower considerably their population growth rates. Togo: Overcoming the Crisis, Overcoming Poverty 102 8.11 Fertility rates are closely related to women's status. Indeed, the most important variable determining fertility behavior is women's education.5 More educated women tend to have fewer children and to have them appropriately spaced. Women with fewer children, in turn, tend to have better fed and better educated children. The opposite, of course, is true: uneducated women easily fall into a vicious circle of repeated childbearing and lower productivity (because of childbearing, lactation and childcare demands) which results in high dependency ratios, poor nutritional standards and uneducated children. If women derive their social value mainly from their maternal role, as it is the case in most of Togo, there will be few incentives for women to try to limit family size. 8.12 This suggests that efforts to control population dynamics would have to be complemented by measures to improve women's condition. Increasing female education is probably the most effective, especially if it is accompanied by revisions in the curriculum--and in teachers' training curricula--that promote a view of women as equal to men and reinforce girls' self-esteem. Giving women economic and legal rights equal to men's should also be a priority goal (see Box 7.2), and should be accompanied by an education campaign to inform women and girls of their rights. The media, the Ministry of Women in Development, the Ministry of Social Affairs, NGOs and women's associations have an important role to play in this sense, which should be supported by appropriate resources from the government and donors alike. 5 For a review of evidence worldwide see: Shireen Jejeebhoy, Women's Education, Autonomy, and Reproductive Behavior: Experiencesfrom Developing Countries,. Oxford, Clarendon Press, 1995. Togo: Overcoming the Crisis, Overcoming Poverty 103 Summary of recommendations to manage demographic growth * Mount a pervasive information and education campaign highlighting problems related to reduced birth spacing, and early and late motherhood. Who? Ministry of Women in Development and Social Affairs, in collaboration with Ministry of Health, Ministry of Education, Ministry of Communication and Culture, donors, NGOs (ATBF especially) and media (espacially radio). * Mount targeted sensitization, information and education campaigns for men and adolescents, stressing the importance of responsible parenthood and, for adolescents, the dangers of early motherhood. Who? Ministry of Women in Development and Social Affairs, in collaboration with Ministry of Health, Ministry of Education, Teacher Associations, Ministry of Communication and Culture, donors, NGOs (ATBF especially) and media (especially radio). * Train health personnel and staff from the Ministry of Women in Development and Social Affairs in family planning. Who? Ministry of Health, Ministry of Women in Development and Social Affairs, donors (to help finance training). * Support and coordinate all public and private institutions active in the promotion of family planning. Who? Ministry of Health, NGO associations, donors. * Repeal 1920 contraceptive laws. Who? Parliament. * Mount a campaign to discourage early marriages. Who? Ministry of Women in Development and Social Affairs, media, Associations Parents d'Eleves, Teachers' Associations, NGOs, Confederation Nationale des Chefs Traditionels. Complementary measures: increase female education reinforce women's and girls' self-esteem give women equal rights, and inform them of their rights. Decentralization 8.13 Article 141 of the October 1992 constitution stipulates that the country is organized in collectivites territoriales following the principles of decentralization while respecting national unity.' Up to now, however, deconcentration--rather than decentralization--continues to be the approach in place. It is in practice a way to reduce the burden of the central government and increase its outreach which consists in delegating some administrative decisions while keeping the bulk of the decisional power 6 The movement toward decentralization started over twenty years ago. Art. 7 of the ordinance no. 75-26 of 28 July 1975 authorizes the creation of regional planning and development committees, while art. 9 creates local committees at the prefectural level. These committees were in principle to have responsibility for the conception, elaboration, monitoring and execution of programs and projects. In reality, committees were never given sufficient autonomy to carry out their responsibilities. (See also: UNDP, Avant-projet de programme national de lutte contre la pauvret6 (PNLP), draft, December 1995.) Togo: Overcoming the Crisis, Overcoming Poverty 104 in Lome. While a number of sectoral ministries (e.g., education, health) appear to be making some progress in achieving greater decentralization, many routine administrative decisions have still to be authorized by Lome and the planning process is still very centralized. Regional Planning and Development Directorates (DRPD) lack autonomy and decisional power, and regional development plans, which should be at the basis of effective decentralization, have not been elaborated. 8.14 A Ministry of Interior and Decentralization was created in response to the 1992 crisis, and the nomination in November 1995 of a Junior Minister (Ministre delegue) for Decentralization augurs well, but resources allocated for decentralization purposes have been meager and, more importantly, progress made in transferring power to regional centers has been limited. Yet, decentralization has a crucial role to play in the fight against poverty, because it is the key to improved delivery and better targeting of public services. Because the voices of the poor are weak, they are less likely to be heard in far away Lome than in the regional or district capital, so services which are centrally planned are less likely to be responsive to the poor's needs and concerns. Should the services be responsive, central management tends to slow down implementation and makes it more difficult to adopt a flexible approach that adapts to local variations. 8.15 Putting into practice the decentralization principles outlined in the 1992 constitution requires an approach that combines institutional coherence with regional specificity. Keeping in mind that there is room for considerable variation in the appropriate design and form of a decentralization process, the following four areas of action should be considered: a) Improve the institutional framework by creating a viable bottom-up structure and by clearly defiming the responsibilities, obligations and prerogatives of central, regional and local authorities. Social capital' should be strengthened through the establishment of committees at the local, departmental and regional level where civil society and government would work together. Each region should have autonomous deliberating and executive bodies, such as a Regional Council (made of elected representatives) and a Govemor. Within this frarnework, the role of central government would focus more on coordination, assistance, guidance and monitoring. b) Strengthen regional institutions by giving them the authority to identify projects to be financed by the central govemment, and therefore included in the PIP. These projects 7Social capital can be simply defined as the capacity of people to work together. It relates to the trust, norms, and reciprocity networks embedded in social organizations. Social capital is a primary determinant of social and economic performance, both at the community and at the regional or national level. For example, research in poor urban communities has shown that the erosion of social capital during economic crisis can break down community-based systems for delivery and maintenance of social and physical infrastructure. See: Moser, C., "Confronting Crisis: A Summary of Household Responses to Poverty and Vulnerability in Four Poor Urban Communities" (ESD Studies and Monographs Series no. 7). Washington DC: World Bank, 1996. For further reading on social capital see: Putnam, R., Making Democracy Work: Civic Traditions in Modern Italy. Princeton NJ: Princeton University Press, 1993. Togo: Overcoming the Crisis, Overcoming Poverty 105 should be identified following a bottom-up approach, whereby village committees make the initial proposition, which is then discussed by the canton, district and regional committees. Village Development Committees (Comites Villageois de Development or CVD), Local Planning Committees (at the district level) and Regional Planning Committees already exist (although not all villages have a CVD), but in many instances they are dormant or they are perceived as imposed by outside political interests, and in any case they have little power. Some NGOs, however, are working with CVDs and it is conceivable that such institutions could be revitalized, depoliticized and empowered to play a leading role in defining poverty reduction interventions, and in contributing to their effective implementation. Such committees should comprise elected representatives of the population (women, youth, food crop farmers, cash crop farmers, artisans, etc.), as well as representatives of regional government bodies and NGOs. c) Mobilize local finances, as decentralization by definition means that localities take over responsibilities and the taxing power or financial resources needed to meet them. New revenue sources should be identified, and old ones strengthened. This should be accompanied by the establishment at local levels of at least the essential public expenditure management systems (e.g., forecast and accounting systems to monitor and control cash flows) and by the adoption of budgetary procedures compatible with regional control over allocations. d) Encourage decentralization of staff by providing detailed descriptions of (new) regional posts, by making regional rotations an integral part of civil servants' carriers, and by providing incentives to work outside Lome. At present, transfers to the regions are often felt like a punishment for a number of reasons: responsibilities in regional posts are ill-defined, regional staff are marginalized as important decisions are taken in Lome, "field" work is not valued, regional centers have inferior infrastructure compared to the capital, Lome offers greater opportunities to earn money in private sector activities (either as second jobs or for other family members). Giving regional posts well-defined responsibilities and real power would help in establishing an organizational culture whereby regional posting is perceived as a privilege rather than a punishment. Fostering the development of regional capitals is a longer-term and much more costly project, but its benefits would go far beyond attracting competent civil servants. 8.16 While decentralization in the public administration is essential, it should be accompanied by decentralization in the private sector as well. It is therefore encouraging that NGOs have started to decentralize: some years back only a few were operating outside the capital or the Maritime region, but by now there is considerable NGO presence in all regions, although NGOs are by no means evenly distributed throughout the country and the greatest concentration is still in Lome. The next step would be for NGOs not only to decentralize, but to form regional associations so as to better coordinate approaches, collaborate and coalesce around issues of common interest. Togo: Overcoming the Crisis, Overcoming Poverty 106 NGOs active in the Savanes are an example (they have formed the Federation des Organisations de Developpement de la Region des Savanes or FODES). Summary of recommendations for decentralization * Create institutions at local, district and regional levels so as to form a viable bottom-up structure within which civil society and government can collaborate. Existing CVD, CLP and CRP could be revitalized. * Establish autonomous deliberating and executive bodies for each region. * Clearly define responsibilities, obligations and prerogatives of central, regional and local authorities. * Establish budgetary procedures compatible with regional control over allocations, including the authority to mobilize finances at the regional and local level. * Identify new sources of regional and local revenues, and strengthen old ones. * Establish public expenditure management systems at regional and local levels (e.g., forecast and accounting systems). * Make regional rotations an integral part of civil servants' carriers. * Give regional posts well-defined responsibilities and real power. * Provide incentives to work outside Lome (for the longer-term: foster development of regional capitals). Regional Action Plans 8.17 Much can be done to reduce poverty with interventions of a national scope planned centrally, but the causes and manifestations of poverty are also context-specific, as they reflect agro-ecological conditions and cultural patterns which are not shared by the whole country. Similarly, the opportunities that can be exploited or created to improve the poor's welfare vary greatly across agro-ecological zones and administrative subdivisions. Because agro-ecological conditions are fairly homogeneous within regions, following regional demarcations is a useful way to group areas with common problems and opportunities. A poverty reduction strategy with a strong regional dimension, as advocated here, would still imply an important role to be played by the central government. In particular, this would: provide general policy guidance and a facilitating environment (e.g., laws and regulations), coordinate and monitor activities in all regions, optimize resources allocation in response to regional priorities and needs, ensure that efficient and consistent implementation processes are in place, facilitate information exchange and cross-fertilization of ideas between regions and with neighboring countries, and ensure inter-regional cooperation. Togo: Overcoming the Crisis, Overcoming Poverty 107 8.18 A regional approach would have two main advantages. First, it would facilitate decentralization and a bottom-up approach, as it would give regional development actors a key role to play in defining action plans and it would shorten the distance between the ultimate beneficiaries and the decision-makers. Second, it would make it possible to tailor interventions in a way that is responsive to the priority problems of each region, thus encouraging a more selective and efficient use of public resources. In the Savanes, for example, where the soudure lasts several months and malnutrition rates are high, the most pressing problem is food security; in the Plateaux, on the other hand, two harvest seasons a year make food security a relatively marginal issue, so poverty reduction interventions should concentrate elsewhere. 8.19 Well-focused, implementable regional action plans for poverty reduction can only be the result of extensive consultations involving grassroots communities as well as all development actors operating in the region. Although this may be time-consuming, the process of elaborating action plans itself is likely to have a positive impact, for it will encourage dialogue and require collaboration, thus building social capital. The results of an initial attempt by government technical services and NGOs to define regional strategies for poverty reduction are summarized in a matrix at the end of this chapter (this work is tentative and is provided mainly as an illustration). There are important differences among regions in terms of the relative importance with which a particular problem is felt and in terms of the solutions envisaged to similar problems, but it is possible to group recommendations under nine general themes. 3 Access to land. Although access to land is an issue that has been identified as a national priority, it has also a markedly regional dimension. So, while everywhere it was suggested that land tenure problems be solved through a local approach first, the massive presence of northern migrants in the Plateaux requires measures that are very different from those required in the Maritime, where the problems of access to land have more to do with high population density and the presence of OTP (see Chapter 3). * Soil degradation. Loss of soil fertility, and therefore reductions in yields, was lamented everywhere. Among the solutions envisaged was the establishment of regional land use plans that would help farmers identify environmentally acceptable and profitable crops. At a lower scale, the promotion of village land management schemes (gestion du terroir) by SOTOCO, DRDR and NGOs was often seen as a possible solution to both environmental and land tenure problems. Organic fertilizers should also be promoted, taking their traditional (and often forgotten) use as a starting point. * Water. Access to water is a pressing problem in the Savanes, but also in a number of districts outside the Savanes (e.g., Haho in the Plateaux). In the Savanes it involves access to potable water (especially in the dry season) as well as access to water for crops and animals. Community participation in planning, managing and maintenance was considered essential for a sustainable solution in all cases. Togo: Overcoming the Crisis, Overcoming Poverty 108 * Food security and food crop production. Cereal banks were seen as a promising way to cope with food shortages in the north. In addition, rural communities should be trained in resource management, including stocks (to prevent post-harvest losses and waste) and land (gestion des terroirs). Diversification (e.g., legumes) and counterseasonal activities (e.g., horticulture) should be promoted. Better access to inputs could be achieved through farmers' organization (done already by some NGOs for bulk purchasing of fertilizers) and micro-finance services (see below). Extension has an important role to play: it should be more demand-driven, more timely, more oriented toward women, and more practical. * Ceremonies. Traditional beliefs and behaviors were often identified as one of the main causes of vulnerability. Ceremonies in particular encourage conspicuous consumption that often leaves the celebrants severely impoverished.! This, in turn, was seen as a manifestation of the more general problem of resource management, whereby households may take short-sighted or otherwise ill-advised decisions in the use of their stocks (e.g., by squandering their grain supply to celebrate a funeral) or their land. Changing traditions is not easy, but intensive campaigns, perhaps targeting traditional authorities, accompanied by training in stock management would help. Traditional forms of communication, such as women's songs during village celebrations or the dunku,9 could be effective vehicles to pass messages. * Credit. Improving access to credit is one of the main priorities for both the rural and urban poor. Although there was no agreement over the role to be played by savings (as a pre-condition, as a compulsory component of repayments, or as a separate service?), there was no doubt that micro-credit programs should be promoted to improve farmers' access to agricultural inputs and means of production (tools, animals, land), and to help artisans and microentrepreneurs start or expand their businesses. Young people should be particularly targeted by these programs. • Transport. Maintenance of feeder roads should be improved by organizing communities and providing them with a minimum of assistance (tools, training), and the establishment of a regional road maintenance fund should be considered. The presence of military road blocks was deeply resented everywhere, but particularly in the north. Road blocks should be eliminated. * Education. Although increasing enrollment rates is important, an effort in this sense would be meaningless if quality is not improved--school should be made relevant and s Ceremonies play an important role in building social cohesion, and therefore strengthening social capital. The issue here is not the existance of ceremonies, but the apparent inflationary nature of ceremonial expenditures. 9 The dunku (literally: "the eye of the village" in Ewe) is the "social communicator" of the village. It is an individual who knows everybody's business in the village, who is universally liked, and who makes fun of people but who is also carefully listened to. In many ways, the dunku is the ideal vehicle to promote community participation, inform, persuade. Togo: Overcoming the Crisis, Overcoming Poverty 109 useful for the poor, especially during the first years. This would mean teaching practical topics (nutrition, hygiene, civic education, basic business arithmetic, family life, etc.), experiment with multi-level classes, provide assistance to "clandestine" schools and turn them into community schools. To facilitate access, pilot initiatives would include providing credit for school-related expenditures and creating satellite schools where schools are too far for the youngest children to walk. Health. The implementation of the Bamako Initiative has greatly improved access to health, and it should be sustained. Drug procurement and distribution, however, are still problematic in some areas and should be improved while continuing to rely on private providers. Pharmaceutical kits should be set up in all villages. 8.20 In addition to these nine themes, two more themes kept re-emerging: women's underprivileged status and youth unemployment/underemployment (not only for graduates), which is one of the main causes of rural-urban migration. Many of the measures proposed above address these problems, but perhaps a more concerted effort focusing on them would be needed. Also, because 80 percent of the poor live in rural areas, the attention has gone mainly to finding solutions to rural poverty, but it is clear that urban poverty requires a distinct approach. Initial discussions focusing on Lome called for measures to create employment, such as support to microenterprises and labor- intensive public works, and measures to improve the habitat, such as social housing programs and waste disposal ("Avant c'etait Lomg la belle, maintenant c'est Lome la poubelle"). Regional Development Funds 8.21 If regional action plans for poverty reduction have to become a reality in the near future, a mechanism has to be established whereby funds to carry out recommended actions are directly available at the regional level, not just to government agencies but to all institutions that have been identified as capable of assisting in implementation. At present, even a concerted effort to decentralize decision-making and give regional government structures direct access to public finances would not represent an adequate solution, because non-governmental structures would be excluded. When resources are scarce, as in the case of Togo, the challenge is to create a system capable of using effectively--and equitably--contributions coming from as many sources as possible (communities, NGOs, professional associations, traditional authorities, private businesses) and in as many forms as possible (knowledge, creativity, technical skills, labor, capital). 8.22 While there is no blueprint to follow to meet this challenge, the experiences of other countries facing similar problems may be helpful. For exarnple, the 6S Project (Se Servir de la Saison Seche au Sahel et dans les Savanes) based in Burkina Faso provides soft loans for a wide variety of activities aiming at improving productivity during the dry season; in Cote d'Ivoire, the PNGTER (Programme National de Gestion des Terroirs et d'Equipement Rural) provides a multi-sectoral fund to stimulate rural development in Togo: Overcoming the Crisis, Overcoming Poverty 110 response to community demand.'" One option for Togo would be to establish five Regional Development Funds (one per region). The structure, operation and cost of Regional Development Funds could only be specified after much consultation with all major actors involved, so what follows should be interpreted more as an effort to stimulate thinking than as a prescription. 8.23 Similarly to social funds, and in line with the approach recommended for the definition of regional action plans, Regional Development Funds would work in a participatory and bottom-up fashion aimed at empowering grassroots communities. Like social funds, they would finance the construction and reparation of small social infrastructure (schools, health centers, water points, feeder roads, bridges), assist in the establishment of money-making projects (credit or grants for micro-enterprises, storehouses, market facilities, small irrigation schemes), and help communities become less vulnerable through capacity building and organization." Their scope of action, however, would be broader than that of typical social funds in two main ways. (a) A wide range of activities capable of improving the welfare of the poor in the region could be financed. So, in addition to traditional social fund activities, Regional Action Funds could, for example, organize fairs to bring business to the area, launch competitions for low cost money-making ideas, pursue twinning arrangements with European schools, conduct marketing campaigns for local products, set up a regional broadcast service, etc. Regional Development Funds should encourage creativity in finding solutions to the problems identified, and infuse a dynamic and entrepreneurial spirit into the fight against poverty. (b) They would be free to use whichever intermediary they choose to deliver services and goods, as long as the guiding criteria are competence and reliability. NGOs, extension services and social development agents (APS) would clearly play a very important role, especially for community outreach activities, because of their experience and interests. But consulting companies, entrepreneurs, professional associations, churches and other organizations or businesses would be considered as well. 8.24 Needless to say, Regional Development Funds should have extremely simple and transparent mechanisms for decision-making and financial management. They would be 10 The PNGTER could represent a particularly useful experience because it combines a number of elements contained in the strategy proposed in this document. It has three components: one focuses on land tenure issues, one focuses on communities empowerment through the definition of alternative schemes for managing village resources (amenagement du terroir), and one provides a mechanism for financing the scheme selected by the community. II For a review of social funds see: Marc, A., C. Graham and M. Schacter, "Social Action Programs and Social Funds" (AFTHR Technical Note no.9). Washington DC: World Bank, 1993. Carvalho, S. "Social Funds: Guidelines for Design and Implementation" (HRO Working Paper no.34). Washington DC: World Bank, 1994. Schmidt, M. and A. Marc, "Participation in Social Funds" (ESD Paper no.4). Washington DC: World Bank, 1995. Togo: Overcoming the Crisis, Overcoming Powrty III autonomous from each other and from the central government, meaning that each Regional Development Fund would have complete authority to dispose of available funds as long as this is done in accordance with pre-established criteria and procedures. In addition, they would be free (actually encouraged) to do their own fund-raising and to conclude independent aid agreements with private, bilateral or multilateral donors. In this sense, funds provided by the government could be considered as seed money. 8.25 An important advantage of this approach would be that it could not only be an instrument of decentralization, but it could also facilitate geographic targeting. For example, all development indicators show that the Savanes is by far the most disadvantaged region, both in terms of population welfare (poverty incidence and severity, enrollment rates, malnutrition rates) and in terms of government assistance (number of schools and teachers, health facilities, health personnel). Regional Development Funds would allow the government to channel extra-assistance to the Savanes. This could be done by adopting criteria for the allocation of funds that take into consideration the population to be served as well as the relative poverty of regions (e.g., using a poverty weight), so that more disadvantaged regions will receive proportionately larger Regional Development Funds. Monitoring Poverty 8.26 Policies and programs designed to reduce poverty require a clear understanding of the nature and extent of the problem. Once policies have been adopted and interventions are ongoing, information is critical to evaluate their effectiveness and fine-tune their design. An accurate measurement of poverty, therefore, is an investment. It is not costless, but it can save considerable money and suffering if it helps to focus limited resources where they will be most effective. Reliable and detailed information on the magnitude and characteristics of the phenomenon will facilitate accurate targeting of poverty-oriented interventions. Monitoring the actual impact of policies on the poor may help reduce wasteful subsidies or review well-intentioned but counterproductive decisions. In addition, the measurement of poverty has a governance dimension, in that it encourages accountability."2 8.27 Large-scale, nationally representative income and expenditure household surveys, such as the EBC, are extremely important to establish base-line data, monitor trends, and conduct in-depth analysis by region, gender, income source and other socioeconomic variables. But serious efforts to monitor poverty cannot depend on occasional large surveys because of their cost, their technical difficulty and the time it takes to produce results. Effective monitoring demands the establishment of data collection systems, that is, established ways to obtain information and feed it to users at periodic intervals. The importance of a continuous and reliable flow of information is most apparent in the food 12Sec: World Bank, "Managing Development: Tlhe Govenance Dimension- A Discussion Paper', SEC M91-820, June 1991. Togo: Overcoming the Crisis, Overcoming Poverty 112 security and nutrition sectors. It is easy to see that if information about food shortages is late in coming, the consequences may be disastrous. Regular data-gathering, however, should not be considered a luxury to be afforded only for the most extreme risks. And it does not have to be expensive or cumbersome. Light and focused surveys are a well- tested way to assess changes in the characteristics of poverty as well as the impact of specific policies or projects (the series of Enquetes Legeres aupres des Menages carried out with UNDP assistance is Benin is an example). Routine administrative records on health, education, agricultural extension, water and sanitation provide important information which could easily be exploited for poverty monitoring. 8.28 For cost-effective poverty monitoring the Government should not only gather its own data, but also take advantage of information available from other sources. NGOs often have information which is under-utilized because administrators (and donors) may not recognize that they could use it, and that the NGOs have it. Secondary level training institutes and university departments may also have useful information which goes untapped. Therefore, the Government should strive to develop channels to share information with donors and NGOs, focusing especially on information gained from implementing (as opposed to planning) poverty alleviation projects. Informal occasions in which different parties will feel free to openly discuss their failures and successes are likely to result in much cumulative learning and more effective interventions. 8.29 Because knowledge is essential for effective action, the recommendations offered below should be carried out in the short term, that is, within a year or so. The estimates included should be considered just as benchmarks. (a.) A new household budget survey should be carried out as soon as possible, provided substantial technical and financial assistance are made available. An interdisciplinary approach would facilitate a comprehensive treatment in data gathering and analysis, and, especially, in interpreting findings. While the main unit of analysis should be the household, the survey should be designed in such a way as to allow for the measurement of intra-household allocations, since this would permit an analysis of intra-household inequality. Ideally, this type of survey should be repeated every five years. (Estimated cost: US$600,000). (b) National-level nutritional indicators should be collected regularly, possibly through a collaboration between the university, the Ministry of Health and OCCGE/CRAN. (Estimated cost: US$90,000). (c) Light and focused surveys combining quantitative and qualitative analysis should be carried out at least once a year to monitor household-level consumption and other poverty-related indicators both in rural and urban areas. Particular attention should go to sampling, especially in the rural areas, to optimize representativeness within the limits of relatively small samples (sampling based on agro-ecological zoning and periodic rotation could help in this sense). (Estimated cost: US$70,000). (d) Socio-economic analysis should be complemented by the analysis of cultural variables which are likely to play a major role in shaping vulnerability as well as in determining the appropriateness and effectiveness of poverty alleviation Togo: Overcoming the Crisis, Overcoming Poverty 113 interventions. For example, the important role played by traditional medicine, food taboos and traditional ceremonies in people's lives suggests that systematic analysis of the relationship between these variables and vulnerability could significantly increase our understanding of poverty and stimulate innovative approaches to its reduction. The social sciences departments of the university could play an important role. (Estimated cost: US$70,000). The Virtuous Circle 8.30 The National Program to Fight Poverty contains a very comprehensive list of interventions that would contribute to poverty reduction. Given the limited resources and implementation capacity presently available in Togo, chapters 7 and 8 sketched a strategic approach that identifies a limited number of mutually-reinforcing interventions. First, three pre-requisites were identified: better governance, macro-policies that promote growth and improved social services. Then four priorities were discussed: improving access to land, managing demographic growth, decentralizing and adopting regional action plans. The establishment of Regional Development Funds was proposed as a vehicle through which regional poverty reduction programs and decentralization could initially be implemented. 8.31 All of the recommended interventions are justified on their own merit, but it should be recognized that there is a synergistic effect, meaning that the contribution of each intervention to the fight against poverty will increase with the implementation of other interventions. Because poverty and vulnerability result from the interaction of a number of factors, an effective strategy should be based on the interaction of measures to change these factors. In other words, if poverty is a vicious circle, getting out of poverty requires a virtuous circle. Togo: Ovcrcosing tdc Crtis, Owcoming Poy 114 REGIONAL POVERTY REDUCTION ACTIONS ISSUE SAVANES KARA CENTRALE PLATEAUX MARITIME Access to -negotiate access to land -negotiate access to land -negotiate access to land -negotiate access to land -negotiate access to land land (women based on customary law based on customary law based on customary law based on customary law based on customary law & youth) at community, district at community, district at community, district at community, district at community, district and regional level; and regional level; and regional level; and regional level; and regional level; -introduce new forms of -introduce new forms of -introduce new forms of - introduce new forms of - introduce new forms of access, such as bail or access, such as bail or access, such as bail or access, such as bail or access, such as bail or rent; rent, or collective land rent; rent- build on A.V.E. rent-)reassure land- -take advantage of titles; -take advantage of experience with 9 yr. owners of their rights agricultural census to -take advantage of agricultural census to bails for metayers; (take advantage of clarify situation and agricultural census to clarify situation and -take advantage of agricultural census) to identify possible clarify situation and identify possible agricultural census to encourage them to rent solutions identify possible solutions clarify situation and land; -encourage NGOs & solutions -encourage NGOs & identify possible - to clarify situation and govt to assist vulnerable -encourage NGOs & govt to assist vulnerable solutions identify possible grs (women, youth) in govt to assist vulnerable grs (women, youth) in -encourage NGOs & solutions organizing for better grs (women, youth) in organizing for better govt to assist vulnerable -encourage NGOs & bargaining power; organizing for better bargaining power; grs (women, youth) in govt to assist vulnerable -establish incentives to bargaining power; organizing for better grs (women, youth) in receive migrants (from -establish incentives to bargaining power; organizing for better west to east), such as receive migrants (Bassar, -establish incentives to bargaining power; wells or schools to Dankpen), such as wells receive migrants, such as -orient agricultural receiving communities. or schools to receiving wells or schools to activities toward less communities. receiving communities. land-demanding ones: fisheries, small animal husbandry, horticulture; -revalue land already I__________ ___________________ __________________ exploited by O TP. Togo: Overcoming the Crisi, Overcoming Poverty 115 ISSUE SAVANES KARA CENTRALE PLATEAUX MARITIME Soil -SOTOCO should -Promote agro-forestry -SOTOCO should -Include in bail contract -Require OTP to return degradation sensitize & train farmers (e.g., CARE Intl); sensitize & train farmers requirement for land to owners in usable on soil degradation and -promote use of organic on soil degradation and afforestation (A.V.E. shape: fish ponds, regeneration; fertlizer (e.g., SOS regeneration; experience); remblayage, -establish regional land Enfance); -establish regional land -promote improved afforestation; use plan identifying -SOTOCO should use plan identifying fallow (e.g., cajanus -establish regional land profitable & environm. sensitize & train farmers profitable & environm. cajan); use plan identifying acceptable crops; on soil degradation and acceptable crops -establish regional land profitable & environm. promote coop NGO + regeneration; use plan identifying acceptable crops; DRDR/SOTOCO on -establish regional land profitable & environm. encourage use of organic pilot village land mgt use plan identifying acceptable crops fertilizer and rock schemes, after training profitable & environmn. phosphate; front line staff; acceptable crops -promote use of organic fertilizer Water -Rehabilitate irrigation -Rehabilitate irrigation -Rehabilitate and -Create village water infrastructure; infrastructure develop water committees; -build small dams infrastructure (e.g., -IEC on potable water (barrage zone de Haho) Borgou); -encourage water harvesting; -install boreholes and wells; -use water committees as entry points for I community devlopment _ Togo: Overcoming the Crisis, Overcoming Poverty 116 ISSUE SAVANES KARA CENTRALE PLATEAUX MARITIME Education -Start primary in local -Start primary in local -Start primary in local -Start primary in local -Start primary in local language on a pilot basis, language on a pilot basis, language on a pilot basis, language on a pilot basis, language on a pilot basis, with progressive switch with progressive switch with progressive switch with progressive switch with progressive switch French; French; French; French; French; -teach topics relevant to -teach topics relevant to -teach topics relevant to -teach topics relevant to -teach topics relevant to the majority (community the majority (community the majority (community the majority (community the majority (community input): nutrition, input): nutrition, input): nutrition, input): nutrition, input): nutrition, hygiene, env/sustainable hygiene, env/sustainable hygiene, env/sustainable hygiene, env/sustainable hygiene, env/sustainable agr., civic rights, basic agr., civic rights, basic agr., civic rights, basic agr., civic rights, basic agr., civic rights, basic business, family life; business, family life; business, family life; business, family life; business, family life; -multi-level classes (train -multi-level classes (train -multi-level classes (train -pilot ecoles de relais for -pilot ecoles de relais for teachers); teachers); teachers); the youngest (CP1-CM1) the youngest (CP1-CM1) -pass and apply law -turn clandestine schools -turn clandestine schools -experiment with credit preventing early in community schools, in community schools, for school supplies marriage for girls; with govt T.A. with govt T.A. (MFFR experience); -turn clandestine schools -turn clandestine schools in community schools, in community schools, with govt T.A. with govt T.A. Food -Set-up cereal banks for -Set-up cereal banks; -Use inter-disciplinary -Encourage farmers to -Provide extension on security and small groups (build on -train communities on approach to extension, organize (don't force horticultural production; food crop RAFIA experience); resource mgt: stock not just agronomists; organization!) to buy -encourage production -train communities on (ceremonies!) and land; -provide extension to inputs and market crops diversification (e.g., resource mgt: stock -in Massif Kabye, women; leguminose) and (ceremonies!) and land develop agr. activities -help farmers get intensification; (gestion des terroirs); requiring little land: organized for fertilizer -increase access to inputs -diversify, develop small animal husbandry, distribution through micro-credit; counter-seasonal horticulture; -train communities on activities (horticulture); -provide extension to stock mgt (prevent post- -provide extension to women. harvest losses); women -provide extension to women Togo: Overcoming the Crisis, Overcoming Poverty 117 ISSUE SAVANES KARA CENTRALE PLATEAUX MARITIME Ceremonies -Conduct intense IEC to -Conduct intense JEC to -Conduct intense IEC to -Conduct intense IEC to -Conduct intense IEC to limit excesses (e.g., see limit excesses; limit excesses limit excesses; limit excesses Cote d'Ivoire); -take advantage of -link with training in -link with training in Kabye custom whereby stock mgt stock mgt women sing identifying "defaulting" husbands; -link with training in stock mgt Credit -Set up micro-credit -Set up micro-credit -Set up micro-credit -Set up micro-credit -Set up micro-credit programs (savings, group programs for access to programs for access to programs for access to programs for access to solidarity etc.) for access agr inputs (savings, agr inputs (savings, agr inputs (savings, agr inputs (savings, to agr inputs and means group solidarity etc.); group solidarity etc.); group solidarity etc.) group solidarity etc.); of production (artisan -organize young artisans -provide start-up capital -set-up micro-credit tools, animals, land) to obtain group credit for to young people to start programs for artisans and tools (e.g., Don Bosco). farm or business SMEs Transport -Fix-up north-south -Fix-up north-south -Organize communities -Organize communities trunk; trunk; for maintenance of for maintenance of -organize communities -repair Bassar-Kabou feeder roads, provide feeder roads, provide for maintenance of road; basic tools; basic tools; feeder roads, e.g., -organize communities -establish a regional road -establish a regional road provide basic tools, agree for maintenance of maintenance fund, maintenance fund, with SOTOCO to help feeder roads, e.g., decentralize decisions; decentralize decisions; with maintenance of all provide basic tools, agree -eliminate road blocks -eliminate road blocks feeder roads (not just with SOTOCO to help cotton); with maintenance of all -eliminate road blocks feeder roads (not just cotton); -build rural roads in the west; -eliminate road blocks Togo: Overcoming the Crisis, Overcoming Poverty 118 ISSUE SAVANES KARA CENTRALE PLATEAUX MARITIME Health -Keep up private generic -Set up iodine deficiency -Provide primary health -Set up iodine deficiency -Ensure women represent drugs procurement program for Douffelgou, care to new villages in program for Amou and in Village Health through NGO; Binah, Koza: iodized oil Sotouboua and Blitta Haho: iodized oil & salt, Committees -continue setting up & salt, IEC, "parrainez IEC, "parrainez un boites pharmaceutiques un goitreux" goitreux"; in villages -improve OCDI's drug procurement & distribution (but leave I out govt) ANNEX 1 Page I of 15 Methodological Notes on the Calculation of a Poverty Line The Data 1. The 1987-89 Household Budget Survey (EBC)2 provides the most recent database on detailed household expenditures. The first phase of the survey (1987) covered Lome and three other regional capitals: Atakpame, Sokode, and Kara (Lome et Grandes Villes). The survey consisted of a number of modules covering, among others, household food consumption, household expenditures, and household composition. These three modules provide a complete database that permits the calculation of a poverty line. The second phase of the EBC, which was conducted in 1989, consisted of a similar survey covering the remainder of the country: the remaining regional capital Dapaong, a sample of the smaller towns (Centres Urbains Secondaires), and a sample drawn from all rural communities (Communautes Rurales). Together these two data sets provide a comprehensive coverage of Togolese households for the 1987-89 period. Unfortunately, the survey provides relatively little data on nutritional status.2 2. The two phases of the survey employed slightly different methods. The first phase (Lome and large cities) took place between February 1987 and March 1998. During this period, household expenditures were recorded during a full month, while food preparation and consumption were observed during one week of that month. Food consumption was not observed in single-person households, because most food in such households is prepared and consumed outside the house. However, the expenditure data also cover single-person households. Own- consumption, that is, food obtained from home production or as a remittance from elsewhere, was not included in the expenditure measurement, while, of course, the consumption of home- produced food is included in the food preparation and consumption observations. The second phase (small towns and rural communities) started six months after the completion of the first phase in October 1988, and continued till November 1989. Own-consumption was measured during this phase, while households were visited twice for a full week, at an interval of six months. Single-person households were covered also for food preparation and consumption. In general, the data of the second phase are more complete than the data collected during the first phase, but on the whole the data collected in the two phases are comparable, also because own- consumption plays a much smaller role in the urban environment. I Enquete Budgetaire, Telegramme 1-13. This series of reports presents detailed tables of the most important subjects covered by the survey. Further reports are still in preparation. 2 Anthropometric data were collected for all household members, but the precision of measurement was not sufficient to allow computation of the usual indices, especially for children. For example, ages are available only as number of years completed, weights are to the nearest kilogram, and heights are in centimeter intervals. However, calculation of the usual Height for Age and Weight for Height anthropometric indices for children under five years of age requires ages in months, weights to at least 0.1 of a kilogram, and heights to 0.1 of a centimeter. ANNEX I Page 2 of 15 3. In 1995, UNDP financed a survey to collect updated information on poverty conditions in Togo by interviewing a subsample of the same households that were interviewed during the 1987-89 survey. Although this new survey constitutes an important first step to evaluate current developments, methodological difficulties render survey results unsuitable for comparisons with many of the estimates obtained from the EBC. The EBC carefully controlled for seasonal fluctuations in household incomes and market prices (the price levels of certain staple foods change considerably in the course of the agricultural year), and a parallel analysis ensured that the monetary valuation of auto-consumption was based on realistic local prices. The scope of the 1995 data collection was insufficient to allow similar calculations, and the survey results for households outside Lome appear to be particularly affected by this limitation. In addition, limiting the sample to previously interviewed households introduced a bias that is likely to have resulted in an under-representation of the more vulnerable households (e.g., older people living alone, families who lost their home, recent migrants). Therefore, such data are useful in defining general trends, but a point by point comparison with estimates based on the EBC survey would be misleading. Such a comparison, however, seems most urgent, because a number of important changes that took place recently (civil disturbances, the nine-month strike, the CFA devaluation) are likely to have affected household expenditures considerably, and may have also altered income distribution patterns. 4. In the absence of a comprehensive up-to-date household budget survey, an indication of the effects of recent events on the poverty status of households can be obtained by looking at changes in the cost of typical consumption baskets and in the earning power of households3. In the case of Togo, data exist that document both changes in the cost of living and changes in household incomes. 3 Changes in the cost of living. Prices have been recorded over the years by the Direction de la Statistique (DS) and by the Direction des Enquetes et Statistiques Agricoles (DESA). For Lom6, a regular price index is calculated on a monthly basis, with observations covering a wide range of consumption items. Outside Lome, observations of prices are made with varying regularity in a range of communities, covering markets in different parts of the country. Although the material is often incomplete, for most products it is possible to compare prices as they evolved in several regions over a number of years. Farm gate prices, or estimates thereof, are available for all regions, and in most cases are recorded on a monthly basis. Combining the price data available from these different sources results in relatively accurate estimates of the changes of the cost of living in terms of current prices. * Changes in household incomes. Agricultural production data concerning the major food crops and all export crops are available as a yearly time series by region. DESA maintains an annual database of food crop production, while the cash crop production is annually recorded by OPAT. These data combined with producer price data provide one of the major sources of If an increase in the cost of living is not met by a comparable rise in income, the number of poor households increases to the extent that the cost of living increases; if the cost of living remains constant, higher incomes result in less poverty, lower incomes in more poverty. ANNEX 1 Page 3 of 15 information on rural income, and can be used to obtain approximate estimates of changes in rural incomes for each of the five regions. However, there is no equivalent proxy for estimating changes in urban incomes. Until such information becomes available, only very tentative estimates based on, for instance, the national accounts. Fortunately, the 1995 UNDP-financed survey provides some important insights into the current conditions of Lome households. Data validity 5. The quality of the data available from the EBC is generally good. The survey was well conceived and quality control has been thorough. The data of the survey are accessible in the form of a well maintained data base --Banque d'Information Permanente sur les Conditions de Vie des MJnages (BIP). The sample size is large enough to estimate a number of parameters at the regional level, enabling the analysis of poverty to gain in relevance as a tool for policy evaluation. With a few exceptions, the data show a consistent pattern, well in line with the findings of independent survey data concerning nutrition. The levels of poverty obtained from the EBC data correlate with the nutritional data collected in 1976-77 and 1988. 6. There exist no independent data sources that can be used to validate the production and price data collected by DS and DESA. The data series collected by these two organizations vary in quality. Data collection takes place at the major urban and rural markets, and records are available on a monthly basis. At times, collection has been very systematic, with monthly collection at all markets; other times there are large gaps, generally because the relevant data sheets were not received in headquarters or because the data was not collected to begin with. To obtain complete series, it is necessary to estimate such missing data, and for a number of items it is necessary to correct estimates for seasonal price differences.4 In the end, the available data are sufficient to enable proper seasonally corrected estimates for the major crops, taking into account that there are regional differences in the timing of annual seasonal cycles for different crops. For most crops, complete price data are available for at least five different years. 7. The data collected for 1991-92 and 1992-93 are necessarily affected by the turmoil in large parts of the country during those years. Retail price data for that period are completely absent, although there is information on most farm gate prices. In most regions, collection of price data in markets and shops resumed in 1993. In early 1994, immediately after the devaluation, prices were highly uncertain, but around April most prices had stabilized again, and the data used to estimate changes in prices concern the only prices from April onwards. Because 1993 prices are also available for most regions, the effect of the devaluation on price levels can be examined to some extent, and it is possible to look separately at the long range trend of prices before the devaluation. Price variations between 1989 and 1994 can thus be checked for unlikely figures that do not correspond to the expected effect of the devaluation (such as 200 percent 4 In some cases, data collection did not take place during several consecutive months. These missing values have to be replaced by values that reflect both the usual seasonal price level for those months and the overall price level for the year in question. ANNEX 1 Page 4 of 15 increases of imported products) and the trend existing between 1989, 1990, and 1993.5 The household data collected by the 1995 survey in Lome indicate that the cost of living index as produced by the DS reflects actual Lome conditions. Selecting a Calculation Model 8. To calculate a poverty line several alternative approaches exist. Common methods are the Cost of Basic Needs (CBN) method and the Food Energy Intake (FEI) method.6 In both cases the poor are defined as those persons (or households) who are unable to satisfy a number of basic needs, usually taken as the possibility to achieve (I) a food intake that satisfies minimum energy requirements, (2) sufficient drinking water, (3) sufficient clothing, (4) sufficient housing, (5) sufficient health care, and (6) sufficient education. Obviously, the definition of these basic needs is to a large extent arbitrary, but by basing the criteria of sufficiency on socially accepted definitions, and by basing the bundle of needs used in the definition of poverty on actual consumer choices, this arbitrariness is somewhat reduced. The CBN method is based on a detailed analysis of the consumption patterns found in the population, but goes beyond that by setting an external norm of what constitutes the minimum budget required. The choice between adhering to consumer choices made by the population or referring to an external norm will always remain arbitrary and open to discussion. The FEI method is an attempt to remove these arbitrary aspects altogether by adopting food energy intake as a primary reference criterion. In this method, people are considered poor if they do not satisfy the minimum required energy intake; the poverty line is defined as the income below which people do not consume the required amount of calories. In this way, the FAO/WHO minimum energy requirements can serve as a relatively objective yardstick. It must be recognized, though, that the intuitive advantages of the FEI method are often offset by its data requirements.7 5 If certain imported products show less than a 100 percent increase between 1989 and 1994, their price must have gone down in the interval 1989 - 1993. In fact this is the case for a large number of goods, possibly as a by- product of economic chaos which resulted in a drop of demand for many goods. 6 For a recent discussion on poverty definitions using these two methods see Ravallion and Bidani, 1994. 7 There are two other main disadvantages with the FEI method. One is that it is insensitive to differences in resouce allocation within the population to which it is applied. The other is that it assumes fairly normal conditions, so in the case of sudden crises (e.g., droughts), when the whole population may experience distress, the FEI method would fail to produce a relevant poverty line. The CBN method is superior under such circumstances. ANNEX 1 Page 5 of 15 Box 1A.I: Other Approaches to Measuring Poverty A pragmatic approach to poverty measurement used in several recent poverty assessments (e.g., Nigeria, Ghana, CMte d'lvoire) is to arbitrarily set the poverty line at two-thirds of the mean income of the population, and an extreme poverty line at one-third of the mean. Although poverty lines calculated in this way can be used for most policy purposes, they bear no relationship to poverty in the absolute sense of the word, and comparisons between countries have relatively little meaning beyond what is usually presented by way of a Lorenz curve or a Gini index. The same measure can be used to calculate a poverty line in rich countries, and would, depending on the distribution of income, give similar numbers of poor as in poor countries, something which is not the case with poverty lines defined according to the CBN or FEI method. In Togo the poverty lines calculated following this method would be 53,500 CFAF and 26,800 CFAF. This would result in a poverty count of approximately 38 percent. An important disadvantage of this type of poverty lines is that it does not allow for meaningful regional comparisons, because the lines calculated for regions depend on the distributions only and do not account for differences in price levels and consumption patterns. 9. When applied to a country with considerable regional variation, overall poverty lines can become meaningless. In Togo, setting a single national poverty line that is applied to the population as a whole results in significant overestimates of the number of poor households in rural communities compared to urban communities, particularly in the north. For example, a single national poverty line of 58,000 CFAF results in 12 percent poor households in Lome and 78 percent poor households in all rural communities, ranging from close to 70 percent in the southern regions to close to 95 percent in the northern regions. These differences reflect to a certain extent differences in consumption patterns and in the cost of living (highest in Lome, lowest in rural areas) and could be misleading in the formulation of poverty reduction policies. It was therefore decided to conduct the analysis for each region separately. Fortunately, the EBC data are sufficiently detailed to compute regional poverty counts, which means that the FEI method represents a sensible approach. The CBN method, which can be also applied by region, is less suitable given the existing data, because a good CBN poverty line requires insight into the choice processes and alternatives considered by poor consumers. In this case, economic and political changes in the past few years were of such magnitude that it would have been inappropriate to combine recent interview data (from the participatory poverty assessment) with pre- 1 990 quantitative survey data. The data requirements of a CBN poverty line are usually smaller, in terms of quantifiable household information, than those of a FEI poverty line, but to estimate a valid poverty line requires a combination of quantitative and qualitative information. Without sufficient insight into the choice process, the calculated poverty line may have little relevance. ANNEX 1 Page 6 of 15 Calculation details 10. The approach followed by the FEI method is to establish the income level at which the food needs of the household are satisfied. The cost of those food needs is equal to the extreme poverty line, and the income level itself is the poverty line. Calculations of this type can be best done on households consisting of more than one person, because data concerning food consumption and expenditures of single-person households tend to be less reliable: more food is consumed outside the household in restaurants or with family, and the consumption pattern is influenced by differences in time allocation; in addition, this type of households often is supporting dependents elsewhere. A preliminary step in the analysis was to examine the data for the most common expenditure patterns. This immediately revealed important regional and rural/urban differences. For instance, poor households were more often consuming sorghum and millet, while rich households generally consumed maize, a finding which runs counter to expectations since maize is cheaper. The main reason for such counter-intuitive correlation (similar examples existed for other staples, such as cassava and yams) was that low incomes were concentrated in the north of the country, where sorghum and millet are grown because they are more drought resistant, while maize does better in the more humid parts of the country further south. 11. This implied that budgets should be analyzed separately for different parts of the country, to account for the different costs imposed by each type of consumption pattern. Furthermore, in each region the urban population appeared to follow a mixture of local food habits and those of other urban areas. Characteristically, more convenience foods are consumed by urban populations, and this has important consequences for the cost of living. Based on these considerations, it was decided to perform separate calculations for the five administrative regions, each of which has a particular combination of agricultural and environmental characteristics linked to distinct consumption patterns. To account for rural/urban differences, each regional sample was further subdivided in three groups according to community types: urban, small town, and rural. Together this made fifteen combinations, but the sample size of small towns in the Savanes region did not allow separate treatment, and was joined with Savanes urban thus leaving us with fourteen different groups (see Table lA.l). Table 1A.l: Population size in 1987-89 Community Type Region Urban Small Towns Rural Maritime 503,944 88,613 480,215 Plateaux 41,478 76,963 510,233 Centrale 39,810 42,829 199,561 Kara 41,330 60,470 278,261 Savanes 38,573 - 312,262 Source: BIP, Direction de la Statistique 12. To establish the minimum food requirements, the usual procedure is to compute the needs of the population using a formula that requires values for a large number of variables: the sex and age composition of the population, the average weight of adults in the population, the fertility ANNEX 1 Page 7 of 15 rate (i.e., the number of pregnant women needing supplementary food and the number of babies needing breastfeeding), the activity level of the population (PAL), the amount of extra nutrition to recover from bouts of illness (which is directly related to the type and frequency of morbidity), and the amount of food needed to allow for catch-up growth in malnourished children.9 Most data necessary for these calculations were not available and it was decided to use approximate values which had been previously calculated for Benin.'" Since calculations were required for each of three environments, minimum food requirements were set at 1,990 calories per day per person for urban, 2,000 calories per day per person for small towns, and 2,080 calories for rural (small towns were placed at an intermediate point, since a part of their population is farming). 13. The next step is to estimate the income levels at which these food needs can be satisfied by examining the relationship between total expenditures and caloric consumption. This assumes that nutritional needs are satisfied when caloric needs are satisfied, which is usually the case when cereals make up a sizable proportion of the diet, as in Togo. A difficulty was that the individual household expenditure data showed large variations, which are due to the fact that monthly expenditures were obtained on the basis of collection during a single month rather than from a monthly average derived from yearly expenditures. As a result, the regressions between total expenditures and caloric consumption were not well determined, with extreme values dominating the outcome. To simplify the procedure it was decided to start with determining the relationship between food expenditures and caloric intakes. This provided a stronger relationship, although for some sub-groups the relationships remained relatively weak, which demonstrates the range of choices open to households, such as to buy cheap unprocessed foodstuffs, instead of more expensive substitutes." 14. The relationships established at this stage provided the equations that were used to calculate the extreme poverty line. In developing countries, food expenditures of poor households usually range in the neighborhood of 70 percent of their total expenditure. The EBC data showed that this approximate relationship also existed for poor households in Togo. Food expenditures of poor households (around the thirtieth percentile) amounted to approximately 60 percent of total budget in urban households, 70 percent in small towns and 80 percent in rural areas. Using this factor to calculate the income needed to cover non-food expenditures gives the poverty line. Table I A.2 presents the results of these calculations. 9 See James and Schofield, Human Energy Requirements - A Manual for Planners and Nutritionists. New York: FAO & Oxford University Press, 1990. 10 World Bank, "Benin: Toward a Poverty Alleviation Strategy" (Report no. 12706-BEN). Washington DC: World Bank, 1994. 11 Several regressions were calculated. To remove part of the individual variation, households were grouped in twenty equally sized groups each representing five percentiles of the expenditure distribution. The calculations were subsequently performed on the average caloric consumption and expenditures for each of these groups. Both linear and logarithmic relationships were tried. The parameters of the equation with the highest R-square were employed in further calculations. The lowest R-squares were found for the small towns in Centrale, due to a relatively small sample size and some obviously atypical households. R square values averaged around 51 percent. ANNEX 1 Page 8 of 15 Table 1A.2: Poverty Lines: Calculations, Assumptions, Lines, and Results Poverty Lines Assumptions (1000 CFAF/year) Households Individuals Caloric Share of Food Absolute Poverty Core Non- Requirements in total budget Poverty Line Line Poor Poor Poor Core Poor Non- Poor Poor Urban Lom6 1,990 60 33.6 55.9 2.7% 9.8% 87.5% 3.8% 13.1% 83.1% Atakpawn 1,990 60 25.2 42.0 4.6% 2.9% 92.5% 7.5% 4.5% 88.0°/e Sokodd 1,990 60 22.7 37.8 3.5% 18.7/o 77.8% 2.7% 14.9% 82.5% Kara 1,990 60 17.8 29.6 1.8% 10.00/% 88.2% 3.7% 11.4% 84.9% Dapaong 1,990 60 21.4 35.7 6.1% 16.0% 77.8% 5.8% 18.1% 76.1% Small towns Maritime 2,000 70 23.4 33.4 10.4% 14.3% 75.3% 9.3% 21.2% 69.5% Plateaux 2,000 70 21.1 30.2 6.8% 5.1% 88.1% 6.5% 8.4% 85.00/% Centrale 2,000 70 21.7 31.0 2.4% 11.2% 86.3% 2.6% 16.1% 81.3% Kara 2,000 70 16.0 22.9 13.6% 14.00/o 72.4% 11.2% 15.2% 73.6% Rural Maritime 2,080 80 28.3 35.4 18.3% 16.4% 65.3% 25.0°/o 19.8% 55.2% Plateaux 2,080 80 21.0 26.3 3.7% o 7.5% 88.9°/o 3.9o/o 8.2% 87.90/o Centrale 2,080 80 23.7 30.0 10.90/o 12.6%/ 76.6%/ 11.8% 13.6% 74.6% Kam 2,080 80 23.3 29.1 34.3% 18.0%/e 47.70/ 39.5% 19.4% 41.1% Savanes 2,080 80 26.4 33.0 44.2% 21.6% 34.1% 47.8% 21.0%/e 31.3% Source: EBC data and staff calculations. 15. Two important remarks concerning these poverty lines are in order. The first concerns the obvious limitations of poverty lines based on household income only. Ideally, poverty lines should include the results of an analysis of public spending, as household welfare may vary significantly according to the level of public spending contributions to their budget. Unfortunately, available data do not allow for such an analysis. In general, though, the regional distribution of public expenditures seems to confirm the pattern calculated here, as the tendency of poverty incidence to increase from south to north is reinforced by the relative low public expenditure in the north compared to the south (on a per capita basis). A second remark concerns the calculation of a single national poverty line. It is possible to approach this problem in two ways. The first would be to compute a weighted average of the lines presented in Table IA.2, using the population size of each group as weights. This average would give 25,700 CFAF per capita per year as a line of absolute or extreme poverty and 35,700 CFAF per capita per year as a poverty line. Another approach would be to consider the urban minimum, with the poverty lines of Lome as a target for the remainder of the country (33,600 CFAF per capita per year for absolute poverty and 55,900 FCFA per capita per year for poverty). Neither of these lines are useful to estimate a poverty count, however, because they ignore differences in the cost of living, and the second implicitly take Lome's expenditure patterns as the norm. ANNEX 1 Page 9 of 15 Changes since 1989 16. After 1989 several important changes took place. The most important of these, the devaluation of early 1994, had a significant impact on the cost of living. To estimate this impact, an effort was made to compute the increase in the cost of living since 1989 for different parts of the country. An official index figure is available for Lome only. On the basis of this index figure, by December 1994 the cost of living had increased to 142%. In other parts of the country the cost of living increased slightly less, mainly because the consumption basket is less dependent on imported goods and transportation costs. Index figures were calculated for the major items in the consumption basket as established by the EBC. Table 1A.3: Price changes of major product categories between 1989 and 1994 (1989=100 Products Plateaux Centrale Kara Savanes Maize 113 105 77 86 Rice 108 123 107 95 Wheat 147 147 147 147 Sorghum/Millet III 101 80 85 Cassava 102 98 69 69 Yam 130 113 72 53 Tubers 109 120 98 98 Sugar 173 168 154 154 Beans 82 98 73 75 Nuts 122 132 109 127 Vegetables 138 122 94 94 Fruits 138 67 96 96 Meats 119 120 108 108 Fish 137 159 145 145 Milk and Eggs 176 195 162 162 Oils 180 135 128 128 Spices 148 76 140 140 Alcoholic beverages 195 188 199 199 Stimulants (coffee, tea) 180 180 180 180 Non-food 144 134 136 136 Source: Staff calculations based on data provided by DS and DESA. 17. DS maintains a regular retail market and shop prices survey in different parts of the country. The data of this survey are often not complete, but in response to the devaluation, a concerted effort was made to systematically collect prices on a monthly basis in all major retail markets throughout the whole country. Similar data could be traced for 1989, except for the Maritime region. In some cases important items were lacking during some months, but missing values could be approximated in almost all cases by using the available seasonal data series from the years for which information was available. Producer prices collected by DESA provided a further opportunity to validate price data. Table IA.3 presents the increase in prices that was recorded for the retail markets in four of the five regions. Using the index figures shown in this table, it is possible to calculate a cost of living index for each of the four regions correcting for the different consumption bundles of consumers in urban, small-town, and rural environments, as ANNEX 1 Page 10 of 15 obtained during the household budget survey. For the Maritime region a similar series was calculated using the changes recorded for agricultural producer prices to estimate the corresponding changes in retail prices. Table lA.4 shows the cost of living index figures that were obtained following the methods outlined here. Table 1A.4: Cost of Living Increase 1989 - 1994 (1989=100) Urban Lome 142 Atakpame 141 Sokode 129 Kara 129 Dapaong 123 SmaUl towns Maritime 137 Plateaux 135 Centrale 126 Kara 122 Rural Maritime 136 Plateaux 133 Centrale 125 Kara 111 Savanes 113 Average Togo 136 Source: Staff Calculations based on 1989 EBC and Table IA.3 ANNEX 1 Page 11 of 1S Methodological Notes on the Rapid Poverty 4ppraisal The Rationale for Rapid Appraisal I1. General sample surveys are relatively inefficient instruments for studying poverty conditions in detail. The reasons have to do with some of the most common characteristics of poverty, namely that poverty often occurs in pockets, that poverty is related to intra-household differences, and that its causes and manifestations tend to be multi-dimensional. (a) General sample surveys are most efficient when the variables to be measured are evenly spread among the study population. The main reason for this is that cluster sampling, which is the common method for containing survey costs, becomes relatively ineffective when the subject of interest is unevenly spread and causes a high variance between clusters. This is common in cases where poverty is correlated to geographical conditions, such as poor soils, lack of water or distance from markets and social services. Besides the obvious statistical argument involved here, there are other important reasons why poor people are often not well represented in sample surveys. The poor, and especially the extremely poor, are often without fixed abode, not registered or difficult to reach. This tends to result in under-representation, or road-side bias, in big surveys. (b) Intra-household differences are not easily identified by household surveys, which by definition tend to concentrate on the household as a whole. The economic characteristics of households are usually defined as some sort of sum-total of the contributions and expenses of individuals, with a bias towards the household head- often automatically assumed to be a man regardless of the actual distribution of responsibilities and power inside the household. There are however, well-documented differences among household members in terms of their access to resources and assets, and these differences tend to be directly related to vulnerability. Attributes like age and gender are probably the most common determinants of intra-household inequality, but their relationship to vulnerability and poverty is notoriously difficult to study in a household survey, unless it is itself the focus of research. (c) The multi-dimensional nature of poverty is probably the most compelling argument for the use of rapid appraisal techniques. While sample surveys such as household budget surveys are effective tools in capturing the economic aspects of poverty, they will tend to reflect the researchers' definition of poverty in that they will measure only the dimensions that the researchers themselves have included in the questionnaire. Pilot testing of questionnaires helps offsetting this inherent bias, but it is highly unlikely that a standard questionnaire can capture some of the dimensions of poverty which do not easily lend themselves to quantification, such as powerlessness or isolation (special or social). Sample surveys are based on standard questionnaires which, by definition, sacrifice richness of information to uniformity in view of quantification. As a result, unexpected information is not easily captured, nor are cultural or political dimensions. Rapid appraisal techniques are far more flexible and therefore provide an essential complement ANNEX I Page 12 of 15 (and validation) to the "dry" quantitative data produced by sample surveys. Moreover, the multi-dimensionality of poverty-and in particular its relationship with vulnerability- makes it very difficult to identify the causes of poverty through standard questionnaires. This is so for two main reasons. One is that socio-cultural and political variables may play a crucial role but go undetected; the other is that poverty, tends to result from a number of factors which interact in ways that statistical analysis may be unable to capture, unless extremely large and unwieldy samples are used. 2 Because of the rigorous procedures specified in sample surveys, such surveys, at least in theory, are reproducible. This is much more difficult to accomplish with rapid appraisal methods, which is one of the reasons why there is always a need to strike a balance between the quantity of data to be collected in a rapid appraisal and the current state of knowledge conceming its subject. The argument here is as follows. If little is known conceming poverty in a country, the rapid appraisal technique necessary has to become more like a survey, and the nunber of communities to be visited has to increase to limit the factor of subjectivity in the selection procedure. If much information is already available, the rapid appraisal can concentrate on understanding and filling the gaps of previous knowledge; the danger of subjectivity in selection is much less, because the available information base provides a safeguard against this problem and allows for purposive sampling. In such cases, the method will not contain more error than the already available data, because it takes those data as its starting point. 3. Obviously, rapid appraisal is not a replacement of the sample survey but an additional instrument. When and where should this technique be applied? This is a matter of cost- effectiveness and of the suspected impact of-weaknesses of the usual sample survey techniques as described above. Sample surveys are costly; rapid appraisal techniques are far cheaper. If part of the data to be collected are already available, the arguments for a full-fledged sample survey approach are relatively meager: the gain to be expected from a dedicated sample survey are much smaller if already much is known. Time limitations are an important factor: it takes easily a year before results from sample surveys become available (it took two years for the first EBC results to become available), while rapid assessments produce results in as little as a couple of months (findings from the field work for the Rapid Poverty Appraisal in November-December were used in the Country Environmental Strategy Paper circulated in draft in March). Another important argument for rapid assessment procedures relates to the desired quality and reliability of the data. Sample surveys are not well suited to areas in which supervision of enumerators is difficult, and in particular when it is important to reach marginal locations which are difficult to access. As poverty is often linked to such variables, the data collected by sample surveys on poor people often lack both in quality (lack of supervision) and quantity (under-representation). ANNEX 1 Page 13 of 15 Sampling Issues 4. Rapid poverty appraisal methods cannot replace sample survey methods in the study of poverty, but they are effective instruments for obtaining data under precisely those conditions that make survey methods more difficult to apply. Instead of concentrating on representation through sampling, rapid appraisal concentrates on relevance and understanding. For this purpose, a rapid appraisal survey will address sample conditions rather than sample cases, often studying the whole population of conditions. What this means in practice is best explained by considering the geographical factors related to poverty. In a rapid assessment, a first analysis considers the geographical dimensions of poverty. If climate or soil types are factors causing low incomes among farmers in part of the country, the assessment will have to include examples of all possible combinations of these factors. Technically, the number of conditions included in the survey will of course have to be kept limited, but by discussing this issue with local experts, studying the relevant maps, and examining previous survey results, a selection can be made that represents the relevant variation in conditions. 5. Appraisal usually follows a hierarchical approach. At each level-national, regional, district, local-a careful analysis considered the geographical and social distribution of poverty, combining the data obtained from different sources such as interviews, official statistics, surveys and studies. Because the data collected were to be used also as an input for the Country Environmental Strategy Paper, particular attention was given to the relationship between poverty and environment, thus ensuring that the sample of conditions contained areas considered vulnerable because of specific environmental problems, such as resettlement and mining areas. A special role was played by experts and key informants, including NGO representatives (local and international), government field workers (especially extension agents), specialized donor staff, university researchers and other people who had intimate knowledge of local conditions, even though they were not directly involved in poverty-related interventions. The appraisal concentrated on visits to rural communities (40 villages covering the five regions) and, to a lesser extent, to urban neighborhoods (only in Lome). These constitute the lowest level of the hierarchy as outlined above, and arguably the most important. 6. It is often argued that a rapid appraisal of the kind outlined here lacks objectivity. Since the method is not based on sampling, it is thought to be biased and to lack the scientific rigor involved in survey methods. These arguments are largely beside the point. By basing themselves on a hierarchical approach, rapid appraisals are population-based techniques, not sampling-based. They do not aim at representing the population by taking a small random sample, but instead rely on taking the whole population of experts (that is, official and non- official authorities, including available documents) as a starting point. From there on the procedure moves systematically, and choices that are made in terms of selection of lower levels depend on the contacts with those met at the higher levels, but cross-checked at each point by less formal interviews outside the hierarchy. 'Me interviews at community level are not different in this respect. These interviews are usually made to cross-check data obtained earlier; they are not meant as a random sample. Even at the very lowest level, selection is systematic: households will be visited in different parts of the community for example, to check on the impact of isolation, differences in opportunity because of access to water, and so on. Or in other cases, ANNEX 1 Page 14 of 15 households will be visited because of certain handicaps, such as women heads of households, widowers, disabled people. The method of selection, at all levels, is best compared to that followed in experimental science: at each step a comparison is made between different conditions, and this comparison suggests conclusions that may be tested later on. Data Gathering in the Field 7. Before starting field work, local researchers (12 in total) extensively discussed the purpose of the data-gathering exercise and possible ways to obtain needed information in a culturally sensitive and practical way. Much insight was gained from these discussions, as the local researchers--all jeunes diplomes and almost all sans emploi--had experience in field work and in many cases were quite familiar with living conditions in vulnerable situations (see Box IA. 1). During visits to rural communities and urban neighborhoods, a multi-pronged interview technique was used to gather information on the conditions of the poor at the individual and household level. Interviews were semi-structured (the themes to cover had been agreed by the whole team), covering a variety of subjects including quick budget appraisals, problem hierarchies and their perceived solutions, assessment of survival strategies, and, very important, the definition of what is poverty as seen by the poor and their communities. Short life histories were also collected in some cases because they show the ways in which people become and stay poor, as well as people's attempts to escape from the poverty trap and the relative success of different survival strategies. This type of informationi is left out by the usual household survey, and longitudinal or time-series data are extremely rare because of their high cost. Short life histories collected during rapid appraisals may provide precious insights into the dynamics of poverty at a very low cost. Box 1A.1: Simple Non-Monetary Poverty Indicators at the Household Level * Quality of soap: commercial soap for the rich, local soap for the others (may use shell of cashews a soap), no soap for the extremely poor. * Use of soap: to wash everything (clothes, dishes, people) for the non-poor, to wash only people for the poor, to wash only adults (no children) and/or only on special occasions (market day, ceremonies) for the very poor. * Roof quality: iron sheets for the non-poor, straw for the poor, straw with holes for the extremely poor. * Door quality: wood for the non-poor, strings or rags for the poor, no door (or insufficient strings/rags) for the extremely poor. * Bouilli (porridge-like food): with peanuts for the non-poor, without peanuts for the poor. * Sauce: with meat or fish for the non-poor, with only vegetables and maybe a bouillon cube or tomat paste for the poor, no sauce (just salt and pepper, or not even that) for the extremely poor. * Number ofdaily meals: three for non-poor, two for poor, one or none for extremely poor. Source: group discussion with Rapid Poverty Appraisal Team. 8. No questionnaires or interview guides were used. Having understood the purpose of the data-gathering exercise and agreed on the main topics of interest, it was up to each individual researcher to use his or her skills to approach people and obtain information. To facilitate ANNEX 1 Page 15 of 15 comparisons and ensure that important information would not be lost, all researchers filled out information sheets for each household contacted and each village (samples are attached). Thorough debriefing and discussion at the end of each day enabled researchers to learn useful techniques from each other, evaluate the reliability of the information gathered (e.g., was there general agreement within one village about the numnber of households unable to eat two meals a day on a regular basis? was it true that the closest well was 5 km away?), venture some quantitative estimates at the community level (e.g., prevalence of maternal mortality, rate of unemployment), formulate research hypotheses, and identify themes to be included (or excluded) the following day. This approach allowed for high flexibility and produced a much greater contribution (and commitment) from researchers than is usually the case. 9. Information gathered during field work was used also to validate the material collected from the expert interviews and the statistical data. By confronting such data with observations and interviews at household and individual level, the bias that is often inherent to expert data and statistics becomes visible. Another advantage is that these "low level" interviews generate hypotheses that can be used to re-analyze the statistical data or during later interviews with experts. A case in point are the workshops held as a first step toward the formulation of regional action plans for poverty reduction. Discussions during these workshops were especially fruitful because the organizers were familiar with findings from the Rapid Poverty Appraisal. Furthermore, such discussions showed that NGO and government representatives active at the regional level have a good knowledge of the manifestations and causes of poverty in their region, of the survival strategies already employed by the poor, and of other possible solutions. Many NGOs in particular demonstrated that they are close to the grassroots and are able to see poverty through the eyes of the poor, as their contributions often paralleled results from the Rapid Poverty Appraisal. Indeed, a working matrix summarizing findings from the field work turned out to be quite similar to the matrix summarizing the output of the workshops (in Chapter 8), hence lending credibility to both exercises. ANNEX 2 Page 1 of 9 Table 1A.1: Value Added for Main Food Crops, 1983-90 Avre Ave 1983 1934 1985 1953485 1986 1937 193 1989 1990 1955-90 CASSAVA Production, '000 tonnes 345 444 474 421 411 355 413 409 593 436.2 Producer Price, CFAF/IKg 35 33 29 34 30 32 26 30 23 28.2 Value of output, m.CFAF 12075 16872 13746 14231 12330 11360 10739 12270 13639 12067.4 YAM Production, '000 tonnes 383 342 364 363 409 360 379 405 391 388.8 ProducerPrice,CFAF/Kg 89 31 64 78 61 63 63 68 63 64.6 Value of output, m.CFAF 34087 27702 23296 28361 24949 22680 23877 27540 26588 25126.8 MAIZE Production, '000 tonnes 145 222 182 183 127 172 296 287 285 233.4 Producer Price, CFAF/Kg 104 75 47 75 73 72 79 60 61 69 Value of output, m.CFAF 15080 16650 8554 13428 9271 12384 23334 17220 17385 15928.8 SORGHUM Production, 000tonnes 79 113 95 97 131 93 119 153 115 123.2 Producer Price, CFAF/Kg 11 89 58 52 76 66 98 69 68 75.4 Valueofoutput,m.CFAF 869 10502 5510 5627 9956 6463 11662 10557 7820 9292.6 PM Total Food Crops Total m. CFAF 86195 102707 75157 88019 81719 77140 95739 96162 87253 87612.6 InputCosts,m.CFAF/b 11127 10676 8417 10073 9076 8393 9565 10316 9680 9406 VALUE ADDED, m.CFAF 75063 92031 66740 77946 72643 68747 86224 85346 77573 78206.6 Source: WB Agricultural Sector Review (Green Cover 1993) ANNEX 2 Page 2 of 9 TABLEAU 1A.2: Budget de fonctionnement, Coefficients budgetaires et Taux de depenses sociales. BUDGET DE FONCIIONNEMENT (en miliers) POURCENTAGES BUDGETAIRES TAUX DE DEPENSES BUDGET SOCL4LES ANNEES (Mdi F) (1) SANTE (2) EDUCAT.(3) AFF.SOC.(4) SANTE(5) EDUCAT. (6) AFF.SOC. (7) (8) =(2=3=4)1(l) 1980 67,27 2837944 12042182 402834 4.21 17.90 0.59 22.70 1981 70,27 2724053 13639045 364713 3.85 19.30 0.51 23.66 1982 72,04 3488144 13905355 381916 5.24 19.30 0.53 25.07 1983 75,80 3403283 14148050 389616 5.50 18.66 0.51 23.67 1984 76,79 3347127 15053043 401424 4.35 19.60 0.52 24.47 1985 81,89 3477845 14529751 389377 4.24 17.74 0.47 22.45 1986 87,28 34999996 15828413 430554 4.00 18.13 0.49 22.62 1987 89,69 3485651 16091425 423384 3.91 17.94 0.47 22.32 1988 89,69 4204716 18554649 525268 4.68 20.68 0.58 25.94 1989 92,49 45733434 19939807 495380 4.94 21.56 0.53 27.03 1990 92,49 4926759 20749547 331917 5.32 22.43 0.59 28.12 1991 92,49 4927337 21604199 689510 5.53 23.35 0.74 29.42 1992 93,63 5156289 24037929 797550 5.57 25.67 0.85 32.03 1993 75,98 3088003 15360316 570106 4.06 20.21 0.75 25.03 1994 121,11 6062686 24958734 891361 5.00 20.60 0.73 26.35 Source: Eguida, "Base de donn6es socio-6conomiques pour le developpement humain durable", PNUD, 1994 ANNEX 2 Page 3 of 9 Table 2A.1: Food Consumption and Poverty Core Poor Region Category Maritime Plateaux Centrale Kara Savanes Total Maize 27.1% 10.7% 3.9% 8.3% 3.5% 11.4% Millet, Fonio, Sorghum 0.7% 4.7% 12.5% 27.1% 37.7% 15.3% Rice 7.2% 7.6% 20.7% 7.8% 9.9% 10.6% Wheat 3.4% 0.7% 0.4% 0.9% 0.6% 1.3% Yam 2.3% 18.7% 6.5% 8.4% 1.5% 7.6% Casava 7.5% 5.7% 13.1% 1.7% 1.0% 6.0% Other Tubers 0.7% 1.2% 0.5% 0.4% 0.1% 0.6% Legumes 4.1% 3.1% 4.9% 4.2% 2.5% 3.9% Fish 21.2% 14.3% 6.2% 6.0% 2.4% 10.5% Meat & Poultry 1.8% 1.3% 0.8% 2.8% 2.7% 1.9% Eggs & Dairy 0.5% 0.5% 0.0% 0.4% 0.3% 0.4% Oil Seeds & Nuts 2.3% 4.2% 6.4% 3.5% 3.0% 3.9% Fats 3.4% 3.3% 1.2% 3.6% 3.4% 3.0% Vegetables 8.3% 7.6% 9.5% 9.3% 9.7% 8.8% Spices & Condiments 3.7% 5.4% 6.5% 6.4% 8.6% 5.9% Fruits 0.4% 0.8% 0.2% 0.4% 0.3% 0.4% Sweets & Sugar 0.8% 0.7% 1.4% 0.2% 0.2% 0.7% Non-alcoholic Drinks 0.4% 0.2% 1.4% 0.4% 3.0% 0.9% Alcoholic Drinks & Stimulants 1.3% 5.9% 2.0% 6.4% 8.3% 4.5% Other foods 3.0% 3.6% 2.0% 1.9% 1.2% 2.4% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Poor Region Category Maritime Plateaux Centrale Kara Savanes Total Maize 26.1% 15.7% 7.8% 8.8% 7.1% 14.2% Millet, Fonio, Sorghum 0.8% 1.8% 8.9% 18.6% 33.4% 10.4% Rice 7.7% 11.0% 11.4% 8.9% 6.9% 9.4% Wheat 3.5% 2.1% 1.2% 1.5% 0.7% 1.9% Yam 2.0% 11.9% 11.2% 8.3% 0.2% 7.1% Casava 6.9% 11.8% 14.2% 4.3% 0.6% 8.3% Other Tubers 1.0% 0.9% 0.0% 0.3% 0.1% 0.5% Legumes 3.3% 3.8% 6.0% 7.1% 3.7% 4.7% Fish 17.3% 9.7% 6.4% 3.8% 2.9% 8.9% Meat & Poultry 4.0% 2.8% 3.2% 3.4% 4.2% 3.5% Eggs & Dairy 0.5% 0.0% 0.1% 0.1% 0.1% 0.2% Oil Seeds & Nuts 2.7% 3.4% 4.7% 4.1% 2.8% 3.5% Fats 3.3% 5.7% 2.2% 1.8% 1.7% 3.2% Vegetables 7.3% 6.7% 6.4% 8.3% 10.5% 7.6% Spices & Condiments 3.8% 4.5% 6.1% 6.3% 9.2% 5.7% Fruits 0.6% 0.8% 0.4% 2.4% 0.8% 0.9% Sweets & Sugar 0.9% 0.5% 1.7% 0.6% 0.7% 0.9% Non-alcoholic Drinks 0.9% 0.4% 0.7% 0.3% 2.4% 0.9% Alcoholic Drinks & Stimulants 4.2% 3.8% 3.1% 9.0% 7.0% 5.0% Other foods 3.0% 2.7% 4.1% 2.2% 4.8% 3.3% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% ANNEX 2 Page 4 of 9 Non-Poor Region Category Maritime Plateaux Centrale Kara Savanes Total Maize 18.7% 11.6% 7.4% 10.0% 7.0% 11.6 Millet, Fonio, Sorghum 0.5% 1.3% 5.3% 8.0% 23.4% 6.0% Rice 8.1% 9.6% 11.4% 10.2% 11.6% 9.9% Wheat 4.8% 3.7% 2.4% 3.4% 1.7% 3.4% Yam 3.9% 11.8% 12.1% 9.6% 2.6% 8.3% Casava 5.3% 6.4% 10.4% 2.7% 1.5% 5.6% Other Tubers 0.7% 1.2% 0.5% 0.9% 0.2% 0.8% Legumes 3.3% 3.1% 4.7% 4.4% 3.5% 3.8% Fish 16.2% 13.9% 5.3% 6.2% 3.8% 9.9% Meat&Poultry 8.1% 6.9% 10.1% 10.2% 8.3% 8.7% Eggs & Dairy 1.9% 1.0% 0.7% 1.3% 1.0% 1.2% Oil Seeds & Nuts 1.3% 2.6% 3.1% 2.7% 1.8% 2.3% Fats 3.4% 3.4% 2.7% 2.2% 2.3% 2.9% Vegetables 6.8% 6.7% 5.7% 6.5% 9.2% 6.8% Spices & Condiments 2.9% 3.9% 4.0% 4.0% 6.5% 4.0% Fruits 1.3% 1.1% 0.8% 1.5% 0.6% 1.1% Sweets & Sugar 1.2% 0.9% 1.6% 0.9% 0.6% 1.1% Non-alcoholic Drinks 2.7% 1.6% 2.3% 2.1% 2.4% 2.2% Alcoholic Drinks & Stimulants 6.8% 7.3% 7.0% 11.3% 8.9% 8.1% Other foods 2.1% 1.9% 2.6% 2.0% 3.1% 2.3% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Source: Staff calculations based on EBC 1987-1989 Table 2A.2: Percentages of Underweight Children, 1984-1992 1984 1985 1986 1987 1989 1990 1992 Maritime 46 38 34 35 33 35 20 Plateaux 31 30 28 28 27 28 14 Centrale 34 31 30 27 25 25 16 Kara 38 34 31 30 29 29 20 Savanes 46 44 43 41 48 48 31 TOGO 39 35 33 32 32 33 21 Source: CRS clinic data. ANNEX 2 Page 5 of 9 Table 4A.1: Employment in the Public Sector, 1986-90* (Number of employees) 1986 1987 1988 1989 1990 (Est) Civi] Servants 20,123 20,630 20,556 20,265 20,467 Category Al (1,590) (1,637) (1,653) (1,621) (1,694) Category A2 (2,113) (2,255) (2,270) (2,250) (2,323) Category B (55,491) (5,891) (6,072) (6,062) (6,406) Category C (7,419) (7,105) (7,101) (7,064) (6,711) Category D (3,510) (3,742) (3,460) (3,268) (3,333) Permanent government employees 8,723 8,329 7,698 7,519 7,168 Technical assistants 82 93 82 86 102 Guards 532 506 499 479 567 Contratual employees 32 27 30 31 25 Teachers in religious schools 2,535 2,528 2,543 2,478 2,523 Others 140 143 144 141 161 TOTAL 32,158 32,256 31,552 30,999 31,013 Source: Ministere du travail et de la fonction publique, direction de la gestion informatique, du personnel et de l'emploi. * Based on the civil service survey completed in December 1983 and regularly updated since then; data are for December 31 of each year, excluding military personnel. ANNEX 2 Page 6 of 9 Table 5A.1: Annual Average Food Balance, 1988-90 (tons) Bank Estimates Government Estimate Estimated Consumption (I) Available Output (2) Balance= (2)-(1) Balance Maize 221,385 246,257 24,872 36,048 Rice (Paddy) 60,148 23,106 -37,042 -23,191 Sorghum+Millet 143,700 169,387 25,687 18,633 Cassava 359,923 400,796 40,873 -6,263 Yams 283,347 254,709 -28,638 11,148 Cowpeas 26,961 15,119 -11,842 -1,233 Note. "Available Output" is net of post-harvest losses. Source: DESA, 1991 (Working Paper 7) Table 5A.2: Togo -- Factors Influencing Crop Output Growth Mean output Mean price Growth per change per Net revenue Net revenue Input year, year, per laborday, per hectare, Subsidy Return to 1980-1990 a 1980-1990 b 1990 C 1990 C level, 1990 d intensification e % % CFAF CFAF % % Cassava 2.4 0.4 868 108,500 11.4 121.2 Yam -1.8 1.6 1,148 315,563 1.8 89.3 Maize 8.0 -0.1 636 54,100 28.2 107.0 Sorghum 4.3 1.8 592 44,410 19.1 83.3 Millet 7.0 1.4 534 34,720 28.2 64.2 Rice 9.3 2.0 639 76.715 21.2 65.5 Beans 2.0 4.4 781 54,675 - 99.1 Groundnuts 1.6 0.3 419 35,604 12.3 123.3 Coffee 3.8 1.9 771 73,250 34.0 34.9 Cocoa -4.3 2.1 932 51,250 66.0 99.1 Cotton 29.7 6.2 750 90,00( 41.4 12.8 Source: Annex tables a Based on a comparison of average output in 1979-81 and 1989-91 b Food crops: average price on rural markets in 1979 compared to same for 1988-90. Coffee, cocoa, cotton: average producer price in 1979-81 compared to same for 1989-9 1. Note: Between 1980 and 1990, the Consumer Price Index increased at an average annual rate of 4.4% c Food crops: Based on prices in 1990, no input subsidies; Cash crops: based on prices in 1991-92 season, no input subsidies. d Improved manual technology, level of subsidy is Wo-W/Wo, where Wo=total input cost per ha, without subsidy and W=total input costs per ha, with subsidy (1989-90) e Refers to the percentage increase in the net return per hectare resulting from the shift from existing to improved manual technology, with elimination of post-harvest losses. ANNEX 2 Page 7 of 9 Table 5A.3: Target Yield Increases for Major Crops Yield (Kg/Ha) Togo Best West African Increase Needed Cotton 1,133 1,290 (Mali) 17% Cocoa 200 410 (Cote d'lvoire) 10% Coffee 300 -- Maize 1,000 1,350 (Ghana) 35% Sorghum 600 1,020 (Nigeria) 70% Millet 600 1,150 (Nigeria) 92% Rice 1,300 1,900 (Nigeria) 46% Cassava 7,600 10,950 (Nigeria) 44% Yam 9,800 11,156 (Benin) 14% Groundnuts 530 1,270 (Ghana) 140% Source: Annex tables * No West African country has a significantly higher mean yield. ANNEX 2 Page 8 of 9 Food Crop Price Variations FCFA per kg Maize 90 80 -X-Mar ~Plat - - - Ctr 70 K~~~~~~~~~~~~~~~~~ a 40 - 30 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 340 Fruit vegetables -X-Mar 290 /l\ / \ ~~~~~~~~~~~~~~~~~~~~~Plat - - - Ctr 240 - -4--- Kar ------ Sav 40, - + - + - -_t Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec ANNEX 2 Page 9 of 9 Table 6A.1: Financial Needs of the National Solidarity Agency, 1995-99 (thousands CFAF) Health Care Aid (health & nutrition) 336,000 SOS- Needy students 216,500 Nutrition (food stamps) 52,700 Capacity-building for self-reliance of low income families 927,000 Creation of income generating activities for unemployed youths 3,274,150 Assistance to the handicapped 215,764 TOTAL 5,022,114 Source: Agence de Solidarite Nationale, Programme de solidarite nationale, ao0t 1994. ANNEX 3 Page 1 of 1 Annex 3: Methodological Notes on Food Security Calculations Annual food balances have thus far been calculated on the basis of a 1973 study of food requirements that makes no allowances for differences between urban and rural consumption, and obviously cannot take into consideration changes in the dietary pattern during the past twenty years. Fortunately, the 1987-1989 EBC survey provides more up-to-date information on consumption patterns. Cereals are the most important staples providing almost 60 percent of the total energy consumption of the Togolese. Root crops (cassava and yams) provide 15 percent of the calories, while oils and fats provide 9 percent of the calories. In general, consumption patterns are in line with agricultural production, meaning that people consume locally-produced food crops and therefore local diets closely follow the ecological and agricultural characteristics of the country, resulting in significant regional differences. For example, the relatively high consumption of oils and fats in Maritime and Plateaux regions is correlated with the cultivation of oil palms. Urban consumption patterns are very similar to those of the surrounding region except that rice and wheat play a far more important role (they provide 15 percent of all calories in urban areas as opposed to only 5 percent in rural areas). The calculations of the food balance are based on two assumptions. First, it is assumed that the consumption pattern of 1987-1989 is still valid. Second, it is assumed that seeds and post-harvest losses add to about 15 percent for most crops, with the exception of yams (35 percent), and legumes (25 percent). This last assumption was also made in previous calculations of the Togo food balance (Agricultural Sector Review, 1993). Other sources assume larger losses for maize and yams, on the basis of FAO studies from the early 1980s (see Examen de l'Intervention Gouvernementale dans la Commercialisation des Produits Vivriers au Togo, avec Recommendations pour l"Action Future de l'Etat. janvier 1984, World Bank WA2DA). IMAGING - ~~~~~~~~Report No: 15526 TO Type: ER