WPS5509 Policy Research Working Paper 5509 Togo Growth Diagnostics Susanna Lundstrom Leonardo Garrido The World Bank Poverty Reduction and Economic Management Network Economic Policy and Debt Unit December 2010 Policy Research Working Paper 5509 Abstract This paper starts by analyzing past growth trends and location as a regional hub, but also make growth in sources of growth in Togo, as well as the structure of the Togo more inclusive. And as economic opportunities economic actors in the country. The second part explores arise for the private sector, there is a need to restructure the current binding constraints to rapid and sustainable the banking sector, which has already started, to smooth growth. For increased international competitiveness distortions in the credit market. Promising sectors and growth, Togo will need to raise efforts to streamline within agriculture that are vital to economic growth, its costly and cumbersome business procedures. However, employment opportunities, and poverty reduction for this to have a sizable impact, Togo must prove to remain important, but will need to overcome a number potential investors that political stability is permanent of coordination failures. Not least due to the history of and that corruption, poor budget execution and government interventions causing economic distortions, mismanagement of state owned enterprises belongs to the government must allow for a stronger role for private the past. As a result of the new government's reform operators and encourage it wherever possible. Finally, agenda and the return of international aid, a window although education does not exhibit constraints to of opportunities for high returns to the still limited economic activity in Togo today, it is of importance to public investments has opened up. This is especially improve the quality of education, not least to profit from true in infrastructure and connectivity services, which and catalyze the opportunities related to Togo's potential would not only take advantage of Togo's geographical as a regional hub. This paper is a product of the Economic Policy and Debt Unit, Poverty Reduction and Economic Management Network. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank. org. The authors may be contacted at slundstrom@worldbank.org, and lgarrido@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Togo Growth Diagnostics By Susanna Lundstrom and Leonardo Garrido PREM Economic Policy and Debt, The World Bank JEL codes: O10, O53 Key Words: Economic Growth, Growth analytics, Country Specific Analysis, Growth Diagnostics, Growth constraints, Competitiveness, Togo. Acknowledgement This report was prepared by a core team comprising of Susanna Lundstrom (team leader and principal author), Leonardo Garrido (coauthor), Juan Pedro Schmidt and Pooja Kacker. The report was prepared under the guidance of Philip English, Lead Economist for Togo. Valuable comments were provided by Olivier Cadot, Edgardo Favaro, Timothy Gable, Elena Ianchovichina, Elke Kreuzwieser, Marshall Mills, Lorraine Ocampo, Gallina Vincelette and the Togo country team. A shorter, revised version of this study was published as the first chapter in the 2010 Country Economic Memoranda "TOGO: Reviving the Traditional Sectors and Preparing for the Future: An ExportLed Growth Strategy", Report No 53711TG, The World Bank. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. 2 Table of Contents 1 Introduction .......................................................................................................................................... 4 2 Methodology ......................................................................................................................................... 6 3 . The Current Economic Framework in Short ....................................................................................... 10 4 Development Dynamics ...................................................................................................................... 11 4.1 Overall Growth Decomposition and Employment Dynamics ..................................................................... 1 1 4.2 Capital and Investment Dynamics .............................................................................................................. 2 1 4.3 Demand Distribution of Growth ................................................................................................................ 5 1 4.4 Economic Sector Dynamics ........................................................................................................................ 6 1 4.5 Exports and Economic Diversification ........................................................................................................ 9 1 4.6 Conclusions from and Implications of the Development Dynamic Analysis .............................................. 1 2 5 Profile of the Economic Actors ........................................................................................................... 22 5.1 Demographic Dynamics ............................................................................................................................. 2 2 5.2 Literacy and Education Level of the Working Age Population ................................................................... 3 2 5.3 Employment ............................................................................................................................................... 4 2 5.4 Employment, Education and the Incidence of Poverty .............................................................................. 5 2 5.5 Conclusions from and Implications of the Profile of the Economic Actors ................................................ 6 2 6 Analysis of Potential Constraints to Growth ....................................................................................... 27 6.1 Cost of and Access to Capital ..................................................................................................................... 8 2 6.2 Rate of Return on Capital ........................................................................................................................... 1 3 6.3 Geography and Regional Potential ............................................................................................................ 2 . 3 6.4 Infrastructure ............................................................................................................................................. 4 3 6.5 Human Capital ............................................................................................................................................ 8 3 6.6 Macroeconomic Stability ........................................................................................................................... 1 4 6.7 Governance, Corruption and Government Efficiency ................................................................................ 4 4 6.8 Barriers to and Costs of Doing Business ..................................................................................................... 6 4 6.9 Market Failures .......................................................................................................................................... 2 5 6.9.1 SelfDiscovery ......................................................................................................................................... 3 5 6.9.2 Coordination Failures ............................................................................................................................. 5 5 7 Conclusions ......................................................................................................................................... 56 8 References .......................................................................................................................................... 58 9 Appendices .......................................................................................................................................... 59 3 1 Introduction Togo is one of the poorest countries in the world. With a per capita GDP equal to 437 current US$ in 2008, the economy was ranked within the world's lowest income decile and in Purchasing Power Parity (PPP) terms it was the world's 10th poorest economy.1 Similarly and in the same year, Togo was ranked 159th among 179 countries appraised in 2008 UN's Human Development Index.2 Low aggregate per capita income level has translated into widespread poverty. In 2006, 62% of the population was poor, with poverty headcount rates in the north exceeding 90% in most rural areas (CWIQ, 2006). Furthermore, an additional 20% of the population technically considered nonpoor, was deemed as vulnerable, which raised the fraction of the population that may eventually fall into poverty to over 80% (PRSP, 2009). Figure 1: GDP per capita development of Togo and LICs 19902007. ( $ ) 400 Low Income Avg. 350 Togo 300 Real US$ of 2000 250 200 150 100 50 0 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 Source: Authors' calculations. DDP Data, World Bank. However, Togo has not always exhibited a grim social and economic performance. In fact, almost half a century ago, Togo began its postindependence economic journey with a fastpaced sustained growth, which permitted its percapita income to almost double in a span of two decades. An economic collapse after 1980 had wiped out most of the postindependence improvements in living standards by the early 1990s. If Togo had maintained its average growth rate from the 1960s (5.8% per year per capita), it would have achieved middlelowincome status today, with a current US$ per capita GDP comparable to that of Philippines, Honduras and Egypt in 2008.3 Instead of catching up, per capita income in Togo has diverged from the average of peer lowincome countries, Sub Saharan Africa and the rest of the world (See Figure 1 and Table A1, Appendix). 1 Among 192 countries with data reported for 2008, and out of 233 registered by World Bank. Cross country comparisons of income levels, regarded as proxies of individual welfare, are more accurate when presented in Purchasing Power Parity terms (PPP) than in current or real US$. In simple terms, PPP is the ratio of prices for the same quantity of a particular commodity in two countries, when the prices are expressed in each country's currency. In 2007, the PPP conversion factor was 224.6 XOF (Communaute Financiere Africaine Francs) per US$, implying that, to buy 1 US$ worth of the PPP basket in Togo, one would have needed 224.6 XOF. In turn, the official exchange rate in Togo was 479.3 XOF/US$. At this rate one would have purchased more of the selected PPP basket in Togo using XOF, than the amount of that basket purchased by 1 US$ in the United States. This is why a per capita income in Togo equal to 379 US$ in 2007 (239 real US$ of 2000) translates into a 808 US$ PPP per capita in 2007 (763 real US$ in PPP of 2005). 2 The Human Development Index (HDI) is a summary measure of human development that is published by the United Nations Development Programme (UNDP). The HDI provides an alternative to the common practice of evaluating a country's progress in development based on per capita Gross Domestic Product (GDP). HDI measures the average achievements in a country in 3 basic dimensions of human development: 1) "A long and healthy life" proxied by the average life expectancy at birth; 2) "Knowledge", proxied by adult literacy rate, combined with primary, secondary and tertiary enrollment ratio; and, 3) "A decent standard of living", proxied by GDP per capita at PPP in US$. 3 Income level classification is that of the World Bank. According to this, lowincome countries are those with GNI per capita (calculated from World Bank's "Atlas method") under US$ 935. Middle lowincome countries are those with GNI per capita ranging from 935$ up to 3,705 US$. 4 Togo's postindependence economic performance has mostly reflected the country's economic policies, with international and local shocks playing a secondary but not unimportant role in outcomes. The country registered a trend of decelerating growth since independence, with long periods of negative growth rates that were briefly interrupted by a period of small pickup in the mid 1990s after the 100% devaluation of the CFA franc. As can be seen in Figure 2, average real per capita income growth was 5.8% in the 1960s, a period characterized by "prudent and marketfriendly policies macroeconomic policies with efforts to modernize the (economic) administration and to adjust the economy from colonial to postindependence structures and a stable international environment" (Gogué and Evlo, 2004). In the 1970s, a decade that witnessed major shift in economic policy, growth fell to 0.1% as a commodity price boom resulted in increased government revenues, which, in turn, prompted a higher participation and intervention of the public sector. Specifically, higher public revenues were used to increase public investment, implement ill conceived projects, and expand the parastatal sector by creating publicly owned manufactures behind heavy import protection. During this period, government expansionary policies and a liquid international capital market led to a substantive increase in public debt. Togo was heavily affected by the onset of the oil shock and debt crisis in the early 1980s. During this period, and despite economic reforms sponsored by the international financial and development community, real per capita income shrank further at a pace of 2 percent per year. The 1990s was a period characterized of high political instability and continued stopandgo in the reform agenda, as the stabilization and adjustment program faced severe obstacles for implementation. Government effectiveness was severely compromised, not only by people's opposition to reforms, but also to increased and widespread corruption. During this period, per capita income recovered at a rate of 0.7% per year. A relatively high population growth rate in Togo (2.3% between 1970 and 2000) contributed to the observed low or negative per capita growth rates during the period. Another factor to be considered is the turbulent political situation that characterized the country since the early 1990s4 and resulted in the withdrawal of international aid until mid 2000s. The lack of aid together with the meager government revenues, resulted in a long period of very low public investment and hence deterioration of public goods. Figure 2: GDP and GDP per capita growth rates and Development of GDP per capita in Togo, 1961 to 2008 (LHS), and Per capita GDP growth rates, Togo and selected ECOWAS countries 20042008 (Real US dollars of 2000) (RHS). 12% 400 7% annual growth from Real US$ 2000 10% 350 6% 5% US$ of 2000 per capita 8% 300 4% 6% 250 3% 4% 200 2% 2% 150 1% 0% 0% 100 1% 2% 50 2% Nigeria LICs Mali GuineaBissau Burkina Faso Togo Niger MICs Cape Verde Gambia, The Cote d'Ivoire Benin 4% 0 Liberia Sierra Leone Ghana Guinea Senegal SSA 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 GDP per capita (constant 2000 US$) Tota real GDP growth (US$ of 2000). HP filter Per capita real GDP growth (US$ of 2000) HP filter Source: DDP data, World Bank. Togo's more recent real per capita income growth rate of 0.1%, 20042008, was the second lowest of the ECOWAS countries5 and much lower than the average growth rate of SubSaharan Africa (3.2%) as well as that of 4 After the oneparty system, people started their fight for democracy in the beginning of the 1990s. President Gnassingbe Eyadema died in February 2005 after 38 years in power and immediately replaced by his son Faure Gnassingbe. International pressure forced him to hold an election, but his victory was soon disputed by the opposition and some international observers. This triggered violence, hundreds of lives were claimed and tens of thousands fled to neighboring countries. In August 2006 an agreement, Accord Politique Global (APG) was signed and parliamentary elections were held in October 2007, starting a new period of reconciliation and focus on economic reforms. 5 The Economic Community of West African States (ECOWAS) is a regional group of 15 West African countries founded in 1975, with the main mission of promoting economic integration among its members. It includes: Benin, Burkina Faso, Cape Verde, Côte d'Ivoire, Gambia, Ghana, Guinea (suspended after 2008 coup d'etat), GuineaBissau, Liberia, Mali, Niger, Nigeria, Senegal, Sierra Leone and Togo. Out of this group, seven Francophone countries (Benin, Burkina Faso, Côte d'Ivoire, Mali, Niger, Senegal, 5 the LICs (4.3%) (Figure 2). IMF (2009b) predicts a GDP growth rate of 1.7% in 2009, which corresponds to a 0.8% fall in per capita terms.6 This means a continued divergence compared to other lowincome countries. Predictions for 2011, given commitments to planned reforms, are slightly more positive with a 4% GDP growth (IMF, 2009a), corresponding to a per capita growth rate of 1.5%. However, that is still a very low pace for a country that wishes to achieve higher income standards within in the span of one generation. In fact, for Togo to reach a middle income status by, say 2030, it would need to grow at per capita growth rate of 4.1%, which would correspond to a GDP growth of 6.6% (see further comparisons in Table A1 in Appendix I).7 In the 2000s Togo has had more success in terms of both political stability and economic reforms, even if a significant impact on growth is not yet seen. Increasing political liberties coupled with higher macroeconomic discipline8 have restored cooperation with the international financial community. As a consequence, total aid grew from 8.2 current USD per capita in 2001 to 18.4 in 2007, which is still low compared to the average aid per capita in LICs (37) and SSA (44) the same year. Moreover, at the end of 2008 Togo qualified for the HPIC debt relief and normalized relations with the international donor community. The Government of Togo has in 2008 achieved nearly all of its structural reform targets agreed with IMF. Missed growth targets are in part due to the global financial crisis and the negative impacts from food and energy price increases, but one needs to take into consideration the nature of the reform so far. Up till now the focus has been on getting some fundamentals correct, such as reinforcing public finances and reforming stateowned banks and enterprises. Next steps involve policy actions expected to have more direct effects on the private sector and growth. However, as will be further argued, this will depend on to what extent the government manages to target Togo's binding constraints to growth. The purpose of this study is to better understand the structural problems that remain in Togo and binding constraints for higher, sustainable growth. The growth analytics is based on a number of approaches and methods, covering the more traditional tools combined with a growth diagnostic approach à la Hausmann, Rodrik and Velasco (2005). A central question asked is why the levels of private investment and entrepreneurship are not higher. Possible answers include limited investment and entrepreneurship because of high financing costs or low returns to capital, defined in a broad sense. The latter will in turn point to problems either in the underlying potential of the country in terms of complementary factors (social returns), market failures or government failures such as lack of public investments or involvement in the market that causes distortion. In order to answer this question different types of data are needed, macro and micro, firm and household level, quantitative and qualitative, data will be thoroughly used. To search for the underlying questions as to what constraints to economic growth in Togo today are, involves first an understanding of the negative growth in the 1980s and its consequences today in terms of the investment environment of Togo. In addition, further understanding is reached by looking at what happens to economic actors that manage to circumvent the constraints that are still present. 2 Methodology The growth analytics in this study is based on a number of approaches and methods it will cover the more traditional tools combined with the HRV growth diagnostic approach (Hausmann, Rodrik and Velasco, 2005). There are three main steps involved in this growth analytics: and Togo) plus Guinea Bissau created the West African Economic and Monetary Union (WAEMU or UEMOA from its French name), an organization established to promote economic integration among countries that share a common currency, the CFA franc. 6 Assuming a constant population growth of 2.5 % per year. 7 Authors' calculations using the Atlas method and World Bank, DDP data. The income group limits are assumed constant at the values in 2007. 8 See for example the letter on intention with IMF in April 2009. 6 1) The Development Dynamics Analysis, including an understanding of major factors that explain the country's past growth trends and trendbreaks, population dynamics and migration trends, overall productivity and employment dynamics in the country, sector dynamics, the stage of economic transformation and diversification, major challenges and opportunities faced, etc. (2) The Profile of the Economic Actors, including a description of income earning activities of self or wage employed, distinguished by sector, geographical area (e.g. rural, urban), etc, and other relevant characteristics such as health and education (3) The Growth Diagnostics. The HRV approach relies on a growth diagnostic tree that is used to organize the thinking about binding constraints to private sector growth in the short to medium run (Figure 3). The first question posed is whether the level of investment and entrepreneurship in a country is low or constrained.9,10 In order to answer this question, from both macro and micro perspectives, firm and household level, quantitative and qualitative, data are needed. Figure 3: Growth Diagnostics Tree Constrained investments and entrrepreneurship Returns to economic Cost of finance activity International Private Local sources Social Returns sources Appropriability of finance of finance Government Financial sector Market failures Domestic savings failures intermediation Macro risks Micro risks Source: Adapted from Hausmann, et al. (2005). Investment and entrepreneurship may be low due to low returns on capital or the high cost of financing. Returns to capital may be low due to low `social returns' because of insufficient investment in complementary factors of production such as infrastructure and human capital, low land productivity linked to poor natural resource management, or just unfavorable geographical conditions. Further possibilities possibly include low `private 9 Note that the focus is in particular how to raise private investments and entrepreneurship. However, this does not imply that the analysis do not cover public investments. On the contrary, the analysis is to a large extent about what types of public investments that are needed in order to release constraints to private investments, i.e. the analysis is about advising the government on reforms or public investments that will boost growth in their particular country. Moreover, we are assuming that even returns to public investments will increase as the binding constraints are addressed. 10 Note also that we assume that investment and entrepreneurship is indirectly affecting total factor productivity (TFP). Hence, even in cases where TFP has been the main problem for growth, factor accumulation may be the answer. Also, low TFP can in itself be the reason for low factor accumulation (i.e. low returns to investments) and the root cause of the productivity problem may still be best understood by looking at the causes of low private investments. 7 returns' to capital as a result of `government failures' (such as taxes, poor property rights, corruption, laborcapital conflicts, macro instability, etc) and `market failures' (such as coordination externalities and learning externalities affecting negatively the country's ability to adopt new technologies). The cost of finance may be high because the country has limited access to external capital markets or problems in the domestic financial market. Difficulties in accessing external capital markets arise for diverse reasons such as high country risk, unattractive FDI conditions, vulnerabilities in the debt maturity structure, and excessive regulations of the capital account. Bad local finance may be due to low domestic saving and/or poor domestic financial intermediation. An exercise in growth diagnostics consists of reviewing and analyzing the factors found along the branches of the growth diagnostic tree in order to ascertain which of these factors are the binding constraints to growth.11 Although all factors may matter for growth and welfare, those that are binding are likely to provide the largest positive direct effect when released, so that even after taking into account secondbest interactions, the net impact of a policy change is positive and sizable. While this methodology seems simple, in practice it is challenging to identify indicators of scarcity that make possible a tradeoff between the potential constraints to growth. When a constraint is binding, one should be able to observe a number of phenomena: (i) the (shadow) price should be high, (ii) movements in the constraint should produce significant movements in growth, (iii) agents in the economy should be attempting to overcome or bypass the constraint, (iv) agents less intensive in that constraint should be more likely to perform well, and vice versa.12 Hence, for each potential constraint it is not enough to just conclude that the levels are low but it is also necessary to observe signs of scarcity, i.e. signs of a current need for the constraint to be released in the particular economic environment of the country in question. Benchmarking the performance of an economy is one way of identifying factors that hold investment and economic activity in a specific country back. First of all, it is important to select an appropriate group of comparative countries. The core group of comparators should consist of countries that share similar characteristics with the country of interest, and are going through similar development states (see Box 1). More so this set of countries should also include good performers in the region (as well as regional and income groups' averages) to see in what dimensions they differ from the studied country. Though the group of comparators may, in fact, vary depending on the question one needs to address.13 While useful for identifying problem areas, benchmarking for growth diagnostics has limitations and should be used with caution and jointly with other methods to shed light on constraints to growth. For example, comparisons are meaningful only if the indicators are constructed following procedures that are consistent across countries, and a change in the set of comparator countries for an indicator may change the conclusions from a benchmarking exercise. Moreover, evidence that a country ranks low on a given indicator does not imply that the area is a `binding' constraint to growth ­ it is only an indication that this might be the case. There might be worse constraints that are the binding ones, or the structure of the economy in the specific country of study is simply not sensitive to that potential constraint. Another popular source of information is enterprise survey data, but this type of data must also be interpreted with caution. Besides general information about the firm (including size, ownership, location, sector of economic activity, etc.) these surveys record perceptions, indicators of performance and objective business indicators, in areas such as infrastructure and services, access to land, crime, finance, business government relations, labor regulations and environment, etc. So called Investment Climate Assessment (ICA) surveys are extracted mostly from private business enterprises and aim to understand, precisely, what are the main constraints to economic activity. However, there are a number of caveats such as: 11 Note that we do not assume we will find one binding constraint but rather a bundle of constraints that all need to be addressed to trigger growth. However, the goal is not to present a laundry list but a few prioritized areas of concern. 12 See Hausmann, et al. (2005), Hausmann, Klinger and Wagner (2008), Rodrik (2004a and 2004b). 13 For example, when one would like to assess whether the cost of capital is high in a country, the group of comparators must include countries with similar credit worthiness ratings. When addressing questions about logistics and transport costs, the group of comparators must include landlocked countries with similar types of traded goods. 8 1) The sample of existing firms included in ICAs is a biased sample excluding the firms that would have existed if the relevant constraints were removed.14 2) Comparing hedonic scales across individuals and countries is rifled with cultural issues, as well as the particular economic structure in a country15. 3) The business enterprise sample may not be representative of the span of business enterprises across the country16. Thus, a growth diagnostics exercise has to be undertaken to confirm or reject the reliability and interpretation of these types of surveys. Box 1: Is Togo in a Poverty Trap? When a country is poor, everything seems to be binding and in addition, having a very low level of income may in itself constitute an impediment to growth. Economic literature refers to these cases as poverty traps, a situation where growth cannot be ignited in a country just because productive inputs are so scarce that sustained growth cannot kick in. The HRV approach does not explicitly deal with poverty trapped countries. The framework assumes well behaved aggregate production function, which are not amenable to explain a country's stagnation at very low level of income. Extending the analytical approach to include potential poverty traps is a challenging task, but maybe a necessary one for gaining understanding on constraint to growth in countries at very low levels of income ­ such as in Togo. The identification and possible growth implications of being in a poverty trap or another particular growth state could be analyzed by approaching growth analytics as suggested in Pritchett (2008). The key question in all growth analytic processes is: "What are the feasible actions in the country's current circumstances that would initiate (or sustain) an episode of sustained, broadbased, rapid growth?" Hence, a necessary first step, prior to a growth diagnostic exercise, is the "Identification of the country's current growth state based on its past and 17 recent performance". Pritchett (2008) defines 6 possible (exhaustive) growth states , which are used to generate empirical transition probabilities from past to current states or from current to desired states. These probabilities are shown to depend on the set of policy actions leading to the next state, but also the initial state. If one were to use Pritchett's criteria, Togo could be classified as a country that has in the course of the last three decades transitioned from a phase of fastgrowing preindependence state, to economic decline post1980, and from there to a state of stagnation in the 2000s. Now, being stagnant at a low level of income is not the same as being in a poverty trap. The difference is subtle, yet, important. Being in a poverty trap implies that a country may not be able to endogenously foster and sustain a moderate to high growth rate even if it properly and systematically address main, binding constraints to growth with the available resources and adequate political support. This is so because the low quality, scarce productive factors and the very low level of technology associated with mostly traditional, low scale, subsistence activities; coupled with extremely weak local demand (due to low income) and foreign demand (due to low competitiveness) make business unable to consistently growth, even when good policies are in place. Some exogenous intervention is needed to help the country jump towards a higher growth path. The question whether Togo is in a poverty trap or simply stagnant at low level of income is not an easy one, but there exist a few indications that Togo is not poverty trapped. First, per capita income in Togo in 2008 remains 30% below its postindependence historical maximum, but still 34% above its preindependence level when a period of strong growth started. Second, Togo is better equipped in terms of endowment such as human capital per capita which has steadily increased in the last half a century. Physical capital has been added at a much slower pace than the growth of employment, resulting in a physical capital stock per worker at the same low level as at independence. However, not lower than the amount available when country's growth began to accelerate in the 1960s. Also, it is perceived that the conditions that were present when postindependence growth ignited in the 1960s are nowadays still in place. This growth diagnostics exercise will hence identify constraints that prevent a shift from current stagnation towards a desired state of rapid, converging growth in Togo. This is not to say that all countries in a given state will have to follow similar recipes (as the extrapolation of lessons is precisely what GD is not about) but, as Pritchett explains "the set of diagnostics and therapeutics recommended are likely to be more similar across countries starting from similar states". 14 As Hausmann et al (2008) explain: "The surviving firms may be politically well connected; form part of financial conglomerates; or be exempt from taxation because of some specific provision, making them less sensitive to problems of governance, finance or taxation, respectively". 15 For cultural reasons, what is outrageous corruption in Sweden may not raise too many eyebrows elsewhere (Hausmann et al. 2008). Also, a specific growth determinant such as education may be more important in the US than in Zambia due to their different economic structures. 16 For instance, Togo's business enterprise survey was restricted to firms in Lomé. Also, the agricultural sector or the informal sector is normally not surveyed, posing problems for lower income countries that rely mostly upon primary activities. 17 The six possible growth states according to Pritchett (2008) are: (i) poverty traps, (ii) a status of economic decline or collapse, (iii) stagnation (or non converging growth), (iv) modest converging growth, (v) rapid converging growth and (vi) growth leaders. 9 3 The Current Economic Framework in Short Box 2: Togo's Economy in Short Together with other countries from the West African Economic Monetary Union (WAEMU) Togo maintains a fixed exchange rate regime (at a rate of , a hard peg to the Euro, with own legal tender (the CFA Franc, at a rate CFAfr 655.957 : 1). The country's real exchange rate has followed a slightly appreciating trend ever since 1994. In that year, the nominal exchange rate depreciated 100% against the euro, a factor that significantly affected prices (62% of consumer inflation between the end of 1993 and 1995) and to an increase in poverty. Terms of trade have been volatile, linked to international prices of the country's main export commodities, and those of imported food and energy (the country imports 100% of its petroleum needs). Terms of trade have slightly deteriorated in the last decade. The country sustains current account deficits that are financed by inflows through the capital and financial account, in particular project grants and concessional loans. As IMF (2009) explains, Togo's current account deficits are in line with the norm for a lowincome African country that receives aid inflows. Nonetheless, the country's ability to sustain the projected current deficits (5 6 percent of GDP) over the mediumterm depends on the expected increase in aid inflows (consistent with donor pledges) or other, nondebtcreating inflows such as FDI. Monetary policy is determined by the regional central bank, the Central Bank of West African States (BCEAO, by its French name), which prioritizes the control of inflation and the maintenance of the peg to the euro, which is guaranteed by the French Treasury. Monetary policy is therefore influenced by that of the European Central Bank (ECB). The BCEAO's discount rate broadly reflects movements in the ECB's key interbank rate, although BCEAO rates are typically higher than ECB rates and take account of the prevailing circumstances of member states, such as inflation and bank liquidity. The country's fiscal stance has improved in recent years, especially since the introduction of an IMF Staff Monitored Program (SMP) from October 2006 to June 2007 and a subsequent threeyear poverty reduction and growth facility (PRGF), in the second trimester of 2008. Donor funding (especially from EU and the World Bank) has increased after full IMF reengagements in the country, following free and fair parliamentary elections in 2007. In fact, Togo began clearing external debt arrears under the SMP, paving the way for fresh funding, and reached a decision point under the heavily indebted poor countries (HIPC) debt initiative at the end of 2008. Togo is essentially a rural, agricultural economy: Half of the output, two thirds of workers and over 80% of exports are linked to the primary sector. Most food crops are produced by subsistence farmers who operate on small family farms and produce items such as corn, manioc, sorghum, yams, cotton and coffee. Leading cash crops are coffee and cocoa, followed by cotton, palm kernels, copra and peanuts. Over 40% of land in the country is arable. There exists a state marketing board for Cotton (Société togolaise de cottonSotoco) that is under a restructuring process under the PRGF. th The services sector employs about 1/5 of the workforce, with key subsectors including commerce, public services and transport and communications. The port of the capital, Lomé, is a key asset and plays an important part in regional trade, especially after conflict in neighboring countries caused a shift of international trade through Togo. The banking sector suffered from the dismal economic performance in the 1980s and 1990s, as they had heavily lent to parastatal enterprises. The subsector comprises nowadays just 1% of GDP. Together with agriculture, trade and transport, secondary sector activities in phosphate mining, agroprocessing, cement and clinker constitute the main economic activities in Togo. In 2007, The government returned the mines to full state control with the formation of the Société nouvelle des phosphates de Togo (SNPT). Industrial production was expected to rise with the return in donor support and 18 improvements in electricity supply , but these prospects were halted by the spike in energy prices and the international financial and economic crisis of 20082009. The 1989 Free Zone (FZ) law gave companies the advantages of a tax holiday, dutyfree imports of materials for production, a less restrictive labor code, and the ability to hold foreigncurrency accounts. By 2005 there were about 60 firms operating in the FZ, employing 9,000 workers. Changes in FZ's fiscal regime were introduced in 2009 that attempt to improve tax collection and a new FZ law is being finalized. Togo's current economic policy framework has as the overarching goals of contributing to poverty reduction and modernizing the economy through promotion of increased participation of the private sector and productive diversification. For this purpose, its government has maintained already for over a decade an earnest process of market oriented, progressively liberalizing economic reforms that attempt to encourage both local and foreign investment. To this end, the current economic policy performance of Togo was commended by the IMF in 2009, which is especially impressive when considering current demanding local and international circumstances. Fiscal and monetary policies have been prudent in the face of exogenous shocks and slower than expected growth in 2008 and 2009. Low recent growth rates have even prompted authorities to accelerate key reforms to promote 18 It is expected that a new, 100mw power station in Lomé, which will roughly double domestic generating capacity will be completed in early 2010. 10 growth through the medium term. As the IMF (2009) explains, the main current policy targets in Togo include keeping macro stability; effectively tackling structural bottlenecks in the financial and export sectors and utilities; and, improving budget execution, strengthening sustainability, and supporting growth. Box 2 gives a short overview of the state of the economy as presented in recent assessments. Important risks to the success of the economic policy have also been discussed in the literature: A potential drain of international capital flows resulting from 2009's international financial and economic crisis, which could further depress growth and revenues; delaying private sector involvement in restructuring stateowned banks and enterprises; and, political uncertainty linked to a runup presidential election scheduled for 2010. 4 Development Dynamics An important first step in understanding Togo's historic economic performance is the analysis of underlying, proximate determinants of growth. Observing the dynamic behavior of productive factors, sectors of economic activity and aggregate demand can provide ideas about sources of growth or stagnation in the country. This first layer of knowledge should be followed by further investigation of particular sectors, activities or groups of individuals that are thought to be fundamental drivers of business.19 In addition, several other areas are worth exploring such as the role of the nominal versus real effective exchange rate and that of terms of trade when the country is exposed to commercial shocks, the demographic dynamics when the country is, experiencing an important demographic transition, etc. 4.1 Overall Growth Decomposition and Employment Dynamics Figure 4: Growth Accounting, Togo, 19812007, and Growth Decomposition by Period, and Total Factor Productivity (TFP) Index, Togo and Benchmark Countries, 1960=1 to 2003 (RHS). China 3 2.5 2 India 1.5 Malaysia 1 Cote D'Ivoire Ghana 0.5 Togo 0 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 GDP Physical Human Per capita Period Labor TFP Growth Capital Capital GDP growth 1980s 1.2% 0.3% 0.4% 2.4% -1.9% -2.0% 1990s 2.2% -0.1% 0.3% 2.1% -0.1% -0.5% 2000-2007 2.0% 0.4% 0.5% 1.8% -0.7% -0.8% Note: Using a Growth Accounting decomposition that assumes a constant return to scale Cobb Douglas production function for Togo we produce an index for TFP, which we compare with a similar index calculated by Bosworth and Collins (2000) for a set of countries Source: Authors' calculations. DDP data, World Bank. A growth accounting decomposition for Togo shows the negative or lack thereof ­ growth, explained by physical capital accumulation, human capital accumulation and TFP, concluding that any substantial positive 19 One may, for example, be interested in looking at the dynamics of the exportable sector, and within it of certain activities that constitute an important source of value added, foreign exchange, innovation, employment and/or fiscal revenues; or, in the behavior of the touristic, financial, information technology, transport and communication sector or other services, when non tradable sector are dominant in the economy; or in the role of remittances for growth when a sizeable fraction of the labor force move within or outside the country searching for better living standards; or in primary activities, be it in the non renewable, agricultural, or other commodities. 11 effect on GDP growth is more or less exclusively explained by growth in labor. Figure 4 suggests that less output has been generated with additional units of input, that is, the growth of the residual TFP has been negative. In fact, Figure 4 shows a loss of almost two thirds in cumulative productivity since the end of the 1960s. Togo's cumulative loss of productivity, as approximated by this residual, is the highest compared to a sample of benchmark countries (Bosworth and Collins, 2000). In what follows, we take a closer look at each of the proxy determinants of growth, namely, physical capital, labor, TFP and in later sections human capital. 4.2 Capital and Investment Dynamics As Figure 5 shows, for the period 19812005, the decline in output per worker has been largely explained by the fall in the capitallabor ratio, and to a lesser extent by a shift of resources across sectors. The former ratio has declined due to the significant increase in labor and the very low increase in capital (a total of only 9.5%, 1981 to 2005). The capital stock per unit of worker currently stands at the same level as in the 1960's. It increased steadily in the 1960s and 1970s at a pace much faster than employment growth but was followed by a substantial deceleration in real investments. Figure 5: Decomposition of Changes in Output per worker, Togo 19812005 (LHS), and Capital Stock Values per Unit of Worker, Togo, 19602007 (RHS). 800 T h o u sa n d s o f C F A o f 2 0 0 0 700 600 500 400 51.64 300 200 100 30.47 0 1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 55.00 45.00 35.00 25.00 15.00 5.00 5.00 Output per worker net of intersectoral shifts Capital Labor Ratio TFP residual net of intersectoral shifts Note: Capital stocks have been estimated following the perpetual inventory method, whereby the capital stocks values in a year "t" is obtained as the sum of capital stocks in year "t1" minus depreciation, plus the investment (gross capital formation) in "t". In turn, the initial value of capital stocks is estimated under a steady state assumption on the growth rate of capital, as the value of initial gross capital formation divided by the sum of the depreciation rate and the growth rate of investment. Source: Authors' calculations based on WDI data, World Bank. Figure 6: Capital Formation in Togo and Benchmark Countries, averages 20002005 (LHS) and Real Gross Capital Formation per capita (in logs) versus Real per capita GDP growth, Averages 20002008 (RHS). 45% 400 20% Real GDP per capita (US$ of 2000) growth 40% 350 35% 300 30% 15% Per capita 250 % of GDP 25% 200 rate, 2000 2008 20% 15% 150 10% 10% 100 5% 50 5% 0% 0 Senegal Ghana Sub S. Africa Togo Benin LowerMiddle Inc. Cape Verde Cote d'Ivoire Mali Low Income GuineaBissau Gambia, The Burkina Faso 0% Togo 5% 0 2 4 6 8 10 Real Gross Capital Formation per capita (US$ of 2000) Real Gross Capital Formation (Real US$ of 2000) per capita, Ratio of Gross Capital Formation to GDP (Real US$ of 2000) 20002008, in logs Source: WDI data, World Bank. In terms of real gross capital formation per capita, Togo has the 11th lowest ratio out of 152 countries with data in the world. The average for the period 2000 to 2005 was 46 USD per capita, compared to 99 USD per capita on average in SSA, and as much as 355 USD per capita in the lower part of the middle income country group (Figure 6, 12 left side). Figure 6, on the right side, plots the investment ratio to total population versus real per capita GDP growth for the period 20002008. Interestingly, one can observe an almost total lack of correlation between the investment ratio to population and per capita GDP growth, i.e. a higher level of investment per capita is not associated with the growth rate of income.20 On the other hand, investment growth and GDP growth are strongly correlated.21 Most of the investment in Togo during the 2000s has been undertaken by the private sector, as the government has not been able to obtain resources neither from abroad (ODA) nor from domestic sources as a result of a low fiscal revenue base. Grants (current transfers to government) averaged just over 1% of GDP between 2000 and 2008. More recently, as Togo has been able to reengage with the international financial community, official aid has increased to 3.9% of GDP in 2009 and an expected 4.5% in 2010. Fiscal revenues (excluding grants) from central government remained under 17% of GDP during the same period, but are at least expected to remain at the same level during these two years. As a consequence, the public contribution to capital formation has and is expected to increase compared to historical levels (Figure 7). Figure 7: Public and Private Gross Capital Formation and National Savings, Togo, 20012010. 18% 16% 14% 12% % of GDP 10% 8% 6% 4% 2% 0% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 GFCF, Private GFCF, Public Gross National Savings Source: International Financial Statistics, IMF. Togo has a substantially lower than expected gross national saving rate given its level of per capita income (Figure 8). This fact translates into a persistent current account deficit (Figure 9) that needs to be financed by external sources. 20 This appears to contradict HRV's steady state fundamental equation for the accumulation of income and capital r 1 , with "g" representing the growth rate of real per capita GDP; "cdot" ratio to "c" is c k g c k the growth rate of consumption per capita; "kdot" ratio to "k" is the growth rate of capital per capita; and, , r,(1), and are model parameters for the opportunity cost of funds, expected social rate of return of investments, the proportion of private returns that are privately appropriable, and a measure of productivity, respectively. On closer reflection, one can observe that such relation between the growth rate of per capita income and the investment ratio holds only in steady state (if such state is ever reached), and once one has accounted for the effect of model parameters, such as productivity, the interest rate, complementary factors of production, tax rates, and many others. One has also to consider potentially different levels of capacity utilization, technologies and employment to population ratios across countries, so the lack of correlation between investment and growth does not really come as a surprise. 21 Real GDP growth vs Real Gross Fixed Capital Formation Growth (USD of 2000), using averages 20002005, gives y = 1.461x ­ 2 0.005 with R = 0.424. 13 Figure 8: Gross National Savings as a share of GDP, Averages 20002007 (LHS), and Correlation between Gross Domestic Savings and Gross Capital Formation (RHS). p g 50% 45% 40% y = 0.0345x 0.0592 R² = 0.0946 35% 30% Percent 25% Ghana 20% Cape Verde 15% Cote D'Ivoire 10% Benin 5% Togo 0% 4 5 6 7 8 9 10 11 12 LN(GDPpc Current US$) Note: Correlation data is from the period 19902007, but varies depending on data availability for each country or group. For Togo the correlation is for 19902005. Source: Authors' calculations, DDP data, World Bank. Deficits in the trade balance and in the service account (interest on debt) are compensated by inflows from current transfers, FDIs, remittances and other sources of financing (Figure 9). Togo has resorted to rescheduling arrears and debt relief to cope with a persistently negative balance of payment, but there is still a remaining BOP gap that needs to be solved in a more sustainable manner. FDIs have increased but not enough to close the gap (Figure 10).22 Remittances have also increased substantially since 2000 and Togo has now one of the largest shares (9.2% of GDP in 2008) of remittances in Africa (Figure 11). Figure 9: Gross Capital Formation, Gross Savings and Current Account Balance, 20012010 (LHS), and Sources of Foreign Financing in Togo as a percentage of GDP, 20012010 (RHS). 30% 20% 15% 20% 10% 10% % of GDP 5% 0% 0% 10% 5% 20% 10% 30% 15% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Trade Balance Net Current transfers Net Income from abroad Gross Fixed Capital Formation Gross National Savings Current account balance FDI & other Cap. Fin. Acc. Financing BOP (Reserves & other) Source: WDI data, World Bank. Figure 10: Foreign Direct Investments, Net Inflows to Togo, 19742006. 100 15 Millions 80 10 60 40 5 20 0 0 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 20 5 40 60 10 Foreign direct investment, net (BoP, current US$) Foreign direct investment, net inflows (% of GDP) Source: DDP Data, World Bank. 22 Note that this relatively meager increase in FDI is a sign in itself that credit but rather expected returns to investments seem to be the constraint, as FDI originate from sources that are not credit constraint. 14 Figure 11: Workers' remittances and compensation of employees, received, Togo, 19742007 (LHS), and Workers' remittances and compensation of employees received, Togo and Benchmarks, 2007 (RHS) (% of GDP). 12 High income Middle income 10 Low income Cape Verde Togo 8 Sierra Leone Liberia 6 Senegal GuineaBissau Gambia, The 4 Nigeria Benin Mali 2 Niger Cote d'Ivoire 0 Ghana Burkina Faso 0 1 2 3 4 5 6 7 8 9 10 Source: DDP Data, World Bank. 4.3 Demand Distribution of Growth An initial postindependence decade of prudent macroeconomic policies in the 1960s was followed by a period of expansionary fiscal policies, debt growth and a questionable public investment program in the 1970s. Government expenditures and (mainly public) capital formation increased in the period at rates close to 10% per year (see Table 1), but did not result in substantial growth as most of purchased goods were of foreign origin. The sudden oil price shock in the early 1980s put an end to an era of easy access to international capital markets at a time when Togo's external debt had already become unsustainable. The 1980s were a time period of contracted government expenditures and public and private investments, with private consumption providing most of the contribution to growth in aggregate demand. The structural adjustment program included measures to reduce the budget deficit mainly through a control of government expenditure and a gradual disengagement of the government from some sectors of activity. However, it contributed to a recession as the reduced government expenditure was not compensated by increased private activity. Capital formation fell in the 1990s while government expenditures remained mostly stagnant, as the ongoing economic liberalization program and reforms were, at best, only partially applied. Policy inconsistencies were a result of social unrest and economic mismanagement, and a wider perception of corrupt public officials led to increasing dysfunction of the public authority. During this same time period, exports also remained stagnant, represented mainly by traditional agriculture commodities and phosphates. Private consumption was, again in the 1990s, the main driver of aggregate demand. Government expenditures, investment and exports have slightly recovered in the 2000s, following a period of increased macroeconomic and political stability and continued liberalization and reforms. A significant portion of the registered growth of aggregate demand in the 2000s can be attributed to the exportable sector (more on this on the following section). However, as a significant proportion of supply has been provided by imported goods, real GDP growth has remained low, and even negative in per capita terms. 15 Table 1: Growth Rates of Aggregated Demand/Supply, by Demand Components, by decade from 1970s. Growth Rates of Aggregate Demand / Supply by Demand Components. By Decades. Imports of Aggregate Exports of Private Government Gross Capital Period Real GDP Goods and Supply / Goods and Consumption Consumption Formation Services Demand Services 1970s 2.8% 10.9% 5.8% -1.1% 9.8% 11.9% 8.5% 1980s 1.1% 1.0% 1.0% 4.9% -2.3% -0.2% -1.0% 1990s 2.3% -1.2% 0.9% 4.1% -0.5% -4.8% 0.3% 2000s 1.8% 4.0% 2.6% 1.2% 3.9% 3.4% 5.6% Source: Staff Calculations based on World Bank, WDI Shares in Real Aggregate Supply of GDP and Demand Components Imports of Aggregate Exports of Private Government Gross Capital Period Real GDP Goods and Supply / Goods and Consumption Consumption Formation Services Demand Services 1970s 64.3% 35.7% 100.0% 40.3% 9.9% 13.4% 25.4% 1980s 58.5% 41.5% 100.0% 45.8% 8.9% 9.9% 26.4% 1990s 65.2% 34.8% 100.0% 60.4% 7.8% 7.3% 22.8% 2000s 65.0% 35.0% 100.0% 60.5% 7.2% 7.7% 24.8% Source: Staff Calculations based on World Bank, WDI Contribution to Growth in Aggregate Demand / Supply by Demand Components. By Decades. Imports of Aggregate Exports of Private Government Gross Capital Period Real GDP Goods and Supply / Goods and Consumption Consumption Formation Services Demand Services 1970s 1.8% 3.9% 5.8% -0.4% 1.0% 1.6% 2.2% 1980s 0.6% 0.4% 1.0% 2.2% -0.2% 0.0% -0.3% 1990s 1.5% -0.4% 0.9% 2.5% 0.0% -0.4% 0.1% 2000s 1.2% 1.4% 2.6% 0.7% 0.3% 0.3% 1.4% Note: Contributions of each variable to aggregate demand growth are computed multiplying period growth rates by period average share of the variable in total aggregate demand. Source: Authors' calculation, WDI data, World Bank. 4.4 Economic Sector Dynamics Togo's steep economic decline and subsequent stagnation since 1980 has reversed initial trends in structural transformation and has created a domestic economy increasingly dependent on primary activities, mainly at the expense of its tertiary sector (Figure 12 and Table 2). In fact, primary activities have doubled their share in value added GDP between 1980 and 2009 and represent almost 50% of value added in this latest year. Services have gone the other way, from 51% of GDP in 1980 down to 25% in 2009, while industry has maintained a relatively constant share of 25% of GDP during the same period. On one hand, value added from primary activities has increased at a decreasing pace in the last three decades, boosted mainly by subsistence activities and the animal sector. On the other, traditional coffee and cotton sectors follow declining long term trends of 2.1% and 2.0% respectively since 1990. Among traditional primary sectors only cacao exhibits a slightly increasing trend during the period. Within the secondary sector, traditional extractive industries also show a downward trend. Since 1990, value added generated by the phosphates sector has declined at an average rate of 6.2% per year. This has been compensated by the significant increase in clinker (lime and cement) production, which has performed well at rates over 9% since 1998. Similarly, most manufacturing activities have been on an increasing trend between 1990 and 2008. Among them, the production of nonmineral metals and metallurgic products have risen each at 12.4% and 11.2% per year during the period, elevating their share in manufacturing production from 12.4% in the former year to over 37% in the latter. Other good manufacturing subsectors are chemicals, beverages, and, to a lesser degree, textiles, electricity gas and water. Value added in the tertiary sector has fallen at an average pace of 0.7% per year since 1990, mainly driven by a debacle in the banking sector at a pace of 11.9% per year and in other commercial services (5.2%). In fact, Banking and insurance sector has declined from an already low 5.9% of tertiary GDP (and 2.6% of Value Added GDP) to an incredibly low 0.2% in 2008. The commercial sector has stagnated and, reduced its contribution to total GDP. On the flipside, transport and communication activities have performed moderately well in the last two decades growing at a 2% rate per year. 16 Figure 12: Sector shares in total value added in Togo (percent), 1965 to 2009 (LHS) and Trends in Growth Rates in Value Added GDP, Togo, 19662009 (RHS). factor Costs. 1965 2009 100% 15% 80% 10% % Chg. 60% 5% 40% 20% 0% 0% -5% 1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 Primary Secondary Tertiary Primary, HP filter Secondary, HP filter Tertiary, HP filter Notes: All calculations use data in constant LCUs. There are small differences in Gross value added at factor cost in the table and the DDP data due to incomplete information on some activities. Source: Authors' calculations based on data from the World Bank, DDP data, and Government of Togo. Table 2: Shares, Growth rates and Contribution to Growth, Togo, 1960s2000s. ,g g , y , y 1960s 1970s 1980s 1990s 2000s Shares in Value Added GDP Primary 32.2 26.2 29.4 44.31 46.50 Secondary 18.9 21.3 21.5 21.24 24.90 Tertiary 48.8 52.5 49.1 34.45 28.60 100.0 100.0 100.0 100.0 100.0 Growth in VA GDP Primary 6.2% 2.8% 4.5% 3.1% 2.5% Secondary 11.2% 7.8% -0.2% 2.6% 3.6% Tertiary 7.3% 3.6% -0.4% -0.8% -2.0% Total VA GDP 7.7% 4.3% 1.1% 1.5% 1.4% Contribution to Growth Primary 2.0% 0.7% 1.3% 1.4% 1.2% Secondary 2.1% 1.7% 0.0% 0.6% 0.9% Tertiary 3.6% 1.9% -0.2% -0.3% -0.6% Total VA GDP 7.7% 4.3% 1.1% 1.5% 1.4% Notes: All calculations use data in constant LCUs. Source: Authors' calculations based on data from the World Bank, DDP and Government of Togo. The main conclusion from this discussion is that any incremental growth in the Togo economy in the last three decades, has been achieved out of growth of the country's labor force. Unable to find productive employment in tertiary activities, or employment opportunities in general in industry, Togolese citizens have found employment in subsistence agriculture or have used their resource endowments in other primary activities. Subsistence activities and the animal subsector generated over 85% of primary GDP and 40% of overall value added in the 2000s (compared to 29% in 1990). However, the ability of several manufacturing activities to thrive must also be highlighted; the cement sector making up for decreased activities in extractive industries from phosphates, and the transportation and communication sectors ability to maintain a modest growth rate. An alternative way to observe sector dynamics and their contribution to GDP is the so called Shapley decomposition, which highlights changes in average productivity, employment or labor reallocation by sector (World Bank, 2007b)23. It is confirmed that agriculture and, to a lesser extent, services, mostly absorbed the main part of the increased labor force between 1981 and 2005, combined with an increased (+17%) but still low labor productivity in agriculture and significantly lowered productivity (64%) in services (Figure 13). Labor productivity was highest within the industry sector, growing by 6% during the period, Even though growth in employment was high for the industry sector itself, the absolute number of job opportunities created was small. In 1981, agriculture 23 Historic employment data in Togo is quit scarce. In the case of value added GDP, this decomposition necessitates of at least two data points for each sector value added variable and for the employment levels, as well as two data points for indicators of population, working age population, labor force and employment. As one is interested in long term trends, it is recommended for data points to be at least 10 years apart, and that they correspond to normal (nonshock) years. Fortunately, employment data by sector in Togo is available for 1981 and 2005 by combining information from the World Bank DDP data base and the CWIQ 2006 survey, and real Value Added data by main economic sectors from Government of Togo. 17 employed more than 65% of the workforce, followed by services, which employed 24% and then industry at just over 9%. In 2006, overall employment had increased significantly but the relative shares of employment in agriculture and industry declined (to 60% and 6% respectively), while the share of workers in the service sector increased to 33%. Figure 13: Employment by Sectors (LHS) and Output per Worker by Sectors (RHS). 500,000 450,000 400,000 1981 350,000 Thousands CFA of 1978 2005 300,000 250,000 200,000 150,000 100,000 50,000 0 Agriculture Industry Services Source: DDP Data, World Bank, and 2006 Core Welfare Indicators. The overall decrease in GDP per capita between 1981 and 2005 was largely due to the drop in service productivity. Combining the employment dynamics and productivity dynamics by sector described above gives us a better picture of the total impact on output per worker (see Table 3). The positive contribution to per capita output of the increased employment share in services was completely offset by the huge decline in output per worker in the same sector. Moreover, even though there was a positive contribution from increases in output per worker within the industry sector, these gains were dulled by a corresponding drop in employment share. The same pattern can be observed in the agricultural sector, where increased labor productivity was offset by a decline in the share of the labor force. The positive effect from intersectoral shifts came from a movement of workers mainly from the agriculture sector to the service sector, which had higher labor productivity than the agricultural sector, and compensated for the decline in the employment share in industry. Table 3: Decomposition of Changes in Per Capita GDP in components (Percent points Per Year). 19812005. Contributions from changes in: Output per Level Sectoral Total worker Employment composition of employment Change in Per Capita GDP: -1.01 Sectoral Contribution -1.5 0.0 0.3 -1.20 -Primary 0.3 -0.3 0.1 0.15 -Secondary 0.1 -0.1 -0.2 -0.14 -Tertiary -1.9 0.4 0.3 -1.21 Participation rate -0.05 Pop1564/Population 0.24 Note: Changes in per capita output are decomposed into (i) changes in output per worker assuming a counterfactual with no changes in employment and its sectoral composition; (ii) changes in the employment level assuming a counterfactual with no changes in output per worker and the sectoral composition of employment; and (iii) changes in the sectoral composition of employment holding everything else equal in a counterfactual. As an example in case (iii) one can consider an decrease in the share of employment in sectors with higher than the average product of labor. Then everything else equal the per capita GDP declines by 1.0%, as labor shifts away from higher productive sector to lower productive sectors. Source: Staff calculations, DDP Data, World Bank, and 2006 Core Welfare Indicators. 18 4.5 Exports and Economic Diversification This section presents an overview of the export structure and behavior in recent years, since it adds important information on Togo's stage of economic transformation and growth prospects. Export growth of 5.6% per year explains almost half of the increase in aggregate supply between 2000 and 2007 (see Table 1)24. The ability of Togo to sustain the higher growth rates observed between 2000 and2007 therefore hinges on its capacity to sustain a growing exporting sector. Growth of exports has been accompanied by diversification of the export basket, as shown by the decrease in the HerfindhalHirschmann index of export diversification (Figure 14)25. Togo is actually the second most diversified country within the WAEMU, something that is owed to a substitution of traditional products such as cotton, coffee and phosphates in favor of an array of new emerging products (Table 4). Figure 14: NonOil Herfindahl Index for Togo, 19802006 (LHS), and Compared to WAEMU Countries, 2006 (RHS). 0.4 0.9 0.8 Scale 01 (Lower = Higher Diversification) 0.35 Scale 01 (1=no diverisifcation) 0.7 0.3 0.6 0.25 0.5 0.4 0.2 0.3 0.15 0.2 0.1 0.1 0 0.05 0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Source: Export diversification data, PRMED, World Bank. Table 4: Export Value of Selected Products (% of total exports), Togo, 19802006. product code 1980 1985 1990 1995 2000 2005 2006 71 Coffee and coffee substitutes 11.0% 14.3% 6.0% 12.0% 5.6% 72 Cocoa 13.9% 7.7% 12.0% 22.0% 223 Oils seeds and oleaginous fruit, wh 2.4% 263 Cotton 12.6% 25.0% 26.7% 22.7% 10.5% 7.2% 271 Fertilizers,crude 46.7% 51.2% 52.1% 28.9% 18.1% 10.1% 14.1% 333 Petrol.oils,crude,& c.o.obtain.from 9.8% 334 Petroleum products,refined 18.5% 5.0% 2.5% 18.1% 13.6% 15.3% 661 Lime,cement,and fabricated construc 8.5% 20.3% 12.2% 667 Pearls,precious& semiprec.stones,u 2.3% 8.2% Grand Total 92.4% 90.8% 88.0% 85.6% 72.9% 66.5% 70.7% Source: Export diversification data, PRMED, World Bank. The shift of the export basket away from classic to new, emerging products has been accompanied by an increase of the income value of such basket in Togo, as defined by EXPY26 an indicator of export sophistication of a country (Hausmann and Klinger, 2006). Higher EXPY, authors hypothesize, allows countries to move more easily in the "product space" towards their "technology frontier". Togo is one of the WAEMU countries that has shown one of the faster increases in the income value of the exportable basket (Figure 15). 24 Notwithstanding, imports of goods and services have also increased, at a pace of 4.0% per year. 25 Explained in Appendix 2. 26 Explained in Appendix 2. 19 Figure 15: Trends in EXPY for Togo and WAEMU countries (LHS), and other benchmark countries (RHS), 19802006,. p 7,000 12000 Benin Burkina Faso 10000 Malaysia 6,000 Cote D'Ivoire China Real US$ of 2000 per capita Cape Verde 8000 Real US$ of 2000 5,000 South Africa Ghana India The Gambia 6000 4,000 GuineaBissau Mauritius Liberia 4000 3,000 Mali TOGO Niger 2000 2,000 Nigeria Senegal 0 1,000 Sierra Leone 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 Togo Source: Export diversification data, PRMED, World Bank. Given the country's income level, Togo's EXPY remains below expectations. As reported by Hausmann et al (2005) there exist an empirical relation between EXPY and a country's income level that is explained not only by construction of the indicator. Once a set of controls as rationalized by the authors have been accounted for, the income value of Togo's export basket remain below that predicted by the country's per capita GDP (Figure 16). Figure 16: Residual of EXPY regression after controlling for independent variables suggested by Hausmann et al (2005). Measure of Unobserved Factors Affecting EXPY (Residuals of OLS Regressions of EXPY on Controls Suggested by Hausmann et al.). 2006 100% Observed EXPY - Predicted EXPY as % of 50% Togo 0% Observed EXPY -50% -100% -150% Note: A negative value indicates that a country's EXPY is below that suggested by its income level once one factors in the values for that country of selected explanatory variables. Source: Authors' calculations. It is however important to note that a significant part of Togo's imported goods are directly reexported to Nigeria as well as neighboring landlocked countries, to the extent that aggregated export and export diversification indicators may not reflect actual economic activity. During the 2000s, according to IMF (2007) about 16% of exports (11% of imported goods) were reexported, mainly to neighboring West African countries. It is however very difficult to estimate any numbers and thereby effects of this type of transit trade. Also, it is important to differentiate between the reexportation to the landlocked countries, which is a type of activity that Togo will continue engaging in and where the Lomé Port is part of the value chain for these products. However, the reexportation to Nigeria is a consequence of unfavorable trade policies in Nigeria, which means that this re exportation is not based on Togo's comparative advantage and may be disrupted as policies in Nigeria changes. Also, the reexportation to Nigeria takes many forms and is therefore less measurable than the other types of re exportation. Finally, over the long term, Togo has experienced deterioration on its term of trade due to stagnant price trends in the country's main exportable products from the 1970s through the early 2000s; although import prices have consistently increased during the same period (Figure 17). However, during the 2000s both prices of traditional commodities and of main imports (including food and oil products) have shown sharp increases leading to an overall stagnation in the terms of trade. 20 Figure 17: Export prices, Import prices and Terms of Trade in Togo (Base 2000=100), 19602005. 300 250 200 150 100 50 0 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Exports price index (goods and services, 2000 = 100) Imports price index (goods and services, 2000 = 100) Terms of trade (goods and services, 2000 = 100) Source: WDI data, World Bank. 4.6 Conclusions from and Implications of the Development Dynamic Analysis Togo's immediate post independence period (19601973) of prudent, modernizingoriented and market friendly economic policies were followed by a decade of interventionist, protectionist and expansionary policies that ended with the international oil shock and external debt crisis in the early 1980s. Economic reforms under an IMFsponsored financial stabilization program in the 1980s did not result in increased and sustained growth, mainly because of the burden of debt, economic mismanagement, and confrontations between population and the government, which ended up in removing the political elite in the early 1990s. Social unrest, public authority dysfunction, economic mismanagement and perception of corruption of public officials remained until late 1990s. The 2000s has constituted a decade of market oriented economic reforms that have been commended by international financial institutions (IMF, 2009). Prudent fiscal and monetary policies have been set in place. The structural reforms have stressed fiscal governance, financial sector restructuring and public enterprise reform, all with an emphasis on measures that more directly impact economic growth. Yet, reforms have not yielded the expected growth. Consecutive food, energy price increases and the recent international economic and financial crises have been identified as at least partially responsible for this lack of growth. There has been very limited capital investments in Togo since the 1980s and total factor productivity has declined rather than improved. Any positive effect on growth can almost exclusively be explained by growth in labor, explaining the negative growth per capita numbers. A potentially positive effect on growth per capita from an improved dependency ratio was negated by a decline in the labor force participation rate. In terms of sector dynamics since 1980, the employment share has increased for services, and decreased for agriculture and industry, while at the same time, the value added share has decreased for services, increased for agriculture and stayed approximately the same for industry. When looked at in its entirety, this indicates that little progress in economic transformation has actually taken place. For instance, the increased labor force either stayed in low productivity agriculture or moved to lowproductivity self employment in the service sector, resulting in a drastic reduction of average productivity in that sector. Export growth explains almost half of the increase in aggregate supply since 2000 and the ability of Togo to sustain these recent higher growth rates mostly hinges on its capacity to maintain this sector. Togo's exports have shown an increasing trend during the last two decades, though not as large as that of its peers from SubSaharan Africa and lowincome countries. In contrast, Togo is the most diversified exports and ranks highly in terms of the growth rate of the export basket's income value. Public capital investments are expected to rise as Togo reengages with the international community. The remaining question is whether the government will direct these investments in a way that removes Togo's binding constraints to growth ­ not least through continued progress within the export sector. 21 5 Profile of the Economic Actors The second step in the analysis involves forming a picture of the economic actors in terms of their resources and income generating activities. This is important as it not only tells us about the potential embedded in the production factors labor and human capital, but also about the feasible growth trajectories that would include a major part of the work force without jeopardizing productivity growth. Togo is a mostly rural country with 37% of the population residing in urban areas, of which over 60% live in Lomé.27 As will be discussed below, socioeconomic conditions, employment and living standards differ widely between urban and rural areas with Lomé standing out even compared to other urban areas in the country. For instance the headcount ratio of poverty is 24.5% in Lomé but varies between 28 and 79% in the urban areas in the other regions. Poverty is even higher in rural areas, with Savanes having an incidence of 90%. However, Togo has still a relatively equal income distribution with a GINI in 2006 of 34% compared to the 44% average for SSA (see also Figure 18). Given the very low average income level, this means ­ as we have seen that a major part of the population is poor. It also means that there is a limited share of the population with enough capital to make major, risky investments even though opportunities may be there. Figure 18: Income shares of the highest and lowest 10%, West African Countries (20012007). 45 40 35 30 25 20 15 10 5 0 Liberia Liberia Senegal Senegal Nigeria Nigeria Sierra Leone Sierra Leone Ghana Ghana Togo Togo Benin Cape Verde Cape Verde Mali Cote d'Ivoire Cote d'Ivoire Mali Niger Niger GuineaBissau GuineaBissau Gambia, The Gambia, The Burkina Faso Burkina Faso Note: The first (second) group of bars is for the highest (lowest) 10%. Data points were from different years for different countries but all within the period 20012007. Togo data is from 2006. Source: DDP data, World Bank. 5.1 Demographic Dynamics Togo has a high fertility rate. However, 4.8 births per woman is not higher than expected in a country with the same per capita income level (see Figure 19). 27 Most of the analysis in this section is based on the Core Welfare Indicators Questionnaire (QUIBB/CWIQ) 2006. The poverty line applied is FCFA 164842, using the FAO/ WHO equivalence scales (see World Bank (2007a)). 22 Figure 19: Fertility Rate (births per woman) vs log of GDP per Capita (constant 2000 USD), 2006 Source: DDP data, World Bank. The potentially large positive effect on growth per capita from the decreased dependency ratio was dampened by a decreased participation rate as well as an increase in unemployment. The share of working age population has increased since 1992, resulting in a declining dependency ratio, but the labor force per working age population has declined due to a decreasing participation rate (Figure 20). Overall the labor force registers positive growth but, as Figure 20 also shows, although employment has grown in every year, unemployment growth has typically been even higher. Figure 20: Dependency ratio and Participation Rates, Togo 19982006 (LHS) and Employment and Unemployment Growth, and GDP growth, Togo 19982006 (RHS). g p p y 20.0 1.00 15.0 0.95 10.0 0.90 5.0 0.85 0.0 0.80 5.0 0.75 10.0 0.70 15.0 0.65 20.0 25.0 Employment growth (%) Unemployment growth Labor Force/Working age Dependency ratio Source: Authors' calculations. DDP data, World Bank. 5.2 Literacy and Education Level of the Working Age Population Overall literacy rate is in line with other subSaharan African countries, but the variation in terms of both literacy rate and education level between urban and rural areas, and across regions, is substantial. The literacy rate for the age group 1545 in urban areas is 84% against 52% in rural areas, and 26% in rural Savanes against 88% in urban Lomé (Figure 21). The education level of urban workers is substantially higher than the one of rural workers. 45% of urban workers have finished junior high or a higher education level, over 40% have finished primary school and only 15% have not even finished primary education. Conversely, half of the rural workers have not finished primary school and 37% just finished primary school. The graph also highlights the substantial variation among the regions for both rural and urban areas, while these distributions in turn are similar across regions. Moreover, the situation with regard to literacy is improving. Literacy rate of younger age groups are substantially higher than the overall literacy, especially in rural areas and most children between 6 and 11 attend school in both rural and urban areas. 23 Figure 21: Literacy rates and Education Level in rural and urban areas, for age group 1545, by regions, Togo, 2006. 120% 120% 100% 100% 80% 80% senior high and above 60% 60% junior high 40% primary 40% none 20% literacy 20% 0% 0% maritime plateaux centrale kara savanes Total lomé maritime plateaux centrale kara savanes Total Source: Core Welfare Indicators Questionnaire (QUIBB/CWIQ) 2006 and author's calculations. 5.3 Employment As in other developing countries, most of the workers are employed in agriculture as selfemployed or unpaid family workers. Table 5 presents the sector and status of employment in terms of wage and self employment in Togo, and two things stand out from the table. First, as we have seen, agriculture employs almost 60% of workers, followed by Commerce and trade, 22%. Second, the high share of self employment and family helpers (91%) points to a high incidence of lowskilled and lowproductivity informal employment for those who cannot find other types of employment. Only in education and health and public administration it is more common to be wage employed, and the share is relatively high also for the mining, construction and transport sector. However, the share of these sectors in total employment is very low. As a result, only around 24% of urban workers receive a salary or are paid on a piece rate basis (27% in Lomé). The reminder are apprentices, nonpaid family workers or selfemployed. Table 5: Wage and Self Employment (LHS, and) Wage and Self Employment and Share of Laor Force (RHS), by Sector, Togo, 2006. Shares across sectors Total Self Wage or Self Total Share on Wage or piece employed across total employed/ piecerate Total rate / family sectors workforce family earner earner helper helper Agriculture 0.9% 58.0% 58.9% Agriculture 5.8% 94.3% 100.0% 10.6% Mining 0.2% 0.4% 0.6% Mining 63.4% 36.6% 100.0% 0.5% Manufacturing 0.5% 3.2% 3.7% Manufacturing 22.0% 77.9% 100.0% 7.1% Construction 0.6% 1.2% 1.7% Construction 37.1% 62.9% 100.0% 4.2% Transport 1.1% 1.5% 2.6% Transport 49.0% 51.1% 100.0% 6.9% Commerce and Trade 1.3% 20.2% 21.5% Commerce and Trade 8.3% 91.7% 100.0% 44.5% Education/ Health 2.0% 0.6% 2.6% Education/ Health 83.5% 16.5% 100.0% 5.5% Administration 0.8% 0.1% 0.9% Administration 97.1% 2.9% 100.0% 2.4% Financial Services 1.7% 4.4% 6.1% Financial Services 34.4% 65.6% 100.0% 14.2% Other Services 0.3% 1.1% 1.4% Other Services 22.7% 77.2% 100.0% 4.1% Total 9.4% 90.6% 100.0% Total 23.9% 76.2% 100.0% 100.0% Source: CWIQ 2006 and author's calculations. Again as in other poor countries, workers tend to be underemployed rather than unemployed. As Table 6 shows, overall unemployment rates are 7.4 and 1.4% in urban and rural Togo, respectively. At the same time 19% of rural and 25% of urban workers, are underemployed defined as those workers who declared to have tried to increase their incomes in the last 7 days. 24 Table 6: Unemployment and Underemployment by Region, Togo, 2006. Unemployed Underemployed Urban Rural Urban Rural Lomé 8.7% 19.7% Maritime 5.9% 2.5% 28.7% 30.0% Plateaux 6.1% 1.0% 12.2% 31.1% Centrale 4.5% 1.2% 8.0% 10.7% Kara 5.0% 1.4% 18.4% 10.6% Savanes 4.9% 0.2% 19.3% 30.0% Total 7.4% 1.4% 18.4% 25.0% Source: CWIQ 2006 and author's calculations. With regard to the sectors of employment, the main difference between Lomé and other urban areas is in the role that agriculture plays for employment (Figure 22). Less than 1% of workers are employed in agriculture in Lomé, with the same number being 10% and higher in the other urban areas. Otherwise the distribution of urban workers among sectors is similar across regions. Trade and commerce usually employs most of urban workers, followed by services, agriculture, manufacturing and transport. Figure 22: Employment by Sector, Urban areas (LHS) and Rural areas (RHS) in Different Regions, Togo, 2006. 100% 100% 90% Mining 90% Other Services 80% Administration 80% administ 70% Other Services 70% mines / 60% Construction 60% construc 50% 40% Education 50% transpor 30% Transport 40% educatio 20% Manufacturing 30% fabricat 10% 20% Agriculture services 0% 10% Services commerce 0% Commerce and Trade agricult maritime plateaux centrale kara savanes Total Source: CWIQ 2006 and author's calculations. Most of the rural workforce is employed in agriculture with the share varying from 56% in Maritime to 92% in Savanes (Figure 22). Other sectors with important shares in some regions include commerce, financial services and manufacturing. 5.4 Employment, Education and the Incidence of Poverty The incidence of poverty is ­ not surprisingly higher for workers with a lower education level in both rural and urban areas (Table 7). Similarly, there are certain sectors that have higher incidences of poverty. In urban areas the ones standing out are Agriculture, and Agriculture and Transport in rural areas. Table 8 presents the distribution of poor by education levels and sectors and shows that among the urban (rural) sectors, Commerce (Agriculture) has the highest number of poor workers. 38% of urban poor work in Commerce, of which more than half has not finished primary school. Commerce and Agriculture together concentrate 69% of the poor in the urban areas. Agriculture together with Commerce, concentrate 91% of the rural poor, with low skilled agricultural workers employing the highest share of poor people in rural areas (54%). 41% of the poor in urban areas have no education and in rural areas 65%. 25 Table 7: Incidence of Poverty by Sector and Education, in Rural and Urban areas, Togo, 2006. Poverty Incidence urban workers Poverty Incidence rural workers none primary junior highr high and a Total none primary junior highr high and a Total agricult 68% 66% 59% 42% 64% 76% 69% 65% 60% 73% mines / 100% 15% 8% 0% 16% 67% 55% 64% 38% 61% fabricat 37% 17% 6% 0% 16% 65% 55% 62% 33% 60% construc 33% 12% 13% 9% 14% 60% 41% 26% 41% 45% transpor 31% 10% 7% 0% 8% 42% 44% 54% 75% 46% commerce 28% 16% 8% 2% 18% 61% 43% 37% 30% 54% educatio 41% 27% 20% 6% 14% 61% 62% 36% 27% 37% administ 0% 36% 15% 6% 14% 100% 47% 51% 41% 48% services 41% 22% 6% 4% 17% 67% 57% 43% 15% 56% autres s 21% 6% 23% 5% 11% 55% 37% 32% 0% 41% Total 37% 21% 13% 6% 21% 73% 63% 56% 39% 68% Note: Using poverty line at 164842 and FAO/WHO equivalence scales. Source: CWIQ 2006 and author's calculations. Table 8: Distribution of the Poor by Sector and Educational Level, Rural and Urban Areas, Togo, 2006. Urban Rural none primary junior high & above Total none primary junior high & above Total Agriculture 15% 11% 6% 32% 54% 21% 5% 80% Mining 0% 0% 0% 0% 0% 0% 5% 1% Manufacturing 2% 3% 1% 5% 1% 1% 5% 2% Construction 0% 1% 1% 3% 0% 0% 5% 1% Transport 1% 2% 1% 3% 0% 0% 5% 1% Commerce 20% 14% 4% 37% 8% 2% 5% 11% Education 0% 1% 3% 4% 0% 0% 5% 1% Administration 0% 1% 1% 2% 0% 0% 5% 0% Services 3% 6% 2% 11% 1% 1% 5% 2% Other Services 1% 1% 1% 2% 0% 0% 5% 0% Total 41% 40% 19% 100% 65% 27% 5% 100% Note: Using poverty line at 164842 and FAO/WHO equivalence scales. Source: CWIQ 2006 and author's calculations. 5.5 Conclusions from and Implications of the Profile of the Economic Actors The importance of looking at the business environment both from the perspective of small firms as 91% of the population is selfemployed and large firms with the purpose of increasing scarce wageemployment opportunities. Trade and commerce employs most of urban workers, followed by Services, Agriculture, Manufacturing and Transport. Agriculture employs most of rural workers followed by Commerce. Poor in the urban areas are concentrated in Commerce (about one third) and Agriculture (about one third), and in rural areas in Agriculture (about two thirds). Hence, constraints for the agriculture and commerce/trade sectors need to be given attention, at least as important sectors in the medium run. At the same time, the concentration of the poor in these sectors point to a questioning of these sectors as the main, longrun vehicle out of poverty. 41% of the poor in urban areas have no education and in rural areas 65%. Still, to what extent education is the constraint to growth and a way out of poverty in Togo at this point needs to be further analyzed (see Section 5.4). 26 6 Analysis of Potential Constraints to Growth Recent enterprise survey data28 for Togo offers an important source of information about the main business constraints faced by firms in Lomé, as declared by interviewed firm representatives29. However, as discussed in Section 2, perception data has a number of caveats and limitations so the reliability and interpretation of each of these will be discussed below, when combined with other types of data. Figure 23 suggests that Political Instability, Access to Finance, Corruption, and Informal practices of competitors, are the main constraints to businesses in Togo. Other constraints often mentioned are Trade and customs, Electricity and Tax rates and administration. Interestingly, potential problems such as Crime/Theft/Disorder, Court, Labor regulations and Labor Skills are more or less never mentioned as major or severe constraint. Even Access to land, Business licensing and permits, and Transport are mentioned by a surprisingly low number of firms, but this result will be further discussed below. When breaking down the ICA result by size of firms (Figure 24) we see that micro and small firms see political instability as the main constraint, while medium firms mostly complain about access to finance and large firms about political instability. Figure 23: Main Constraints to Business (ICA), Togo, 2009. Percent of Firms perceiving a category as the main constraint. 30% 25% 20% 15% 10% 5% 0% Source: ICA data, World Bank. Figure 24: Main Constraints to Business (ICA), by Size of Firms, Togo 2009. Percent of Firms perceiving a category as the main constraint. 40% 40% 40% 35% Micro and Small Firms 35% Medium Firms 35% 30% Large Firms 30% 30% 25% 25% 25% 20% 20% 20% 15% 15% 15% 10% 10% 10% 5% 5% 5% 0% 0% 0% Source: ICA data, World Bank. 28 www.enterprisesurveys.org Enterprise survey data is available for over 100,000 firms in 117 countries. The surveys include indicators that benchmark the quality of the business environment across countries. They are commonly referred to as "Investment Climate Assessment" (ICA) indicators. Togo's business enterprise survey was concentrated in the capital city, Lomé, and includes 300 enterprises, of which 145 are micro firms (04 workers); 99 are small (519 workers); 41 are medium (2099 workers); and 15 are large (over 100 workers). These firms are also broken down into 19 firms within the Free Zone (FZ) and 281 nonFZ. Considering the survey structure, the 300 firms in sample represent over 1,700 firms in Togo. 29 Enterprise surveys always include this type of question, providing a list of 15 candidates from the topics mentioned above. The interviewed can provide up to 3 answers: The top, second and third most serious obstacle to business. Note that we in this study report the number of firms expressing a category as the most severe constraint, while the Investment Climate Assessment of Togo 2010 report both the firms that saw it as a main and a severe constraint. 27 6.1 Cost of and Access to Capital If the cost or access to capital is constraining growth, we should observe signals of high demand for financing relative to supply, and, possibly, low domestic credit to private sector. To determine this one should look at (a) Price signals: Trends in real interest rates, and their levels, compared to benchmark countries, (b) Rationing: The proportion of firms seeking for loans that are rejected, or of those that are discouraged from applying (c) Coping mechanisms: Whether individuals are trying to get around the constraint, by, say, obtaining loans from informal lenders (d) High returns for firms or individuals who were able to obtain financing, and/or survival of individuals that do not depend on credit. Furthermore, to determine if supply is low one may look at (a) domestic savings, (b) financial intermediation and (c) access to international financing (loans, aid, etc). It is important to notice that low private credit is not in itself evidence of high cost or access of capital as the reason might be constraints in the returns to investment, resulting in low demand for credit. Concluding that there is a low supply of financing must be done in a relative sense to the revealed demand as those would go hand in hand in a functioning market. Credit to the private sector as a share of GDP is very low in Togo, but not lower than expected given Togo's GDP level (Figure 25). Also, when looking at the relatively low lending interest rates combined with the low levels of credit to the private sector, Togo does not appear to be a liquidity constrained country due to a high cost of credit. Figure 25: Domestic Credit to Private Sector (percent of GDP) vs log of GDP per Capita (constant USD), 2006 (LHS), and Domestic Credit to Private Sector (% of GDP) vs Real Interest Rate, 20002008 (RHS). 250.0 Mean Real Lending Rate = 14.5% Domestic Credit to private Sector / GDP 200.0 150.0 100.0 Mean Dom Cred to Priv / GDP = 50.5% 50.0 Togo 0.0 0 10 20 30 40 50 60 70 Real Lending Rates (% points) Source: DDP data, World Bank, and Lending rate information from Banks in Togo On the other hand, the ICA data has access to finance as the top constraint to business in Togo. 25% of medium firms and over 30% of large firms had ­ by a small margin over political instability concerns access to finance as a major or very severe constraint to business. We further constructed an index of "financial repression" that summarizes a set of objective constraints to access to credit from ICA data. It turns out that over 50% have suffered from some form of financial repression, a number that is high but not much higher than other ECOWAS countries (Figure 26). Rather, a lack of demand is revealed when noticing that of the firms in Togo that do not have a loan, as many as 77% never applied30. From those firms that did apply (23% of total) about 38% were rejected (Figure 27) a problem that appears to be higher for medium and large firms. Hence, ICA based indicators signal the existence of financial problems that somewhat are not embedded in interest rates, which makes us shift our attention to the area of access to finance. 30 Out of them, two thirds did not apply due to insufficient funds (1%), the value of required collateral (30%), or discouragement from high rates, bad maturity or other financial requests from the banks (36%). 28 Figure 26: Percentage of Firms that Think Finance is a Major Constraint to Business (LHS) and Index of Financial Repression (RHS), Togo and Benchmarks. 80% 0.7 70% 0-1 (1=highest repression) 0.6 60% 0.5 50% 0.4 40% 30% 0.3 20% 0.2 10% 0.1 0% 0 Ghana Mozambique Botswana South Africa Togo Mauritius The Gambia Senegal Note: The Financial Repression Index, which is constructed on a scale 01 (higher index means higher repression) represents the fraction of firms that do not have a loan either because they applied for it but got rejected or because they were discouraged from applying (this, in turn, as a result to high cost of credit, insufficient funds offered, bad maturity of the loans, or the amount of required collateral). Source: ICA data, World Bank. Togo data from 2009. Figure 27: Demand for Loans and Reason for Declined Application, by size of firm, Togo, 2009. All firms 100% Rejected [African Avg] (Small/Medium/Large) 37.7% Private Comm. Bank 47.1% (80.0%/14.5%/5.4%) (50.2%/25.7%/14.4%) (34.4%/57.1%/100%) Applied 22.8% (19.9%/32.5%/40%) State Bank Not rejected 27.5% With a loan / credit Without a loan / credit 62.3% (28.1%/35.7%/0%) line line (49.8%/74.3%/85.6%) 18.5% 81.5% From? (14.5%/35.0%/33.3%) (85.5%/65.0%/66.7%) Non Bank Financial Inst 23.5% Did not apply Collateral required? (34.3%/7.1%/0.0%) 77.2% Yes = 81.6% (90.6%/66.6%/60.0%) (80.1%/67.5%/60.0%) Other source Value of Collateral? [African Avg] 1.9% Africa = 126.1% (124.9%/127.1%/137.6%) Why? (3.1%/0.0%/0%) Insufficient loan / High collateral / Discouraged No need a loan maturity interest rates 34.9% 34.9% 0.5% 29.7% (39.4%/12.0%/11.1%) (28.6%/60.0%/89.0%) (0.6%/0.0%/0.0%) (31.4%/28.0%/0.0%) Source: ICA data, World Bank. Financial intermediation issues appear to be a matter of concern in fact, IMF (2009a) highlights banking sector vulnerabilities as a key short term challenge for the economy. Several private banks are severely undercapitalized, primarily reflecting nonperforming loans to the troubled cotton and phosphate sectors. In fact, the share of nonperforming loans is the highest in West Africa (Figure 28). Furthermore, Doing Business 2010 data ranks Togo 150th among 185 countries on counts of public registry and legal rights related to the acquisition of credit31. 31 The index is constructed out of (i) Credit information (depth of credit information index, public registry coverage, private bureau coverage), and (ii) Legal Rights (the degree to which collateral and bankruptcy laws protect the rights of borrowers and lenders). 29 Figure 28: Nonperforming Loans, 2008. ... and the financial system suffers from the highest share of NPLs due to ailing public banks. Nonperforming loans in Togo and WAEMU, 2008 30 25 Share of total loans 20 Share of GDP 15 10 5 0 ni n so ire au al i ge r g al go Be Fa vo ss M Ni ne To a d 'I Bi Se r kin te ne a Bu Cô ui G Source: IMF (2009a). At the same time, there are measures rejecting financial intermediation as a major problem in Togo and the banking sector reform is already a priority for the government and multilateral financial institutions. First of all the government attempts to secure nonperforming loans. Three stateowned banks were fully recapitalized with government securities in December 2008, and measures are taken to speed up the privatization of four state owned banks. Togo has more than a dozen commercial banks, complemented by subnational financial establishments and regional banks (PRSP, 2009). IMF (2009b) notes that despite a still relatively limited debt market, Togo's access to both domestic and regional markets has improved the latest three years and will improve further with the planned reforms. Figure 29 show the Banking Sector Size Index, a crude measure of the level of financial intermediation which also takes the level of GDP into account, and does not identify Togo as an outlier. Moreover, looking closer at the Doing Business 2010 data on legal rights, credit information and public registry coverage ­ important pillars for a healthy credit system ­ they are low but do not stand out compared to the same measures in Togo's neighboring countries (Figure 29). Figure 29: Banking Sector Size Index, 2005 [LHS] and Legal Rights, Credit Information and Public Registry Coverage, 2009 [RHS]. High income Middle income Low income Cape Verde Nigeria Senegal Mali Burkina Faso Benin Togo Cote d'Ivoire Ghana Niger Gambia, The GuineaBissau Sierra Leone 0 1 2 3 4 5 6 7 8 Note: The index is the average of the scaled indicators within the dimension of size: Deposit money bank assets to GDP, Central bank assets to GDP, M2 to GDP, Total System Deposits. This means that the index already takes into account the level of GDP for the country in question. Source: Financial Sector Development Indicators, World Bank, and Doing Business 2010, World Bank. Microfinance has covered some of the credit needs, especially in rural areas (IMF, 2009a). By June 2008, microfinance institutions in Togo had savings equivalent to 15% and credits equivalent to 14% of the banking system (PRSP, 2009). The number of beneficiaries were around 540 000, about 8% of the population. This can be compared to averages of 0.5% in ECA up to 2.5% in South Asia, but the country variation is large (Gonzales and Rosenberg, 2006). Also related to credit to smaller actors in the economy is the increased household credits, mainly from the regional banks (IMF, 2009b). 30 Figure 30: Percentage of Loans from Financial Institutions, Togo and Benchmarks. 1.20 1.00 0.80 0.60 0.40 0.20 0.00 Banking sector State banks Non Bank Financial Institutions Other sources Source: ICA data, World Bank. Togo data from 2009. As explained in Section 3.2 above, Togo has sustained current account deficits in excess of 5% of GDP, with low domestic savings and insufficient foreign sources of savings to close the investment gap. In the medium term this gap is expected to be financed by grants and highly concessional loans. There are clearly some indications that limited international finance may have contributed to the historical downturn in GDP per capita in Togo but, given the increase in FDIs, remittances and aid lately, there is no strong case for international finance to be the current constraint to growth in Togo. It might however be an important ingredient in a package that addresses the binding constraints and hence make productive use of well needed capital for investments. Concluding, the analysis of cost and access to finance suggests that the low level of credit is most likely a result of low demand, and hence a consequence of low returns to investments. Problems related to financial intermediation are serious but are being addressed by the government. We will now turn to the analysis of returns to investments in search for the binding constraints to growth. 6.2 Rate of Return on Capital Togo's return on capital appears to be low, and has followed the dismal economic performance of the last decades32. While international comparisons of the rate of return level may not be accurate due to uncertainties on the actual monetary values of capital stocks aggregate capacity utilization, and differences of price level of investment, among others, returns in Togo do appear to be comparatively low, as judged by behavior over time of Total Factor Productivity (Figure 4) and by the apparently low margins received by leading activity sectors, including agriculture, phosphates and cement. Overall trends of rates of return also offer valuable insights (Figure 31). The sustained fall in returns was temporarily reversed by the 100% depreciation on the CFA against the euro in 1994. But they have stagnated ever since, as a higher yield in local currency units did not really lead to a scaling up of nontraditional economic activities. To better understand the rate of return to capital we now turn to the underlying factors determining returns to investments. 32 The rate of return of capital has been estimated as the ratio of income accruing to capital to the value of capital stocks. The former is calculated as the product of total GDP (in real US$ of 2000) times the share of capital on factor costs GDP (assumed constant at 35%). The latter is calculated using the perpetual inventory method, with data series on real gross capital formation (19602008) and assuming a depreciation rate of 5% per year. 31 Figure 31: Rate of Return to Capital, Togo, 19602008. ogo: ate o etu o Cap ta : 960 008 18% Income Accruing to Capital / Estimated 16% 14% 12% Capital Stock 10% 8% 6% 4% 2% 0% 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 Source: Authors' calculations. WDI data, World Bank. 6.3 Geography and Regional Potential Togo's economic prospects are not constrained by its geographic characteristics. That is, when we regard geography in a narrow sense, involving the country's physical characteristics, access to international markets, resource endowments and its degree of exposure to natural hazard phenomena. However, our opening statement needs to be made with modesty when we extend the concept to consider the economic geography dimension. That is, the one that involves the country's current access to dynamic, high income markets with high income, growthenhancing products. We develop these ideas below. Togo is a small, modestly populationdense country resulting in a small domestic market, but many other characteristics compensate for this. With an area of 56,785 square km it only ranks as the 132nd biggest country in the world. This feature, however, does not impose a physical limitation to conduct business, as the country has excellent access to neighboring Ghana, Burkina Faso and Benin; is a coastal economy with a relatively well developed deep sea port in Lomé; is a mostly flat country with savannas in the north, not very sloped central hills, southern plateaus and low coastal plains, all of which extend through 6 distinct geographic regions; and in spite of being small, is not as densely populated as one may think. With 120 people per square km in 2007, population density is higher than the average for LICs (58.5 people/sq. km), ECOWAS (53.3) and SSA (32.3), but below the world simple country average. In this regard Togo ranks 80th among 225 economies. Togo is also relatively urbanized compared to other lowincome countries with almost 40 percent of its population living in cities (20 percent in Lomé). This is an advantage in terms of public investments that may have a very high payoff as they efficiently can reach a large number of the population. Figure 32: Arable Land, WAEMU countries, 2006. 50 1.2 45 40 1.0 Hectares / person % of lanf area 35 0.8 30 25 0.6 20 15 0.4 10 0.2 5 0 0.0 Senegal Ghana GuineaBissau Mali Niger Togo Liberia Cape Verde Nigeria Gambia, The Burkina Faso Benin Cote d'Ivoire Sierra Leone Arable land (% of land area) Arable land (hectares per person) Source: WDI data, World Bank. 32 Togo has also the highest fraction of arable land to country area within ECOWAS (Figure 32). Despite its population density, the amount of arable land per person is second highest in the region, only behind Niger. At the same time, Togo has relatively limited access to water resources, with the fraction of irrigated land (as a percentage of total croplands) being one of the lowest in the world. Water pollution is also a matter of concern in terms of health hazard and impact on the fishing sector, and deforestation linked to agricultural practices is also an issue to be accounted for. Based on Collier and O'Connell (2004)'s classification of countries, Togo is considered to be a "resourcepoor coastal economy" which is generally considered as positive.33 The authors' crosscountry empirical work shows that, overall, countries in this category tend to grow faster than resourcerich economies and than resourcepoor, landlocked economies. However, in the reduced sample of SSA countries, resourcepoor coastal economies still do better than resourcepoor landlocked economies but grow slower than resourcerich countries, indicating that the impact of "syndromes"34 and of the Dutch disease associated with availability of primary commodities is less strong in SSA countries. Yet, another classification by World Bank (2008) has Togo as an agricultural based economy35 showing its continued reliance on primary commodities and traditional activities and the mostly rural character of its population. Note however, that many countries that are resourcepoor are in Collier and O'Connell's classification also considered to be agricultural based economies. These concepts are not at odds as those countries with a sizeable base of minerals, energy and forest, and exports (not GDP) counts for the resource rich country classification. Togo is not immune to hazards of natural origin, but natural disasters recorded in the country have not been as frequent or as severe as in other SubSaharan African countries. EMDAT36, the most comprehensive database on disasters of natural and technological records a total of about 1,200 deaths and 770,000 affected in Togo between 1965 and 2005, most of them linked to epidemics, droughts and floods (Table 9). This is not an insignificant number, but as shown in the table the number of killed and affected taking population size into account is much smaller than the average for SSA. Table 9: Summary of Natural Disasters since 1990, Togo. Togo SSA excl. Togo Population 2000 5,247,486 670,817,822 Killed 65-05 1,172 1,061,604 Affected 65-05 766,161 420,367,666 Killed/year 29 26,540 Affected/year 19,154 10,509,192 Killed/1000000 people 6 40 Affected/1000 people 3,650 15,666 Source: EMDAT: The OFDA/CRED International Disaster Database. 33 A country is deemed to be resourcerich if all the following are true: 1) Current rents from energy, minerals and forests exceed 5% of GNI; 2) A forward moving average of these rents exceeds 10% of GNI; and, 3) The share of primary commodities in exports exceeds 20% for at least a 5year period following this initial year. 34 "Syndromes" are linked, in general, to the absence of political stability and of market friendly policies. 35 World Bank's 2008 World Development Report that focus on the subject of Agriculture for Development, classifies countries as "Agricultural based", "Transforming" or "Urbanized" based on the contribution of agriculture GDP real growth to an economy's total GDP growth, and on the fraction of rural poor to total poor population in the country. Thus way, countries where agriculture contributes with more than 33% of income growth and have a fraction of rural poor population to total poverty headcount higher than 85% are considered to be agricultural based. Togo's figures are 42% in the former variable and 86% in the latter by 2007 and 2006 respectively. 36 Centre for Research on the Epidemiology of Disasters (CRED) at the University of Louvain, Belgium, 33 Moreover, Togo has relatively easy access to neighboring markets. The NorthSouth axis linking Lomé with the northern commercial town of Dapaong provides an adequate transit means to Sahel countries and Benin, and Togo's capital, Lomé, is literally door to door to Ghana's boundary on the west and has direct road access leading to neighboring countries capitals Accra and Cotonou. Furthermore, in the last decade, the port at Lomé has be able to accommodate an increasing freight activity that has resulted from diverted trade from Cote D'Ivoire and other factors discussed below. Potential access to dynamic external market should account for market size in order to ascertain possible demand limitations to a country growth prospect. Togo is a very small economy of 6.5+ million people with low percapita income. Possibilities for sustained growth are therefore strongly linked to potential demand from abroad. One initiative to overcome this geographical constraint has been to join the Economic Community of West African States (ECOWAS), created in 1975 by 15 countries to achieve "collective self sufficiency" through a single trading block by means of economic and monetary union. The West African Economic and Monetary Union (WAEMU) was created in 1994 and has now eight members, and its purpose is (i) greater economic competitiveness, through open markets and harmonized legal environment, (ii) convergence of macroeconomic policies and indicators, (iii) creation of a common market place, (iv) coordination of sectoral policies, and (v) harmonization of fiscal policies. WAEMU is in general considered to have made substantial achievements towards these goals. However, Togo mainly exports lowincome products to lowincome countries. As mentioned in Section 4 above, the income value of its export value, as defined by Hausmann et al. (2005) is low (even though increasing), mainly due to the fact that few of the country's traditional top exports, including cocoa, cotton, fertilizers and coffee, are lowincome products (low PRODY products). In addition, Togo's main trade partners are mostly lowincome countries: Neighboring Ghana, Benin and Burkina Faso represent over 50% of Togo's exports, and lowincome countries in general absorb over 2/3 of Togo's products. Moreover, Togo competes with neighboring ECOWAS partners in the international market for selling its traditional products, other than phosphates and cement. Lastly, the predicted trade flows from Togo to highgrowing economies are actually low considering Togo's economic size and the country distance to more dynamic economies. This will be further discussed when analyzing constraints to selfdiscovery below. 6.4 Infrastructure Figure 33: Logistic Performance Index (LPI) Infrastructure vs log GDP, 2007. Note: Data is based on feedback from logistical operators doing business in the specific country, and is supplemented with objective data on the performance of key components of the logistics chain in the country. Source: LPI 2007, World Bank. Using overall measures, Togo's infrastructure is better than expected given its GDP level, but (i) given its growth strategy as a regional hub there will be a need for continuous improvements of main links, and (ii) infrastructure linking rural and urban areas will be necessary if the growth strategy is to be inclusive. First of all, few firms complain about transport as a major constraint to growth in the ICA data ­ no matter the size of the firm. Keep in mind though that the firms included in the ICA survey were located in Lomé. Furthermore, Figure 33 shows to what extent Togo deviates from the expected level of traderelated infrastructure according to the World Bank 34 Logistic Performance Index (LPI). Togo is ranked number 119 out of 150 countries in the overall LPI, but specifically for infrastructure the ranking improves to number 90. Note however, that this data is from 2007 and the 2008 floods in Togo destroyed a large part of the infrastructure. IMF (2008) estimates an export loss of about 3% of GDP from the destruction due to the flooding. The Port of Lomé is the only natural deep water port in West Africa, and the importance is expected to rise together with the increasing number of large container ships in the region. The port, surrounding infrastructure and the related transport sector (representing about 6% of GDP) serves landlocked countries in the region but also other costal countries when local capacity is limited or the quality of service is lacking. 90% of Togo's external trade goes through the Port of Lomé. In 2008, about 60% of the imported goods had destination Togo, while the other 40% was transit goods. The port traffic in terms of tons increased on average by 9% per year during the period 20032008 and so far there are no signs of a downturn in activity due to the global crises. The recent strong growth is a consequence of several factors such as (i) change in transit traffic due to the crises in Cote d'Ivoire, (ii) relocation of firms from Cote d'Ivoire to the Free Zone, (iii) congestion at other ports such as Cotonou in Benin and Tema in Ghana, that has otherwise made more infrastructure investments than the Port of Lomé, (iv) the economic activities connected to the expansion of the Free Zone, (v) relatively efficient service of the Togo corridor compared to other road corridor options, not least as a consequence of improved management and reduced number of controls (World Bank, 2009). For this growth to continue it is of importance to focus on the constraints in terms of port efficiency and connected infrastructure, as well as the attractiveness of the Free Zone, rather than relying on the external discrete events that has so far played in favor of the Port of Lomé. The road density in Togo is moderate compared to its neighbors' and of those a fair amount is paved compared to other countries at similar GDP levels (Figure 34). Of the 11 700 km road, 3 000 km are national roads and around 9 000 km rural roads. However, Togo is also struggling with bad quality roads due to lack of resources for repairs (only 200300 km are repaired each year), and no new road has been developed for a decade. In 2005, roads in good condition were only 33% of the national network, 13% of the nonpaved roads, and 7% of the urban roads in either category ­ numbers that are expected to be even lower after the impact of recent rains. This problem may become acute both for Togo as an international trade node and for profiting from the port, but also for internal connectivity and the development of commercial agriculture. Resources to maintain and, as economic activity takes off, developing Togo's infrastructure will be crucial ­ for the economy as a whole but agriculture in particular (see further under coordination failures). However, several infrastructure projects are in the pipeline funded by the increased aid to Togo. An alternative, or rather complement, to road transport could be a well developed rail network. However, Togo's rail network is close to nonexistent and is only used for limited transport of cement and coal. Given Togo's potential in mining, the underdeveloped rail network could be limiting growth. Togo has two international airports that are both used below their capacity. This may have become more an important issue as a result of the recent international economic and financial crisis. Figure 34: Road Density (km road per sq. km), 20012006 (LHS), and Roads, paved (% of total roads), 20002006 (RHS), Togo and Benchmarks. High income Middle income Low income Cape Verde Togo Senegal GuineaBissau Niger Gambia, The Mali Nigeria Ghana Benin Cote d'Ivoire Sierra Leone Liberia Burkina Faso 0 10 20 30 40 50 60 70 80 90 Note: Data points were from different years for different countries but all within the period 20002006. Togo data is from Assiongbon and Chausse (2008). Source: DDP data, World Bank. 35 Electricity consumption in Togo is at the level expected for its GDP per capita level but only a third is covered by domestic production, and the gap is increasing (Figure 35). That consumption is at expected levels given the level of GDP despite a high price, signal a strong demand within Togo. The Togo Sources of Growth (2008) study report that the average electricity price in Togo is 100 FCFA/KWh, which is substantially higher than in SSA (56 FCFA/KWh), Latin America (30 FCFA/KWh) and South Asia (17 FCFA/KWh). The Free Zone firms in Togo are very capital intensive compared to many other free zones and expectedly, electricity is in the ICA data mentioned as one of the major constraints for large firms. However, electricity as a constraint is not cited as one of the major constraints by medium and smaller firms and access in terms of number of days to set up a connection seems to be in line with other countries (Figure 36). Figure 35: Energy use (kg of oil equivalent per capita) vs log of GDP per Capita (constant 2000 USD), 2006 (LHS) and Electricity Consumption and Production in Togo, 19712005 (RHS). Source: DDP Data, World Bank. Figure 36: Average Number of Days Waiting for Services (Electricity Connection, Water and Telephone Line), Togo and Benchmarks. 300 250 200 days 150 100 50 0 Cameroon Mauritius Togo Senegal South Africa Mozambique Ghana Botswana Burkina Faso The Gambia Electricity connection Water Telephone line Source: ICA data, World Bank. Togo data from 2009. Nevertheless, the problem remains for the larger firms and the quality of service provided by the stateowned electricity company (CEET) has been very low with production frequently disrupted by electricity shortages.37 A recent energy crisis emerged from several reasons: (i) institutions ­ the concession for distribution has long been disputed and the contract was cancelled in 2005, (ii) technical ­ the draught reduced the available hydroelectric power production, and (iii) financial ­ increased payment arrears to Ghana and Cote d'Ivoire power suppliers due to higher operating costs from more fuelbased power generation. Electricity constraint has recently been relaxed through investments in new generators but is only a short term solution to an underlying structural problem in the noncompetitive energy market. Of the electricity production in Togo, about half of it comes from hydroelectric sources and half from oil sources, while options of natural gas production are currently considered. One long term 37 CEET is responsible for the transport and distribution of electricity while the Communaute Elecrique du Benin (CEB) assures most of the production (90 percent, and the rest by CEET). 36 option for Togo to be more self dependent is the construction of the Adjarala dam, but the efficiency of such an investment will depend on the energyintensity of Togo's economy going forward. The low percentage of the population with mobile cellular subscription in Togo seems to be more of a demand or monopolistic behavior issue than lack of access. The relatively low mobile subscription in Togo (see Figure 37) is not reflected in a low percentage of the population covered by a network (Figure 38). Also, as shown in Figure 36 above, the number of days needed to wait to get a telephone line is very small in Togo compared to other African countries. However, the high price (see again Figure 38) does indicate that there is either low demand and hence not a concern, or noncompetitive price behavior by the state monopoly (in terms of installation and the international gateway which is handled by TOGOTELECOM). It needs to be noted though that the number of subscribers increased from 55,800 to 708,000, 2000 to 2006, so the price does not seem to be the constraint at this point and may change as the market conditions change. A decline in the price of telecommunications will be crucial for further increases in FDIs as firms currently face by very high prices for international calls, be it from a cell phone or a fixed line, compared to neighboring countries.38 In terms of internet access, Togo has lately made significant progress compared to the other West African economies (Figure 37). Figure 37: Mobile cellular subscriptions, 2007 (LHS), and Internet Users, 19962006 (RHS) (per 100 people). High income Middle income Low income Gambia, The Cote d'Ivoire Ghana Senegal Cape Verde Nigeria Benin Mali Togo GuineaBissau Liberia Sierra Leone Burkina Faso Niger 0 20 40 60 80 100 Source: DDP Data, World Bank. Figure 38: Population covered by mobile cellular network (%), 2006 (LHS), and Mobile cellular prepaid tariff (US$ per month), 2008 (RHS). High income High income Middle income Middle income Low income Low income Cape Verde Cape Verde Togo Senegal Togo Sierra Leone Burkina Faso Gambia, The Benin Benin Cote d'Ivoire Ghana Niger GuineaBissau Nigeria Burkina Faso Mali Nigeria Senegal Cote d'Ivoire Gambia, The Niger Ghana Mali 0 20 40 60 80 100 0 5 10 15 20 Source: DDP Data, World Bank. 38 Togo Sources of Growth (2008) report that fixed line phone calls to the West African region from Togo is 354 FCFA per minute compared to 250 from Cote d'Ivoire, 130 from Senegal and 148 from Mali. The rate calling from a cell phone is about 300 FCFA per minute. Calling to Togo from for example the US is again 1.6 to 4.2 times more expensive than calling to Benin, Ghana, Cote d'Ivoire or Burkina Faso from the US. 37 6.5 Human Capital Some of Togo's human capital indicators are relatively weak but do not stand out compared to the rest of the region. Figure 39 suggests that Togo fares relatively well with respect to health ­ infant mortality rate is the second lowest in the group and the share or birth attended by a skilled health worker is the second highest. With regard to education, Togo has a relatively high primary enrollment rate but the primary completion rate is relatively weak, which indicates high repetition rates and a high number of dropouts. Figure 39: Health and Mortality Indicators, Togo and benchmark countries, 2007. 180 100 Total enrollment, primary (% net) 160 Mortality rate, infant (per 1,000 live births) 90 Primary completion rate, total (% of relevant age group) 140 Births attended by skilled health staff (% of total) 80 120 70 60 100 50 80 40 60 30 40 20 10 20 0 0 Source: DDP data, World Bank. Despite the prolonged economic stagnation, public expenditures per capita have been maintained at least a few years back. Public health expenditures are more or less constant and even though public education expenditures as a share of government revenues and GDP may have a downward trend, the expenditure per capita is more or less constant (Figure 40). Figure 40: Public health and education expenditure, Togo 20022006. 2 Health expenditure, public (% of GDP) (lhs) 12 5 25 Health expenditure, public (% of government expenditure) (rhs) 1.8 4.5 Per capita public health expenditures (const US$) (rhs) 10 1.6 4 20 1.4 3.5 8 1.2 3 15 1 6 2.5 0.8 2 10 4 0.6 1.5 Public spending on education, total (% of GDP) (lhs) 0.4 1 5 2 Public spending on education, total (% of government expenditure) (rhs) 0.2 0.5 Public spending on education, per capita (const US$) (rhs) 0 0 0 0 2002 2003 2004 2005 2006 2000 2001 2002 2003 2004 2005 2006 2007 Source: DDP data, World Bank. Total health expenditure per capita is relatively low, and the share of private health expenditures to total expenditures is high compared to the benchmark countries (Figure 41). The high share of private expenditure indicates a response to a lack of public health care but may be an expected result of a conscious health policy. Around half of the population in urban areas seeking treatment chooses a private facility. In rural areas, the same number is one third. Access to health facilities seems not to be a major issue in most parts of the country the absence of a health facility as a reason for not seeking treatment in rural areas was only given in around 10% of the cases. The vast majority of those who considered treatment necessary but did not seek treatment rather did so because they considered treatment to be too expensive. Satisfaction with the treatment is higher in private clinics, especially in urban areas where 82% of those treated in a private facility are satisfied with the treatment versus 65% in public facilities. 38 Figure 41: Public and Private Health Expenditure per Capita, 2006 (RHS), and Type of health specialist visited, satisfaction with treatment and reasons for not seeking treatment, Togo 2006 (LHS). 50 Went to see health specialist Total Urban Total Rural 45 Public Type of institution 40 35 Private Yes public 6.3% 7.3% 30 of which satisfied with treatment 65.4% 73.5% 25 private 7.4% 3.7% 20 of which satisfied with treatment 81.9% 77.1% 15 10 No 86.2% 88.9% 5 Reason for not seeing health specialist 0 not necessary 88.0% 87.4% too expensive 11.4% 11.7% not available/ too far 0.2% 1.5% other 0.7% 0.5% Source: DDP data, World Bank, and author's calculations using CWIQ 2006. The high repetition rate has created a vicious circle, leading to an additional drain on limited public funds. The repetition rate is the highest of all the benchmark countries for primary and the third highest for secondary schooling, and is part of the reason for as well as an implication of the high pupil teacher ratios (Figure 42). Figure 42: Repetition Rate and PupilTeacher Ratio for Primary and Secondary Schooling, 2007. 60 40 Pupilteacher ratio, secondary 35 Primary Secondary 50 Pupilteacher ratio, primary 30 40 25 20 30 15 20 10 5 10 0 0 Source: DDP data, World Bank. Despite some problems, Table 10 below suggests that skills are not a current constraint to economic activity as both unemployment and underemployment are high for all levels of education. However, there is an important difference across education levels. Unemployment increases with the education level, indicating that the better educated, and probably better off, prefer to be unemployed looking for employment than to work and be underemployed. Conversely, the lower educated and poorer segments of the population have to accept whatever employment they can find and, therefore, are more often underemployed. So it seems that Togo still has potential to use more of the educated workers and to use the ones already employed in a more productive fashion. Table 10: Unand Underemployed by Education Level for Rural, Urban Areas and Lomé, Togo 2006. Unemployed Underemployed Urban Rural Lomé Urban Rural Lomé None 3.9% 0.6% 5.4% 19.4% 24.5% 22.5% Primary 7.4% 2.3% 7.8% 20.7% 25.2% 21.2% Junior high 7.4% 3.9% 8.3% 15.7% 21.4% 17.6% Senior high & above 11.4% 2.3% 12.2% 13.7% 18.6% 14.6% Total 7.4% 1.4% 8.5% 17.8% 24.2% 19.0% Source: CWIQ 2006 and author's calculations. There is a strong correlation between low education levels and poverty, already discussed in the Profile of the Economic Actors section and confirmed in the regression results presented in Table 11. Compared to households 39 in which the household head has not finished primary education, the risk of poverty decreases for households in which the head of household has finished at least primary school and junior high school, in rural as well as urban areas. However, the main reason for the substantial number of poor is probably again the absence of productive employment possibilities. Three main reasons speak for this: First, the probit regressions indicate that the risk of poverty in urban areas depends statistically significantly on the region, even after controlling for education. This indicates that workers with the same level of education face different possibilities for productive employment across the regions. Similarly, rural workers in sectors other than agriculture also face a lower risk of poverty than workers in agriculture. Second, the substantial share of workers who are selfemployed or work as unpaid family helpers points to the absence of more productive employment opportunities. Third, and as mentioned, Table 10 shows that workers of all levels of education are affected by under and unemployment. Table 11: Marginal effects for probit regression on the determinants of poverty, urban and rural areas, Togo 2006 Urban Rural Coef. z Coef. z primary 0.04 1.87 0.07 2.48 junior high 0.05 2.11 0.09 2.56 senior high and above 0.11 6.09 0.38 2.79 Agriculture 0.07 1.86 Mining 0.02 0.29 0.10 0.89 Manufacturing and utilites 0.00 0.07 0.01 0.19 Construction 0.01 0.43 0.12 1.33 Transport/ communication 0.01 0.20 0.12 1.83 Commerce and hotelery 0.18 3.48 Education/ health 0.00 0.08 0.17 2.33 Public Administration 0.03 0.71 0.21 2.17 Services/ banks/ insurances 0.02 0.52 0.17 1.32 Other services 0.00 0.04 0.11 0.86 HH size 0.04 7.65 0.10 15.33 Lomé 0.53 7.52 Maritime 0.08 3.55 0.16 2.66 Plateaux 0.10 5.75 0.37 6.54 Centrale 0.07 3.28 0.09 1.62 Kara 0.04 1.12 0.06 1.22 Age 0.00 3.13 0.00 2.21 sick last 4 weeks 0.04 2.03 0.12 5.55 underemployed 0.02 1.14 0.05 2.00 Land per capita 1.67 1.98 Source: CWIQ 2006 and author's calculations. Mincerian wage regressions reveal the existence of some premium ­ as expected for additional years of education, but these are very low compared to other SSA countries39. This regression of earnings on a proxy for human capital (the education of the labor force), controlling for a set of individual, community, household and regional, characteristics, shows a positive nonlinear relation between educational attainments and the dependent variable (Figure 43 and Table A4 in Appendix 1). For the average population, the mean marginal return to education stands at 5.2%, which is quite close from the findings of Kuepie et al (2008) who, by means of a similar specification estimated returns to education to be around 5% in Lomé. Figure 43 also shows that wages, given the level of primary education completion rate, are very low in Togo compared to other WAEMU countries further indicating a low demand from firms for more educated work force (See Appendix, Table A4 for summary econometric results from the Mincerian specification). 39 There are a number of methodological problems when comparing regressions on returns to education across countries, but Psacharopoulos (1994) can give us a sense that in most SSA countries the return for different levels of education is usually above 15%. 40 Figure 43: Returns to Education in Togo (Mincerian Analysis) (LHS), and Wages vs Primary Education Rate in Togo and benchmark countries (RHS). 0.6 700 Education and Wages in WAEMU Countries BEN % C h an g e in 0.5 600 Avg.Monthly Salary (USD) . E a rn in g s COT 0.4 500 SEN 0.3 400 0.2 300 BUR MAL TOGO 0.1 200 NIG 0.0 100 GNB 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 0 Years of Education 0 10 20 30 40 50 60 70 Primary Completion Rate (%) Source: Authors' calculations, LSMS data, and IMF (2009b). Finally, the ICA data reveals that none of the interviewed firms had any concerns regarding skills of the labor force. In fact very low proportion of interviewed enterprises revealed much concern regarding the availability of skilled workers or labor regulations in Togo. In addition to this it is worth mentioning that over the long run, as the economy of Togo deteriorated, human capital has improved ­ suggesting a low correlation between human capital and growth for Togo at this point in time. In light of this, we cannot conclude that education is currently binding constraint to growth in Togo. That, of course, does not exclude a need for specific skills or a need for reforms given expected future demand as the economy develops. 6.6 Macroeconomic Stability Togo has been affected by the financial crisis, but its impact has been moderate. Before analyzing potential macroeconomic constraints it needs to be clear what in the current situation is due to the global crises and what is due to actual constraints in the Togolese economy (see Box A 1 in Appendix 1 for more details on the effect of the crises).40 Growth has been estimated at 3.1% in 2009, 0.9 percentage points lower than the previous growth projection for 2009. The dependence on commodity exports and official and private capital inflows has to some extent constrained growth during the crisis (IMF, 2009a). Remittances, a significant source of income for Togo (almost 10% of GDP), are expected to be further affected by lower global demand as well as unemployment and lower wages in Europe and the US. Another possible negative impact is a lower than expected level of FDIs. On the other hand, Togo's limited integration with international finance and crosscountry linkages within the banking system may soften that negative impact of the crisis. Also, donor support is not expected to drop significantly. Finally, even though the potential growth rate will be affected, this may not translate into lower overall growth than precrises as public investments and a rebound in agricultural production is expected. Hence, the impact of the financial crises on Togo is expected to be more of a delay in rapid growth and poverty reduction but this can be serious enough given the current low levels of economic activity (IMF, 2009b).41 As mentioned, Togo has made significant progress in terms of macroeconomic stability. Togo is in line with the PRGF targets and the membership in WAEMU, as well as the regional central bank (BCEAO) that controls its monetary policy creates stability in terms of both monetary and fiscal policies. Table 12 confirms what was already discussed in the section on cost of capital that Togo has a current account deficit that needs to be covered from 40 The trade balance and the current account is according to IMF expected to improve in 2010 as global demand recover in combination with increased productivity, led by sectors such as phosphate and cement (IMF, 2009a). 41 The current financial crisis is expected to hurt lowincome countries as a second wave after the more advanced economies have been hit affecting the demand on the global market. However, to what extent different lowincome countries will be affected by this lower demand and lower commodity prices depends on factors as the level of international integration (goods, services, migration, etc), the import or export dependency of affected commodities (such as oil, food, and minerals), tax structures, etc. Also, the short term effect may be very different to the medium term effects as countryspecific factors may aggravate, mitigate or offset the direct effects. 41 foreign sources. Even though FDI and remittances have been on the rise, there is still a gap that is now predicted to be filled by an increasing amount of foreign aid. Table 12: Balance of Payment Items, Togo, 20032007. Column1 2003 2004 2005 2006 2007 2008 Domestic Primary Balance (% of GDP) 4.5 3.2 0.5 1.1 0.2 0.4 Trade Balance (% of GDP) 14 15 16.9 19.6 10 Current Account Balance (% of GDP) 4.2 3 5.3 6 3.9 Gross int. reserves (months of imp) 2.3 3.6 2 3.2 3 Source: International Financial Statistics, IMF. Fiscal policy has meet and even gone beyond its targets, although this is partly a consequence of limited execution on the expenditure side. The domestic primary balance was +0.4% of GDP in 2008, and tax and custom revenue collection were higher than expected. Current expenditures represented 87% of total expenditures while investment spending represented 13%. Capital spending was under executed (among the lowest in SSA as reflected in the low public investments in Figure 44), mainly due to capacity constraints and complex procurement procedures. This may turn out to be an important indirect binding constraint as the public investments expenditures are supposed to finance crucial elements of future growth in Togo (see further Section 5.6 below). Figure 44: Public Capital Investments, WAEMU, latest available data (% of GDP) 14.0 12.0 10.0 8.0 6.0 4.0 2.0 - Niger Burkina Mali Benin Senegal Cote WEMU Togo Faso d'Ivoire Total Domestically financed Foreign financed Source: IMF (2009a). Togo qualified for the HIPC initiative and will upon completion point qualify for the Multilateral Debt Relief Initiative (MDRI). Togo's NPV of debttorevenue ratio in end2007 was 309%, and to reach 250% requires 270 million USD in HIPC relief. The MDRI is then expected to reduce the debt with another 404 million USD. The development of an interim PRSP and now a PRSP, as well as the progress on triggers related to macroeconomic stability, public financial management, governance, debt management and social sectors ­ monitored under the IMF's PRGF program and IDA's Development Policy Operations ­ have made it possible for Togo to make significant progress towards the completion point, which the government is hoping to reach in early 2010. Togo stands out in terms of the share of arrears in the external debt, but many of these arrears have been cleared during 2008 and the remaining is under process of being cleared or rescheduled (World Bank, 2008). Both trade partners and sources of remittances are relatively diversified, lowering Togo's macroeconomic vulnerability. As previously reported, remittances have started to make a significant contribution to Togo's economy. Of the remittances, 30% come from Europe (mainly France), 20% from the US and 15% from within the WAEMU. The diversity of countries is also reflected in the number of trading partners (Figure 45). 42 Figure 45: Togo's Main Trading Partners, 2008. Source: IMF (2009b) Inflation has been fairly stable and single digit in Togo since the 1980s, with the exception of 1994. As can be seen in Figure 46, the recent increase seems to be related to the increase in the world prices of oil and food, affecting oil and food importers such as Togo. Inflation was up at 15.8% in August 2008, but fell to 8.4% by the end of the year ­ hand in hand with falling oil and commodity prices (IMF, 2009a). Hence, there does not seem to be a structural domestic problem needed to be addressed and the oil and food prices have now started to decline. Moreover, the inflation trend during the 2000s follow that of the other WAEMU countries and, the levelwise higher, SSA average (IMF, 2009a). Figure 46: Inflation in Togo and World Food and Oil prices, 19672008 300 45 40 250 35 30 200 25 20 150 15 100 10 5 50 0 5 0 10 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 Food, US$ price index (1990 = 100) Crude oil, $/b, 1990 US$ Inflation, consumer prices (annual %) Source: DDP Data, World Bank; IMF (2009a); Global Economic Monitor, World Bank. Since the devaluation, Togo's real effective exchange rate follows that of WAEMU which has appreciated somewhat since the 1980s but with no signs of a serious overvaluation. The exchange rate dropped significantly in 1994, following the devaluation, but has since been on an upward trend. However, estimations by IMF presented in Figure 47 show that most longrun changes can be explained by changes in the fundamentals, leaving little reason to assume an over evaluation. The recent gap has been due to inflation pressure and the Euro's appreciation against the dollar ­ trend that has now reversed and together with improved terms of trade is expected to close the gap (IMF, 2009a). Taken together, macroeconomic stability cannot be argued to be a binding constraint to growth in Togo at this point in time. 43 Figure 47: Togo's real effective exchange rate (REER) expected REER (EREER), 19802008. Note: To calculate the EREER, IMF used the FEER approach with the following fundamentals: terms of trade, government spending, openness, and real GDP per capita--all variables that are significantly correlated with the REER and share a cointegrating relation. Source: IMF (2009a). 6.7 Governance, Corruption and Government Efficiency Governance indicators such as political instability and corruption were the most often cited major constraints in the ICA data. Political instability seem to affect all types of firms while corruption is more of a problem for medium and smaller firms who are assumed to pay a larger share of their total profits as informal payments. On practically every measure of governance, Togo's performance is the poorest compared to neighboring countries and has worsened over the period 1996 to 2007. Box 3 shows that in 2007 the government has become slightly more effective in `controlling corruption' (relative to 1996)42 and the `rule of law' indicator suggests Togo is doing better. However, on all other counts of governance, especially `regulatory quality' and `government effectiveness', Togo has fallen behind its level in 1996. `Voice and accountability' and `Political instability' are as big concerns in 2007 as in 1996. At a first glance it is therefore hard to rule out governance issues as an explanation for the low growth performance in Togo. What needs to be noted though is that growth has started to take off lately without any signs of improvement in most of the governance variables, suggesting that they are not binding at this point in time. In terms of overall budget execution the rate has been acceptable for current expenditures but there are major shortcomings in investment execution. The Togo PRSP (2009) reports that the overall rate of budget execution was only 52% in 2008. In particular, the rate of execution of investments in the transport sector was only 17%, and 2% in the energy sector, while it looked better in for example the agriculture sector where the rate was 85%. However, this may rather be explained by the nature of projects within the different ministries. It has been argued that the limited absorption capacity may be a result of the socioeconomic crises in Togo that deteriorated financial, institutional and human resources available for the government. It should be noted though, that the Government of Togo has shown an impressive ability to meet reform targets, making it questionable if capacity is a major constraint within the government at this point. 42 Note however, that Togo scores only 2.7 in 2008 Corruption Perception Index by Transparency International resulting in a country rank of 121 out of 180 countries. The CPI score relates to the perception of corruption as seen by business people and country analysts, and ranges between 10 (highly clean) to 0 (highly corrupt). 44 Box 3: Governance Indicators: Togo and Benchmark Countries 0.75 Voice & Accountability Government Effectiveness 0.35 Ghana 0.25 Benin 0.00 Senegal 0.35 0.25 MIC Burkina Faso Nigeria 0.70 0.75 SSA SSA 1.05 Nigeria 1.25 Togo 1.40 Togo 1.75 1.75 2000 2002 2003 2004 2005 2006 2007 2008 2000 2002 2003 2004 2005 2006 2007 2008 Political Stability Regulatory Quality 0.35 0.75 Benin Ghana 0.00 0.25 Ghana Senegal MIC Burkina Faso 0.35 0.25 MIC Burkina Faso Senegal Benin 0.75 0.70 SSA SSA 1.25 Togo 1.05 Togo Nigeria 1.75 1.40 2.25 1.75 2000 2002 2003 2004 2005 2006 2007 2008 2000 2002 2003 2004 2005 2006 2007 2008 Rule of Law Control of Corruption 0.35 0.35 0.00 Burkina Faso 0.00 Ghana Senegal Senegal 0.35 MIC Benin 0.35 0.70 Benin SSA 0.70 1.05 Togo 1.05 Togo 1.40 1.40 Nigeria Nigeria 1.75 1.75 2000 2002 2003 2004 2005 2006 2007 2008 2000 2002 2003 2004 2005 2006 2007 2008 Source: Governance Database, World Bank. The ownership structure and management of state owned enterprises (SOEs) have also been a reflection of the poor quality of governance. The phosphate sector is operating at less than a quarter of its capacity and IMF (2009a) argues that GDP would increase by as much as 2% if full capacity is reached. The phosphate society OTP/IFG has been replaced by the Societe Nouvelle des Phosphates du Togo (SNPT), but the authorities have so far resisted proposals to identify a new strategic partner which will be needed to bring in the necessary capital and technical expertise. The main cotton company (SOTOCO) was highly criticized and has now been replaced by a new company in which the producers hold 40% of the capital. However, here too the government has delayed plans to 45 introduce a private partner. Several private banks required a temporary government bailout but here the authorities have more clearly underlined their commitment to privatization. The container terminal at the port has been turned over to a private operator and the government is supporting an extension of the port to include a second container terminal. . In short, the government has acknowledged the constraint presented by stateowned enterprises but also remains somewhat skeptical of privatization as a result of past experiences, notably in the phosphates and electricity sectors. We will now turn to barriers and costs of doing business, which is another dimension of governance in the sense of the rules set up by the government for formal businesses. 6.8 Barriers to and Costs of Doing Business Competition with the informal sector is in the ICA data mentioned as one of the major constraints, especially for larger firms, even compared to other countries (Figure 48). This by itself suggests that, in general, the type of regulations and administrative barriers you meet as a formal firm might be an important constraint in Togo. Furthermore, the figure presents the perception of Togo firms when it comes to a number of governmentbusiness relation categories and they are all more negative than the average of other ICA surveyed countries in SSA. Figure 48: Percentage of Firms that Think Competition with the Informal Sector is a Major Constraint to Business (LHS), and Perception of BusinessGovernment Relations Categories as a Major Constraint to Business, Togo and Benchmarks (Higher is worse) (RHS), Togo and Benchmarks. 100% Tax 90% Administration 80% 0.80 70% 0-1 (1=WORST) 0.60 60% Courts 0.40 Tax Rates 50% 40% 0.20 30% Togo 20% 0.00 10% 0% SSA Licensing and Corruption Togo Ghana Mauritius Senegal South Africa Burkina Faso Cameroon The Gambia Botswana Mozambique Permits Political Instability Source: ICA data, World Bank. Data for Togo from 2009. Figure 49: Economic Freedom Index and its components, Togo and SSA, 2008 (LHS), and Index of Economic Freedom, Togo and Selected Comparators, 19992009 (RHS). Sub-Saharan Togo Africa 2009 Overall 53.1 48.7 Business Freedom 51.9 36.4 Trade Freedom 64.9 70.6 Fiscal Freedom 72.2 53.7 Gov't Size 70.7 86.6 Monetary Freedom 71.4 81.5 Investment Freedom 42.1 30.0 Financial Freedom 40.9 30.0 Property Rights 32.6 30.0 Freedom from Corruption 29.0 23.0 Labor Freedom 55.6 44.7 Source: Authors' Calculations based on data from The Heritage Foundation. 46 Togo's overall `economic freedom' score in 2009 is very low even though it has improved slightly since 1999. Out of 46 countries in the SubSaharan Africa region, Togo is ranked 33rd. Figure 49 suggest that Togo scores especially low in terms of Freedom from Corruption but in a relative sense, Togo does worse in terms of Business Freedom. The figure also shows that Togo has made some progress since beginning of 2000 but is still at a very low level. The sources for this are twofold. First, going back to 1999 Togo started from a low level, much lower than all comparator countries, so some catching up was expected. Also, this gap continued to narrow until 2005 not only because Togo showed improvements but also on account of poor performance of benchmark countries. In the past two years, while the other countries have picked up, Togo has fallen behind. This section looks closer into some of the components of economic freedom.43 Figure 50: Business Freedom Index, 2009. Source: Doing Business Indicators 2010. Table 13: Monetary and NonMonetary Costs of Doing Business in Togo, 2009. Starting a Closing a Registering Enforcing Paying Taxes Business Business Property Contracts Cost (% GNI per Cost (% Cost (% of Total tax rate (% Cost (% of claim) capita) of estate) property value) profit) Ghana 26 22 1 33 23 Nigeria 78 22 21 32 32 Burkina Faso 50 9 13 45 83 Senegal 64 7 21 46 27 Togo 205 15 13 53 48 Benin 156 22 12 73 65 Sub-Saharan Africa 100 20 10 68 49 Middle-income 31 18 6 43 33 Starting a Closing a Registering Enforcing Paying Taxes Business Business Property Contracts Procedures Procedures Procedures Procedures Time (number)/Duration (number)/Duration (number)/Duration (number)/Duration (years) (days) (days) (days) (days) Ghana 8/33 1.9 5/34 33/224 36/487 Nigeria 8/31 2 13/82 35/938 39/457 Burkina Faso 4/14 4 4/59 46/270 37/446 Senegal 4/8 3 6/124 59/666 44/780 Togo 7/75 3 5/295 53/270 41/588 Benin 7/31 4 4/120 55/270 42/825 Sub-Saharan Africa 9/46 3.4 7/81 38/306 39/644 Middle-income 9/41 3.1 6/57 35/353 39/657 Note: Middle Income includes both lower-middle and upper-middle income countries as defined by the World Bank Source: Authors' calculations based on Doing Business Indicators 2010. 43 Note that the concept of Economic Freedom is much wider than what is the focus of this section, but it is also important to look at indicators of the business environment in this broader sense as they all interact in a market economy and together affect the incentives for investments and expansion. Some of the categories refer to aspects that are discussed in other sections. 47 Togo's business environment in a more narrow sense of direct costs and barriers to firms, is notably constrained compared to benchmark countries, especially for starting a business (see Figure 50 and Table 13). According to `Doing business in 2010', Togo was ranked as low as 165th out of 183 countries in terms of the overall score of ease of doing business. Even though these numbers may not apply to all firms in reality ­ some firms may have means or connections to avoid some of them ­ they do speak for a cumbersome general business environment meeting an average firm in Togo. Looking specifically at protection of property rights, a fundamental basis for growth, Togo is well below many of its neighboring countries (Figure 51). According to the Heritage Foundation's 2009 Index of economic freedom "the judicial system is subject to strong influence from the executive." Moreover, poorly defined inheritance laws make property ownership difficult and lack of transparency in the judicial system thwarts resolution of property disputes (see also the Investor Protection Index in Figure 51). The average loss for a business is 47.5% of the value of the contract in dispute costs (Togo Sources of Growth, 2008). `Registering a property' is another limitation for starting or expanding a business. Togo is doing okay in terms of number of procedures but the time to get a permission of 295 days is substantially longer than other countries (see Table 13 below). It is ranked as low as 155th out of 183 countries in terms of ease of registering property in the 2010 Doing Business Index, 147 for Investment protection, and 154 for Contract enforcement. Note however, that "Courts" as a major or severe constraint was mentioned by very few firms in the ICA data, questioning if protection of property rights is a binding constraint to growth in Togo at this point ­ but will potentially become if not addressed. Figure 51: Property Right Index (LHS) and Investor Protection Index (RHS), Togo, 2009. Source: Doing Business Indicators 2010. For `Starting a business', Togo is one of the lowest ranked in the world (170th out of 183 according to Doing Business data). The cost of starting a business in Togo is more than double the average cost in SSA and 6.6 times more expensive than in a typical middle income country. The number of procedures required is 7, which is an improvement from the 13 needed the previous year44, but in parallel the time has increased with about 40%. It takes on average, 75 days to start a business in Togo compared to a 36 day world average.45 44 Business startup was eased by setting up a onestop shop that eliminated six procedures and lowered costs by almost a fifth. 45 To obtain a license the entrepreneur needs to go through a lawyer, the Centre de Formalite des Enterprise de la Chambre de Commerce, the Ministry of Commerce or the Export Zone. However, authorization is given by the Ministry of Commerce and the Direction Generale des Impots. Within the Ministry of Commerce the application is going through several instances and it is the Minister himself that signs the approval, something that causes severe delays. Moreover, the argument for this system is to be able to track the development of the private sector but little is done and there is no clear agreement on how to transfer the information to the relevant sector ministry. (Togo Sources of Growth, 2008). According to the Doing Business Index 2010, the main step that delays the startup is the deposit of all the documents for company registration and payment of fees at the Guichet Unique: CFE. 48 Labor market rigidities are present in Togo, but are not perceived as a major constraint by firms. It ranks as low as 159th out of 183 in the 2010 Doing Business data. Two measures are presented: a rigidity of employment index46 and a redundancy cost measure. The `Rigidity of Employment Index' has worsened in Togo since last year and is now 54 for Togo, substantially higher than the average for the world (27) and even SSA. Note however that no firms complain about labor market rigidities in the ICA data, making it hard to argue that this would be a binding constraint at this point. The only dimensions that stands out as better in the Doing Business data was `Trading across borders' (rank 87) and `Closing a business' (rank 97). This is also confirmed in the overall economic freedom score where the component `Trade Freedom' is doing relatively well (see Figure 49). Togo has the lowest cost of exporting and importing (US$ per container) compared to other West African countries, and the documents required to export are the lowest in Togo and Senegal (Figure 52). Table 14 on logistical performance confirms that Togo performs higher than expected in some traderelated dimensions, such as infrastructure, logistics competence and domestic logistics costs. However, Togo has the fifth lowest rank of the 150 countries in the sample when it comes to timeliness. The overall conclusion is however that this structure clearly encourages the firms within the Free Zone and reexportation, but for the benefits to reach Togo's businesses as a whole there is a need for improvements along other doing business aspects as well. Figure 52: Trade costs, West Africa, 2009 Source: Doing Business 2010. Table 14: Logistic Performance Index (LPI) for West Africa, World Ranking 2007 International Logistics Domestic logistics Overall LPI Customs Infrastructure shipments competence Tracking & tracing costs Timeliness score rank score rank score rank score rank score rank score rank score rank score rank Gambia 2.52 77 2.25 89 2.33 76 2.67 69 3 46 2.33 99 3 67 2.5 132 Benin 2.45 89 1.8 142 1.89 134 2.78 62 2.56 75 2.89 54 3.22 24 2.78 107 Nigeria 2.4 93 2.23 96 2.23 92 2.49 92 2.38 96 2.36 97 2.9 83 2.69 114 Senegal 2.37 101 2.38 76 2.09 108 2.09 130 2.73 62 2.3 103 3.09 45 2.63 123 C ote d'Ivoire 2.36 102 2.22 98 2.22 94 2.13 128 2.38 97 2 128 3 68 3.25 61 Liberia 2.31 105 2.4 72 2.14 101 2.83 57 2 127 2 132 3.2 31 2.43 134 Mali 2.29 109 2.17 106 1.9 132 2.23 119 2.21 114 2.38 92 3.05 53 2.88 93 Guniea-Bissau 2.28 116 2.14 111 2.25 89 2.22 120 2 132 2.22 114 3.14 35 2.86 94 Togo 2.25 119 2.1 117 2.25 90 2.4 101 2.4 94 2.2 115 3.33 14 2.11 145 Burkina Faso 2.24 121 2.13 115 1.89 133 2.67 71 2.33 102 2.13 122 2.67 118 2.25 143 Ghana 2.16 125 2 129 2.25 91 2.25 115 1.75 146 2.25 110 2 149 2.5 133 Niger 1.97 143 1.67 145 1.4 149 1.8 145 2 134 2 133 1.67 150 3 85 1.95 144 1.58 149 1.83 140 1.82 143 1.91 140 2 135 3 69 2.64 121 Sierra Leone Note: Data is based on feedback from logistical operators doing business in the specific country, and is supplemented with objective data on the performance of key components of the logistics chain in the country. Source: LPI 2007, World Bank. 46 The rigidity of employment index is the average of three subindices: difficulty of hiring, rigidity of hours and difficulty of redundancy, and the Redundancy cost indicator measures the cost of advance notice requirements, severance payments and penalties due when terminating a redundant worker. 49 Finally, `Paying taxes' (including taxes, import fees and other compulsory yearly charges for an average firm) ranks low in the Doing Business Index (155th out of 189 countries), but has improved lately.47 The rank has improved since last year due to improvement in total tax rate as a percentage of profits. It was decided in 2008 to reduce import fees from 37 to 30% for the industries, and from 40 to 33% for the other sectors.48 Also, looking at the ICA data, few firms are mentioning the tax rate as a constraint, while the tax administration is still cumbersome for a number of firms. Tax revenues as a percentage of GDP are shown in Figure 53, and Togo does not come out as an outlier ­ not even when considering the expected level given its GDP per capita which is lower than the actual level. Figure 53: Tax revenue (% of GDP) vs Log GDP per capita (PPP, 2005 int. dollars), 2007. Source: DDP data, World Bank. If the problems of doing business in Togo are binding, one may expect to see higher growth among the firms within the Free Zone as it was created precisely to ease some of those. The Free Zone was inaugurated in April 1990 and is currently including around 60 firms employing around 9000 formal and informal workers. Companies that have been approved include laborintensive industries, leading technology companies, companies making extensive use of local raw materials, enterprises providing subcontracted services for exports and enterprises producing inputs for companies listed above, among others with the main activities being food and chemistry. Togo's Free Zone is atypical compared to other export processing zones as capitalintensive rather than labor intensive industries dominate. In general, companies located in the Free Zone must export all of their productions, but can, subject to authorization, sell up to 20% of their production in Togo. The Free Zone is not a specific geographical area, even though many of them are located close to the Port of Lomé so the benefits are not necessarily connected to infrastructure to the same extent as in other countries. Nevertheless, compared to companies outside the Free Zone, contractors enjoy numerous benefits, including customs benefits and exemptions from various taxes and charges and more flexible labor market freedom (See Box 4). 47 The total amount of taxes or the total tax burden is referring to the sum of all the different taxes and contributions payable after accounting for deductions and exemptions. The taxes withheld (such as sales or value added tax or personal income tax) but not paid by the company are excluded. The taxes included can be divided into 5 categories: profit or corporate income tax, social contributions and labor taxes paid by the employer, property taxes, turnover taxes and other small taxes such as municipal fees and vehicle and fuel taxes. 48 Having said that there might be some questionable taxes such as the payroll tax of 7% in a situation of high un and underemployment, and the fixed minimum rate of 300 USD irrespective of the size of the firm, disproportionally affecting small firms. 50 Box 4: Benefits enjoyed by firms in the Togo Special Export Zone Customs benefits: Exemption from import duties, the VAT and the taxes on equipment required for the instal lation and operation of the company, including office furniture, spare parts, raw materials, semifinished products and consumables; Reduction of 50% of taxes and import duties on vehicles; Exemption from all export duties and taxes for all products that have been imported or manufactured in t he zone. Tax benefits: Tax rate of 0% during the first ten years and 15% from the 11th year after the date of the business accreditation; Reduction by 2 % on taxes on wages over the life of the company; Foreign shareholders are exempt from taxes on divi dends for the first ten years. Financial benefits: Approved companies can transfer capital for investment and commercial transactions to countries outside of the CFA zone; Expatriate employ ees, sharehol ders and nonnational companies can transfer funds to countries outside the CFA zone. Accredited companies can hold accounts in foreign currency. Other benefits: Companies installed in zone have the freedom to set prices, margins and rents on transacti ons between them and foreign markets have preferential tariffs on port services rate have freedom of hiring and di smissal of the Togolese staff or expatriate; can install their own network of telecommunications, satellite stations and radio system; enjoy more favorable rates for post of Togo and TOGOTELECOM; can use a own energy source for their exclusive use and benefit from more favorable rates for energy produced by public services. Note: A new formula to promote the integration with the rest of WAEMU was developed in parallel and became active in 1996. Moreover, in the future, all exports to the subregion will be subject to ECOWAS import duties. Source: Agbodji (2009). Figure 54 shows how the FZ firms' share of exports, but also imports, has increased. In 1992, when the Free Zone had just started the share of imports was 0.6%, and of export 1.1% but the same shares was in 2004 10.3 and 23.9, respectively. If one were to exclude phosphate and "les culture de rente", the Free Zone represent between 50 and 70% of Togo's export. Note also that the firms within the Free Zone demand intermediate goods mainly from nonFZ firms, so even the sectors outside may be positively affected by the zone. Figure 54: Free Zone's Share of Total Imports and Exports, 19912004 (LHS) and Structure of Demand for intermediate goods per industry (%). Ind. Food Ind. Food Ind.Che. Ind. Che. Oth. ind. Oth. ind. nonFZ FZ nonFZ FZ nonFZ FZ Agriculture: Food 62 16 11 0 2 0 Agriculture: Annual Crops 6 6 0 0 10 0 Industry: Food, nonFZ 21 56 0 40 0 0 Industry: Food, FZ 0 0 0 0 0 0 Industry: Chemicals, nonFZ 5 9 10 54 12 28 Industry: Chemicals, FZ 0 0 0 0 0 0 Other Industries, nonFZ 3 9 31 6 68 62 Other Industries, FZ 1 0 0 0 2 0 Services, merchandize 3 5 48 0 6 10 Total 100 100 100 100 100 100 Note: Demand from sectors presented in the columns of products in the sectors presented in the rows. Source: Agbodji (2009), data from SAZOF and MCS, Togo.. Looking at data between 1995 and 2002, the average labor productivity within the Free Zone was lower than the average outside, while average total productivity was higher within the Free Zone than outside. This suggests that low labor productivity was more than compensated for by high productivity of capital. Related to this, the average salary within the Free Zone is two to four times lower than in domestic firms outside the zone. However, this is mainly explained by the fact that the type of firms within the Free Zone demands lowqualified workers. In general, as seen in Table 15 the firms in the Free Zone are more capital intensive, but have a higher share of formal employment than the firms outside the zone (the Free Zone employs about a third of the total number of formal workers in Togo). (Agbodji, 2009) Hence, the relative significance of the firms in the Free Zone is increasing ­ both in terms of export shares and productivity which begs the question what constraints these firms are able to avoid. 51 Table 15: Factor Intensity per Sector (%). Sector Formal Informal Private Public workers workers capital capital Agriculture: Food 0,07 24,48 75,45 Agriculture: Annual Crops 30,90 69,10 Industry: Food, nonFZ 6,80 67,70 25,49 Industry: Food, FZ 53,15 32,43 14,41 Industry: Chemicals, nonFZ 2,98 43,01 54,01 Industry: Chemicals, FZ 11,99 7,59 80,41 Other Industries, nonFZ 16,11 71,69 12,19 Other Industries, FZ 68,70 ,18 31,12 Services, merchandize 15,02 81,49 3,49 Services, nonmerchandize 70,32 29,68 Total 17,92 48,94 28,93 4,31 Source: Agbodji (2009), data from MCS, Togo. The ICA data seems to suggest that the main reason for firms to join the Free Zone is the fiscal and administrative advantage. ICA data in Figure 55 shows that firms in the Free Zone tend to first of all have concerns in fewer areas than nonFZ firms. FZ and nonFZ firms are more or less equally concerned about issues of access to finance and political instability, but nonFZ firms are concerned about for example corruption, tax administration, custom and trade regulation, and business licensing while these are hardly mentioned as the main constraint by FZ firms. Given the higher growth of firms within the zone, this suggests that easier administrative procedures for running a business are binding constraints for many firms in Togo. Note however that the Togo ICA (2010) looks at the number of firms identifying a category as a main as well as a severe constraint and then compares the result between FZ and nonFZ firms and conclude that complaints about business related constraints are not significantly different between the two groups. Given a number of differences when including both the main and severe constraint, the result should be taken with caution. Moreover, in the survey it was never specified if the FZ are responding having their particular investment climate in mind, or the general investment climate in Togo. Figure 55: Main Constraints to Business, FZ firms vs nonFZ firms, Togo, 2009. 35% FZs 30% Non FZs 25% 20% 15% 10% 5% 0% Source: ICA data, World Bank. 6.9 Market Failures Undertaking profitable economic activities requires the availability of inputs specifically needed for such activities, but these inputs may not be available in a market despite a strong demand (current or potential). These are tradable inputs, such as raw materials, equipment and other capital goods; and nontradable inputs, such as skilled workers, able entrepreneurs, infrastructure services and other services. In most of the cases, the provision of tradable inputs implies less of a problem for firms, as long as there is a good enough infrastructure (roads, ports, etc.) to make available these goods from far away when they are not locally produced. On the other hand, the provision of nontradable services represents more of a problem as it may take long time for a market for these services to clear if at all in response to a shortage of an input. In particular, markets may not even send signals for scarcity of certain production factors if there is no perceived demand for them, as the activities that 52 would actually require them in their production process never get to come to life, precisely due to the lack of these factors. It is the typical chicken and egg vicious circle, as explained by Hausmann et al. (2008). Two categories of market failures are identified within this type of problems: Coordination failures and self discovery. Coordination problems arise when inputs and activities that are necessary to undertake other economic activities are underprovided or not provided at all, precisely because there is no demand from those yettoexist activities, or they have the character of public goods and there are hence limited incentives for an individual firm to start those activities. Self discovery problems arise from lack of information about what goods or services, not yet locally provided, can be profitably produced; this is so because the process of "discovering" new goods entails initial costs for entrepreneurs, that may not be able to be recovered, because of high uncertainty on demand, or because the possibility that new entrants in the market (after the product has been found to be profitable) will, by means of competition, wipe away potential excess benefits for the entrepreneur (Hausmann and Rodrik, 2004). Thus way, new profitable products will be underprovided exante, and overprovided (in a sense that free entry will eliminate excess profits) expost. Both, coordination failures and self discovery problems are said to justify some sort of public policy. 6.9.1 SelfDiscovery The task of identifying potential constraints to growth from lack of selfdiscovery is a most difficult one, especially as it involves assessing activities that are not in the country's current production sphere. However, as Hausmann et al. (2008) explain, one can think of a series of tests applicable to actual production data which refer to the degree of sophistication of existing activities; their relative position compared to products closer to the "production frontier"; and the number of new, successful activities. These tests permit to check whether, expost, things are taking place in the country's innovation arena. If innovation is indeed happening as a widespread phenomenon in the economy, then market failures associated with lack of selfdiscovery may not be a major problem. This is, low exante entrepreneurship (before discovery time) and excess expost entry into the new market (hypothesized under laissez faire conditions) are not a binding constraint in the country. A practical problem that arises in attempting to make these tests is the lack of highly disaggregated production data necessary to pinpoint activities by degree of sophistication and discoveries. That is why it has been suggested (Klinger and Lederman, 2006) the use of export data instead. Having in mind these caveats49, a battery of export indicators for diversification, sophistication, proximity to technological frontier and discovery, as defined in works by Hausmann et al. (2005) and Hausmann and Klinger (2006), and Klinger and Lederman (2006), can be used to assess whether new higher quality activities are taking place within a country. We have in Section 4 already defined proxies for the degree of sophistication of products (PRODY) and the export basket (EXPY) and concluded that Togo exports lowincome products but is improving. An additional indicator refers to the "density" of exported products and their position in the "open forest". Products that are "closer"50 to other higher income value products (as defined by PRODY) are said to be in a "denser" part of the product space, implying that the ability of the country to "jump" (as one measure of selfdiscovery) to these higher income value products is actually higher. 49 COMTRADE provides a good source of information, where goods can be disaggregated in thousands of subcategories, 49 depending on the used nomenclature . An additional advantage of using export data, as opposed to production data is that exporting can be seen as a sort of demonstration that a firm was able to pass a "quality" test in the international market. Being purchased in the world market, a locally produced and exported good can thus be regarded as one produced under satisfactory minimum standards, with entrepreneurs being able to break into a more competitive stage. On the flipside, using exports data omits considering the possibility that the goods were actually successfully produced within the country with good quality standards for already a long time and that the inability to bring them to the international market may be linked to some form of obstacle or intervention. It also restricts the analysis to tradable goods and is silent in the ­equally important service area, including the touristic, and information technology sectors. 50 "Closeness" between product groups A and B refers to the probability that a country exporting product group "A" will also export product group "B". In turn, this probability is derived from the export record, by product group, from the rest of the world. It is the minimum of the pairs of conditional probabilities going in both directions as an inverse measure of distance: min{P(A|B),P(B|A)}. See Hausmann and Klinger (2006) 53 The evidence for Togo shows that products who did not have "Relative Comparative Advantages" (RCA) in the early 1980s but were located in the "denser" part of the product space (were close to many other products with higher income value) did "jump" to become products with RCA in the early 2000s. This in contrast to products that were more isolated in the product space (see Figure 56). Figure 56: Histogram of Density for Jumps to New Products (i.e. products with RCA in 2000/04 but not in 1980/84) and No Jumps Products (i.e. no RCA in either period), Togo. 90% 80% 70% 60% Relative Frequency 50% 40% 30% 20% 10% 0% -6 -5.5 -5 -4.5 -4 -3.5 -3 -2.5 -2 -1.5 ln(density) No RCA in 80-84 and RCA in 00-04 No RCA in 80-84 nor in 00-04 Source: Authors calculations, based on WITS, COMTRADE data. Moreover, between 19801984 and 20002004, Togo's export basket significantly approached or became surrounded by groups of products of implicit high income value, positioning the country in a much better position to jump to a basket of products with even higher level of sophistication. In other words, using Hausmann et al (2005) terminology, the "option value" of the countries unexploited opportunities has increased as the socalled "open forest" for new products has become denser, more attainable. This circumstance provides a key element in the analysis of the role of the export diversification path for growth in Togo. As the empirical records show, the increase of export sophistication (measured by positive changes in EXPY) is affected by the opportunities provided by the current productive structure, and changes in EXPY are in turn positively associated with growth. Togo has discovered quite a significant number of new exportable goods already. We define discovery in a similar (but not identical) way as proposed by Klinger and Lederman (2006)51. We use three moving windows to identify whether or not a country was exported by period. If a product was exported in two successive periods (windows), we say the product is an "established" one. If it was not exported in an initial period and was exported in the next period, we call it a "discovery". If a discovered product in the second window was also exported in the third window, the country is added to the "established" category. If a product was discovered or established in one period but not exported in the next period we call it a "disappearance".52 Togo has been able to discover a total of 134 new product groups between 1980 and 2006. It departed from just 43 "established" product groups 1983 1985 up to 173 in 20042006.53 51 Innovation as defined by Klinger and Lederman (2006) inside the technological frontier, this is, the one that refers to the successful production within a country of goods or services already discovered and produced elsewhere, as opposed to innovation "at the frontier" which involves the generation of new products and services based on R&D. 52 We loosely define the time windows to be equal to 3 years each. We consider a product to be "exported" during a period if the two following criteria are met: The product is exported in most of the window (at least 2 out of 3 years in the window) and the average value of exports during the period exceeds 10,000$ per year. This definition more or less resembles that of Klinger and Lederman (2006). In their work, an initial period is used "to confirm that the good was never before exported (1994­96); a window during which time the discovery can emerge (1997­2002); and then a final period when the discovery is confirmed to be an established export (exported for at least $10,000 in both 2002 and 2003)" 53 This compares with a maximum number of 266 categories in the SITC at the 3 digit level. 54 Figure 57: Discoveries and disappearing products in Togo, 19852006 25 20 15 10 5 0 5 10 15 20 Total discoveries Total disappearances Total net discoveries Note: Based on SITC 3 digit data Source: Authors calculations based on COMTRADE data. An alternative definition of "emerging products", confirming Togo's innovation activities, can be derived from the idea of RCA, already referred to. About 20% of Togo's exports comprising 31 product groups from SITC at 3 digit level have been identified only in 20002006 (see Table A 2 in Appendix I). Exports of this group have increased 14.5% between the periods 199296 to 200206. The implied income value of this export basket was 30% above the EXPY for the traditional or "classic" products group. Despite these positive developments, we cannot go as far as to say that Togo is steadily moving to close the gap with the technology frontier. The contribution of discoveries to exports remain relatively low and Togo's growth prospects are highly dependent on its traditional products as judged by the strong correlation observed between the growth rate of real GDP and the evolution of its Terms of Trade explained above. And, as Figure 12 above shows, the share of Togo on world exports is on a declining trend. On the other hand, we cannot argue that self discovery is the current binding constraint to growth in Togo. 6.9.2 Coordination Failures Turning to the second type of market failures, one would easily suspect coordination activities in Togo to be a constraint as the very existence of successful special economic zones indicates what could happen if those types of constraints are removed. The Free Zone in Togo has, as argued, shown to be successful but the benefits of the zone is not directly related to overcoming coordination failures but rather a package of lower cost and less cumbersome business regulations (see Box 4 above). The zone is not a geographical area with surrounding infrastructure (even though most of the firms are located in Lomé), do not have special middlemen or marketing services, no government intervention in terms of input markets, no research or product development program, no advisory services, no special credit treatment, etc. Moreover, Togo is a relatively densely populated, small country where internal coordination is expected to be less of a constraint compared to other countries but external coordination is more important. This has however been acknowledged by the government that has acted, not least through participation in ECOWAS and WAEMU for regional economic and trade coordination (see Section 5.2). And as mentioned, the West African economic integration efforts are viewed as very successful. However, coordination failures are clearly a constraint in the agricultural sector. The sector is dominated by small holders, technically unsophisticated production focused on risk minimization and selfconsumption. Increased agricultural growth requires a number of support structures, such as road infrastructure, market information mechanisms, research, advisory support, etc. Previous government attempts to provide the farmers with this support structure did not succeed, but this is a symptom of government failures rather than a lack of impact from well implemented market coordination effort. For example, the cotton production increased steadily up until 2003 when it dropped sharply, partly due to a drop in the international price of cotton but also due to mismanagement of the national cotton company SOTOCO. Without an alternative coordination function, cotton production has still not recovered and the problems have spread to other crops that are cocultivated with cotton. Another example is the coffee/cacao sector that in terms of downstream impact (marketing, packaging, export, etc) profited from the 55 liberalization in 1996 and saw increased returns for planters. However, production has stagnated due to the inefficient institutions established to provide technical extension services, cutting of 60 percent of the planters from crucial advisory services. Alternative institutions to support the agricultural sector is yet to be developed and those are expected to be a mix of public and private initiatives (see further a more detailed discussion in the agricultural section of the Togo Country Economic Memoranda, 2010). 7 Conclusions As is the case in lowincome countries, many issues appear to be constraining economic activity in Togo. However, a first step is to exclude those potential constraints that do not show signs of being binding at this moment. Among those are 1) human capital (however, it may need to be addressed today for future needs), 2) geography, 3) macroeconomic stability, and 4) selfdiscovery. It is noteworthy that these conclusions are based on the given level of stocks and flows at the time of the analysis, and any subsequent decrease in these levels may change their status as binding constraints. A second step is to determine if the constraint appears to be on the cost or return side of investments and entrepreneurship, which in the case of Togo, evidence indicates problems related to returns of investments despite a continued need to improve access to capital and financial intermediation. Infrastructure measures give mixed signals in terms of scarcity at his moment in time. For instance, few firms complain about overall infrastructure and the supply in relation to countries at similar development levels is not exceptional. However, much of the Togo infrastructure was destroyed in recent flooding to the detriment of the greater economy.. The mixed signals are also indicative of the fact that some types of infrastructure present greater and different challenges than others. Moreover, infrastructure as a constraint must be understood in the context of Togo's growth strategy as a regional trade hub in addition to its goal of broadbased growth and the need to connect firms outside Lomé. Taken together, Togo must ensure that the increased focus on infrastructure remains as donors reengage and support public investments. Access to credit for the private sector is yet another potential constraint that gives varied signals. Credit emerges as a severe and highranked constraint to business according perception surveys at an enterprise level; however, this is the case in most countries surveyed. To this end, it is unclear if the low amount of credit is a result of lack of demand (i.e. given the level of economic opportunities today, few firms are prepared to take a loan and make risky investments), or a lack of supply mainly in terms of financial intermediation now that international finance has returned to Togo. In this sense, the possibility of financial intermediation as a potential binding constraint is ­ in a similar fashion to infrastructure ­ already being addressed through several measures. At this point, it is crucial to, at the very least, implement the necessary banking sector reforms and continue to pay close attention to access to credit as a constraint. Improvements will have to be continuously realized as economic opportunities and demand for credit arises. A prominent area in terms of constraint to growth in Togo is governance issues, in particular political instability, corruption, and governance efficiency. In nearly every measure of governance, Togo's performance ranks last when compared to neighboring countries with very little documented progress to date. Political stability has improved since the latter part of the 1990s and more prudent macroeconomic reforms and policies have been implemented. However, if and how these improvements translate into increased trust among investors is debatable according to perception data from domestic investors and the limited FDIs recorded. Not least, the low governance efficiency in Togo has been traced to mismanaged SOEs and lack of public investment execution. As with several of the measures, this may be more connected to previous governments and whether more sustainable and thorough reforms happen in the future remain to be seen. Governance issues are also related to investor trust in businessgovernment relations, which according to the analysis is an obvious binding constraint Togo needs to address. Foremost, barriers and costs of doing business are high, especially when starting a business, and compare poorly to countries at the same development level. Even though the overarching issue of doing business is critical to tackle, priorities are important. For instance, tax 56 administration is an area that needs continued attention while labor market regulations seem to be less of a problem. Trading across borders is a relative strength compared to other countries and other constraints, however if Togo is able to leverage this strength, with additional improvements, into a role as a regional hub remains to be seen. Moreover, security of property rights does not seem to be binding today but building trust in this area will be crucial for future growth. As an overall indicator of business costs being a constraint in Togo, "informal practices of competitors" ­ i.e. those not following the formal government rules for businesses is mentioned as one of the top constraints by firms. Hence, decreasing doing business costs as well as taking measures to ensure that key reforms are trusted by investors emerge as priority when address binding constraints to growth in Togo. Special attention needs to be given to the agricultural sector because it has a substantially higher output and productivity potential going forward, and is still the most important sector in terms of GDP and employment. The previous institutional structure no longer exists due to several government failures, leaving the sector with coordination issues and lack of infrastructure. The replacement structure is developing slowly at best, but does grant more influence to farmers and the private sector in the organization than in the provision of services, which hopefully will avoid continued government failures in the sector. Looking ahead, for increased international competitiveness and rapid growth, Togo will need to ramp up efforts to streamline its costly and cumbersome business procedures. However, for this to have a sizable impact, Togo must prove to potential investors that political stability is permanent and that corruption, poor budget execution and mismanagement of SOEs belongs to the past. As a result of the new government's reform agenda and the return of international aid, a window of opportunities for high returns to the still limited public investments has opened up. This is especially true in infrastructure and connectivity services, which would not only take advantage of Togo's geographical location as a regional hub but also make growth in Togo more inclusive. Also, as the economic opportunities arise for the private sector, there is a need to restructure the banking sector, which has already started, to smooth distortions in the credit market. Promising sectors within agriculture that are vital to economic growth, employment opportunities and poverty reduction remain important, but will need to overcome a number of coordination failures. Not least due to the history of government interventions causing economic distortions, the government must allow for a stronger role for private operators and encourage it wherever possible. Finally, even though education does not exhibit constraints to economic activity in Togo today it is of importance to improve the quality of education, not least to profit from and catalyze the opportunities related to Togo's potential as a regional hub. Togo appears to be slowly but surely diversifying away from traditional business, and in the process discovering new products that have been successfully placed in the international market. These new products present a path to growth for Togo, so the government should pay special attention to issues confronting these products as a guide to reform. Case in point is the success shown in the Free Zone and the positive incentives it created should be encouraged and potentially be applied to firms outside the zone. However, there is a need to look at the cost side of the success of the Free Zone, and to understand the puzzle of why mostly capitalintensive industries seem to be able to profit from the advantages within the zone, despite Togo's relatively low wages. 57 8 References Assiongbon, E. and J.P. Chausse (2008). Togo: Croissance Agricole et Reduction de la Pauvrete. Draft, November 8, 2008. Azariadis, C. (1996). The Economics of Poverty Traps ­ Part One: Complete Markets. Journal of Economic Growth, 1: 449486. CWIQ (2006). Core Welfare Indicators Questionnaire (CWIQ)/ Questionnaire des Indicateurs de Base du Bienêtre (QUIBB), Direction Générale de la Statistique et de la Comptabilité Nationale, Republique Togolaise 2006. Gogue, T. A. and K. Evlo (2004). Explaining African Economic Growth Performance: Togo Case Study, Mimeo, Universite de Lome. Lome, Togo. Gonzales and Rosenberg (2006). The State of Microfinance ­ Outreach, Profitability, and Poverty: Findings from a database of 2600 microfinance institutions, CGAP. Memo. Hausmann R. and B. Klinger (2006). "Structural Transformation and Patterns of Comparative Advantage in the Product Space" Working Paper No. 128, Kennedy School of Government, Harvard University, Cambridge, MA. Hausmann, R. and B. Klinger (2007). "The Structure of the Product Space and the Evolution of Comparative Advantage", CID Working Paper No 146, Kennedy School of Government, Harvard University, Cambridge, MA. Hausmann, R., B. Klinger and R. Wagner (2008). "Doing Growth Diagnostics in Practice: A `Mindbook' ", CID Working Paper No 177, Kennedy School of Government, Harvard University, Cambridge, MA. Hausmann, R., D. Rodrik and A. Velasco (2005). Growth Diagnostics. Kennedy School of Government, Harvard University, Cambridge, MA. March 2005. IMF (2009a). Togo: Staff Report for the 2009 Article IV Consultation and Second Review of the Poverty Reduction and Growth Facility Arrangement. International Monetary Fund: Washington, DC. IMF (2009b). Togo Selected Issues. International Monetary Fund: Washington, DC. Kuepie, Nordman and Roubaud (2008). Education and Earnings in Urban West Africa. Journal of Comparative Economics, September 2008. PRSP (2009). DSRP Complet 20092011, Document complet de Strategie de Reduction de la Pauvrete, Version Preliminaire, Mars 2009. Republique Togolaise. Psacharopoulos (1994). Returns to investment in education: A global update. World Development, Volume 22, Issue 9, p. 1325 1343. Rodrik, D. (2004a) "Why we learn nothing from regressing economic growth on policies", Narch 2005, Cambridge MA: Harvard University, pp. 114. Rodrik, D. (2004b) "Industrial Policy for the Twentyfirst Century," mimeo. Togo Sources of Growth (2008). Togo: Les Sources de Croissance er le Climat d'Investissement. Working Document for the preparation of the Poverty Reduction Strategy Paper, Version Preliminaire, November 17. World Bank (2007a). Profil de la pauvreté et de la vulnérabilité au Togo, 2007 World Bank (2007b). The Role of Employment and Labor Income in Shared Growth: What to Look for and How, PREM Poverty Reduction Group, World Bank, Draft December 2007. World Bank (2008). Togo: Enhanced Heavily Indebted Poor Counties (HIPC) Initiative Decision Point Document, November 2008, International Development Association (IDA) and International Monetery Fund (IMF), Washington, DC. World Bank (2009). Togo: Mémorandum Economique du Pays sur les Sources de Croissance et Etude Diagnostique sur l'Intégration Commerciale (EDIC) du Cadre Intégré pour l'Assistance Technique Liée au Commerce. Mission Principale du 18 mai au 9 juin, 2009, Aide Mémoire. World Bank (2010). Togo Investment Climate Policy Note. Draft Report. Finance and Private Sector Development, Africa Region. World Bank, Washington, DC. 58 9 Appendices Appendix 1: Additional Figures, Tables and Boxes Table A 1: Togo Income Benchmarking g , , , Years to achieve pc growth of countries at historic growth Historical Growth rate rate: PC Growth rate 2007 GDP required to reach per capita 2007 GDPpc of (real US $ 1980 2000 2006 1980 2000 2006 resp. countries in Country OF 2000) 2007 2007 2007 2007 2007 2007 20 years Togo 239 1.4% 0.4% 0.5% Burkina Faso 260 1.8% 2.4% 1.0% 6 21 16 0.4% Benin 328 0.4% 0.7% 1.5% 23 78 58 1.6% Senegal 509 0.4% 1.6% 1.9% 55 187 140 3.8% Nigeria 472 0.2% 3.6% 4.0% 50 168 126 3.4% Ghana 307 1.0% 3.1% 4.2% 18 62 46 1.2% Mauritius 4700 4.1% 3.2% 3.9% 219 736 550 14.9% LICs 415 1.2% 3.2% 4.3% 40 136 102 2.8% SubSaharan Africa 598 0.0% 2.4% 3.7% 67 227 169 4.6% MICs 1997 2.6% 4.9% 6.9% 156 525 392 10.6% Source: Authors' calculations. DDP Data, World Bank. Box A 1: Expected Transmission Channels of the Global Crisis to Togo.1 The global crisis is having or may have the following spillover effects on Togo: Lower demand and prices for exp orts: Lo wer global growth may adversely impact demand for traditional commodity exports, such as cotton and phosphate. Their s hare in total exports declined from 50% at the end of the 1990s to just 13% in 2008 and could fall further with the crisis. Cement exports, which represent about 20% of total expo rts, are expected to continue to increase, but declining regional demand should no t be ruled out. Decreasing tra derelated activities: The activity of the Port of Lomé and the related trans port sector may suffer fro m lower trans it and reexport of good s to Burkina Faso, Niger, and Mali. Early in 2009 there was no sign of declining activity, but there may be a delayed impact. Lower remittances: If eco nomic con ditions wors en in host countries, particularly in continental Europe, remittan ces, which make up 10 to 15% of GDP, could drop off sharply and jeopardize domestic demand (which is particularly important for construction). As of December 20 08, n o dropoff was eviden t, however. Lack of interest of strategic inves tors: To go is no t highly dependent on private capital flows, but delays in attracting strategic investo rs , which are key to finalizing restructuring of the banking and phosphate sectors, would constitute lost opportunities to en hance growth. For example, low international demand for phosphate may delay investor interes t, which is crucial to obtain fresh capital and raise output. Lower aid flows: Togo is not heavily depen dent on donors flows--its fore ignfinanced investment and so cial spen ding are among the lowest in the WAEMU--but it is expecting an increase to supp ort economic revival. Declining fiscal revenues: A greater than exp ected slowdown would likely redu ce revenues below projections. The resulting financing gap would pose policy challenges. Togo's exposure to the crisis may, however, be mitigated by the following countryspecific fa ctors: Low interna tional exposure of the fin ancial s ector: The banking sector is mostly publicly and regionally owned, with low exposure to the most affected markets. Non eth eless, a recent pickup in hous ehold credit could reverse, if regionally owned private banks--the most active in this market--cu t back lending in Togo to rebalance their portfolio s. While there is no evidence so far of disruptions in international trade finance or the interbank marke t, banks are seeing more strict requirements from correspondent banks. Microfinance institu tions, which represent 15% o f total credits, have experienced d ecreasing funding from international institutions, however. Second round effe cts (e.g., rising N PLs) are als o expected to b e moderate given the recent res tructuring, but the autho rities are committed to heightened overs ight to minimize the risks. N o major decrea se in co mmitted dono r suppo rt: Pub lic investment in 2009, financed in large part by dono r su pport, is expected to pick up du e to new commitments. Togo is largely reliant on multilateral donors--the W orld Bank, the EU , and the AfDB--who se disbursement is mo re stable and which togeth er have committed budget s upport of abo ut 2 % of GDP to help Togo meet its 2009 fin ancing needs. Subs is tence ag riculture th at accoun ts for about 25% of GDP might be an important shock abso rb er. Staff projects a rebound in subsistence produ ction following the floods in 2008. Declin ing energy and food prices are expected to reduce inflationary pressures, improve the terms o f trad e, and narrow the current account deficit in 2009. ___________________ 1 For furthe r details see Selected Issues Paper. 59 Table A 2: Groups of Products with Revealed Comparative Advantage in Togo, 199296 vs 200206. "The Marginals" (with export shares>0.1%, each) No RCA in either 1990-1996 or 2000-2006 "The Classics" RCA in both 1990-1996 & 2000-2006 Classification: SITC2-3 Digits Growth 2000-06 / 1990-96 Share 00/06 Classification: SITC2-3 Digits Growth 2000-06 / 1990-96 Share 00/06 Code Product Name Togo World in Togo Code Product Name Togo World in Togo 11 Meat,edible meat offals, fresh, chi 46.7% 4.2% 0.19% 35 Fish,dried,salted or in brine ; smo 14.3% 1.9% 0.5% 57 Fruit & nuts(not includ. oil nuts), 4.4% 4.7% 0.42% 36 Crustaceans and molluscs,fresh,chil 7.5% 3.2% 0.4% 98 Edible products and preparations n. 19.1% 8.1% 0.28% 71 Coffee and coffee substitutes 6.0% 1.2% 1.9% 112 Alcoholic beverages 21.7% 6.0% 0.20% 72 Cocoa 19.2% 5.3% 7.2% 522 Inorganic chemical elements,oxides 41.8% 6.5% 0.12% 75 Spices 10.0% 5.7% 0.1% 553 Perfumery,cosmetics and toilet prep 31.1% 9.9% 0.13% 81 Feed.stuff for animals(not incl.unm 4.2% 4.1% 1.1% 634 Veneers,plywood,improved or reconst 39.4% 4.5% 0.11% 222 Oil seeds and oleaginous fruit,whol 23.4% 5.5% 1.6% 635 Wood manufactures,n.e.s. 8.2% 7.7% 0.11% 223 Oils seeds and oleaginous fruit, wh 5.8% 4.7% 0.1% 642 Paper and paperboard,cut to size or 35.3% 6.7% 0.16% 263 Cotton 1.3% 0.5% 14.1% 653 Fabrics,woven,of manmade fibres 38.9% 1.5% 0.30% 271 Fertilizers,crude 4.2% 1.0% 11.4% 657 Special textile fabrics and related 19.1% 5.8% 0.10% 292 Crude vegetable materials, n.e.s. 1.0% 3.9% 0.3% 674 Universals,plates and sheets,of iro 23.1% 7.4% 0.27% 334 Petroleum products,refined 27.0% 10.6% 18.1% 678 Tubes,pipes and fittings,of iron or 29.8% 6.6% 0.17% 424 Other fixed vegetable oils,fluid or 13.5% 7.3% 1.2% 697 Household equipment of base metal,n 31.1% 7.3% 0.20% 661 Lime,cement,and fabricated construc 39.5% 6.2% 10.3% 723 Civil engineering & contractors pla 55.3% 8.8% 0.12% 941 Animals,live,n.e.s.,incl. zooanima 3.1% 5.4% 0.2% 778 Electrical machinery and apparatus, 31.6% 8.4% 0.23% 781 Passenger motor cars,for transport 43.9% 7.9% 2.93% Total ------> 68.4% 782 Motor vehicles for transport of goo 33.2% 7.3% 0.85% 785 Motorcycles,motor scooters,invalid 48.6% 6.0% 0.21% 786 Trailers & other vehicles,not motor 49.9% 8.9% 0.15% "Emerging Products" No RCA in 1990-1996 & RCA in 2000-2006 821 Furniture and parts thereof 15.5% 9.2% 0.11% Classification: SITC2-3 Digits Growth 2000-06 / 1990-96 Share 00/06 845 Outer garments and other articles,k 39.5% 6.7% 0.16% Code Product Name Togo World in Togo 846 Under garments,knitted or crocheted 28.3% 7.8% 0.15% 22 Milk and cream 37.9% 4.9% 0.4% 851 Footwear 19.8% 4.6% 0.21% 23 Butter 79.8% 3.2% 0.1% 893 Articles of materials described in 13.2% 8.0% 0.21% 34 Fish,fresh (live or dead),chilled o 21.5% 4.8% 1.1% 971 Gold,nonmonetary 7.6% 5.1% 0.25% 42 Rice 33.0% 4.7% 0.4% 46 Meal and flour of wheat and flour o 40.7% 2.6% 0.5% Share of Marginals (each with an export share>0.1%) on total exports -----> 8.4% 54 Vegetab.,fresh,chilled,frozen/pres. 7.4% 4.8% 0.4% 56 Vegetab.,roots & tubers,prepared/pr 10.5% 4.4% 0.1% 58 Fruit,preserved,and fruit preparati 25.2% 4.5% 0.4% Marginal Products with Share in Togo's Overall Exports 2000-2006<0.1% 61 Sugar and honey 33.6% 4.1% 0.3% Classification: SITC2-3 Digits Growth 2000-06 / 1990-96 Share 00/06 62 Sugar confectionery and other sugar 32.6% 7.0% 0.1% Togo World in Togo 91 Margarine and shortening 46.6% 10.0% 0.5% 151 groups of products, each with share in Togo's 111 Non alcoholic beverages,n.e.s. 41.2% 10.1% 0.2% 2.80% total exports <0.1% during 20002006 247 Other wood in the rough or roughly 35.8% 0.5% 1.7% 248 Wood,simply worked,and railway slee 39.1% 3.1% 0.4% Total share of the marginals ---------------------------------------------------------------------> 11.2% 269 Old clothing and other old textile 37.8% 5.8% 0.2% 273 Stone,sand and gravel 45.6% 5.2% 0.3% 282 Waste and scrap metal of iron or st 32.4% 9.8% 1.0% "Disappearances" RCA in 1992-1996 & No RCA in 2000-2006 288 Nonferrous base metal waste and sc 17.4% 6.8% 0.4% Classification: SITC2-3 Digits Growth 2000-06 / 1990-96 Share 00/06 335 Residual petroleum products,nes.& r 62.2% 10.2% 0.2% Code Product Name Togo World in Togo 423 Fixed vegetable oils,soft,crude,ref 91.2% 6.1% 0.2% 245 Fuel wood (excluding wood waste) an 23.9% 7.7% 0.0% 554 Soap,cleansing and polishing prepar 40.4% 8.4% 0.3% 667 Pearls,precious& semiprec.stones,u 11.5% 9.4% 0.5% 562 Fertilizers,manufactured 17.8% 5.1% 0.6% 844 Under garments of textile fabrics 18.8% 2.0% 0.0% 583 Polymerization and copolymerization 47.9% 8.3% 2.7% 896 Works of art,collectors pieces & an 2.5% 4.3% 0.1% 591 Disinfectants,insecticides,fungicid 51.1% 6.1% 1.2% 652 Cotton fabrics,woven 43.1% 3.0% 2.9% Total share of disappearances -----------------------------------------------------------------> 0.6% 656 Tulle,lace,embroidery,ribbons,& oth 41.7% 5.2% 0.1% 673 Iron and steel bars,rods,angles,sha 51.1% 6.1% 1.6% 677 Iron/steel wire,wheth/not coated,bu 51.9% 7.3% 0.1% 721 Agricultural machinery and parts 74.1% 6.2% 0.3% 783 Road motor vehicles,n.e.s. 59.9% 7.7% 0.6% 899 Other miscellaneous manufactured ar 15.4% 8.4% 0.5% Total ------> 19.8% Source: PRMED staff calculations based on UN/COMTRADE data. Table A4. Mincerian regression of income on education and a set of controls Mincerian regressions of income per individual, Togo 2006 Number obs. 1507 Prob F>0 0 Dependent variable: Individual income, in logs Rsquared 0.37 Coefficient tstatistic Years of Education 0.05 5.24 *** Age 0.03 6.01 *** Age, Squared 0.00 4.51 *** Male (Dummy=1) 0.61 20.01 *** Married (dummy=1) 0.05 2.51 *** Literate (dummy=1) 0.03 0.10 Default is: Education type, General 0.08 0.09 Professional Education type, Technical 0.13 0.11 Private Employee (dummy=1) 0.10 1.87 * Salaried worker 0.07 3.01 *** default sector is: Primary Sector 0.14 0.18 Tertiary Secondary Sector 0.12 0.04 constant 2.32 17.08 *** *** = Significant at 1% ; ** = Significant at 5%; * = Significant at 10% Source: Authors' calculations based on Core Welfare Survey, 2006. 60 Appendix 2: Export diversification and sophistication indices HerfindhalHirschmann index of export diversification This indicator is calculated for country "i" as the sum of squared values of product "j" shares in total country export (Xsh), adjusted by the number of products exported by the country in year "t", as follows: Xshi ,t 2 j HH i ,t 1 1 ji ,t ji ,t In this equation, the adjustment of 1 minus the sum of exported products "j" is intended to scale the HH index to the ratio 0 to 1, with zero meaning full diversification and 1 meaning a concentration of exports in just one product. EXPY index of export sophistication The first step is to construct the socalled PRODY's for each category of service exports. "PRODYj" is the income value associated with the export good "j", and is constructed by using the export (x) share of a country (i) in world exports of "j" divided by the sum of shares of "j" in world exports of "j" across all countries exporting the good.54 These ratios are multiplied by the exporting countries' respective per capita income level (Y) and the result is summed up across all countries. In other words, the PRODY becomes the weighted average of per capita GDPs, where the weights represent the revealed comparative advantage in good "j" for each country.55 PRODYs are constructed for each export goods category, for each year of available data, and are by construction the same for all countries. xij Xi PRODY j i Yi xij i Xi EXPY is then the weighted income value of goods exported by a country, computed as the sum of PRODYs using as weights the share of the particular service in the country's total goods export basket. EXPYs are constructed for each country and for each year with available data. xij EXPYi PRODY j j Xi xij Xi is hence the valueadded share of commodity j in the country's overall export basket. 54 55 The rational for using revealed comparative advantages as weights is to control for country size when ranking the services. 61