Policy Research I J7 o | WORKING PAPERS, Macrooconomic Adjustment and Crowth Country Economics Department The World Bank January 1992 WPS 83, Real Overvaluation, Terms of Trade Shocks, and the Cost to Agriculture in Sub-Saharan Africa Ibrahim A. Elbadawi The observed decline of agriculture and the general worsening of economic conditions in Sub-Saharan Africa are linked to economic distortions, which limit growth. Policy Rese arch Working Papers disseninate the findings of work in piogres and encourage the etchange of ides among Bank stuffand allothersinterested indevelopmentissueashesepapers, dstuibutedby t Research Advisory St ffcarrythenamesofthethors, reflect ony thairviews, nd shed beused n. iteds ccordingly net.ndings, interions, o nd conclusionsma the authors'onyshould not be att ibuted to Lhc Wodd Bank, its Boatrd of Directors, its management, or any of its mcmber countries. Policy Research lcrooconotala Adjustment and Growth WPS 831 This paper - a product of the Macroeconomic Adjustment and Growth Division, Country Economics Department - is part of a larger effort in the Department to study the linkages between agriculture and macroeconomic policy in Sub-Saharan Africa. Copies of the paperare available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Victoria Barthelmes, room N 11-025, extension 39175 (66 pages). January 1992. Starting from the premise that agriculture should Preliminary analysis of evidence in Sub- be pivotal in the structural transformation and Saharan Africa links the observed declines in economic development of Sub-Saharan Africa, agriculture and the general worsening of eco- Elbadawi addresses two reiated issues. nomic conditions to economic distortions. A more rigorous analysis, using data from the The first issue is the extent to which policy- Sudan - an African country with a sizable induced distortions influence the structure of agricultural economy - strongly supports the incentives for agriculture (with direct distortions predicdons of Easterly's endogenous growth induced by policies aimed directly at agriculture model (1990), which posits the deleterious distinguished from indirect policies aimed at the effects of economic distortions on growth. economy's macroeconomic management). The second issue is how these distortions affect agriculture's growth, given other growth fundamentals. The Policy Research Working Paper Series disseminate the findings of work under way in the Bank. An objective of the series is to get these findings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official Bank policy. Produced by the Policy Research Dissemination Center Table of Contents 1. INTRODUCTION . ..................................................... 1 1.1 Government Intervention and Agriculture in Sub-Saharan Africa: An Overview . .................................................. 4 2. THE EQUILIBRIUM RER AND m DETERMINANTS ....................... 13 2.1 The Model ................. ..................................... 14 2.2 An Error-Correction Equation for the RER ............................ 20 3. AN EMPIRICAL REAL EXCHANGE RATE MODEL FOR THE SUDAN ......... 23 3.1 The Econometric Estimation ........................................ 25 4. REAL OVERVALUATION AND THE TAXATION OF AGRICULTURE IN SUDAN ........................................................... 38 4.1 Equilibrium RER and Overvaluation .................................. 38 4.2 Direct and Indirect Taxation of Agriculture ............................. 40 5. AN EMPIRICAL MODEL OF ENDOGENOUS AGRICULTURAL GROWTH IN SUDAN . .......................................................... 48 6. CONCLUSIONS . ...................................................... 52 REFERENCES ........................................................... 55 APPENDIX . ............................................................ 58 List of Tables Table 1.1 Average Ratios of Producers Prices to Border Prices for selected African countries (converted at nominal exchange rages), 1970-86. ..................... 10 Table 1.2 Selected Macroeconomic Indicators for Sub-Saharan Africa .................. 11 Table 1.3 Economic Structure and Performance in Sub-Saharan Africa ................ 12 Table 3.1 Key Indicators of the Agriculture Sector Role in the Sudan Economy .... ...... 30 Table 3.2 Sectoral Distribution of GDP, Exports and Labor in the Sudan .... ........... 31 Table 3.3 Sudan's Real Exchange and its Determinants ............................. 32 Table 3.4 Rates of change in some of Sudan's key macroeconomic indicators (percent) ..... 33 Table 3.5 Variance - Ratio Statistics of K-Differences of the RER and its Fundamentals (equation 3.12') ......................... ............................ 34 Table 4.1 Equilibrium Real Exchange Rate and Overvaluation in Sudan ................ 44 Table 4.2 Taxation of Sudanese Agriculture ..................................... 45 List of Figures Figure 3.1 Variance-Ratio Statistics of K-Differences of RER and its Fundamentals (equation 3.12') .................................... 35 Figure 3.2 Rate of Change of RER in the Sudan ................................. 37 Figure 4.1 Equilibrium Real Exchange Rate and Overvaluation in Sudan ..... .......... 46 Figure 4.2 Real Overvaluation and Indirect Taxation of Agriculture in Sudan .... ...... 47 *The author would like to thank William Easterly for helpful comments, the usual disclaimer of course, still applies. Tle views expressed here are not necessarily those of the World Bank or affiliated organizations. The author also would like to acknowledge able research assistance from Ayda Kimemia. 1. INXTR'RODUCIION "While an agricultural-led growth strategy entails substantial risks, its success greatly reduces the most onerous risk humans face-the risk of dropping below the poverty line" John Mellor (1990). Over the past few years, the case for agriculture as the engine of growth for the poor countries of the developing World, especially in Sub-Saharan Africa, has gained considerable support as lessons and evidence gained from past experiences and analytical and methodological advances started to influence our understanding of the process.' Citing past and recent evidence obtained from large sets of data from several developing countries, many authors (eg. T. Schultz, Kuznets, Mellor, Lal, and Johnston) have vigorously and persuasively argued for this particular strategy of economic development. This position is su.cinctly represented by the following three arguments advanced by Mellor (1990). First, because of its large initial size and the powerful stimulus it provides to the growth of other sectors, accelerated agricultural growth hastens decline in the relative size of agriculture by fostering more rapid growth in the nonagricultural sector. Second, concentration of public investment in agriculture tends to accelerate diversification within agriculture and in the rest of the economy. Third, agricultural growth by raising rural incomes and providing the oipportunity to earn higher incomes, makes the farming of high-risk,low-income, and possibly enviroi ,ntally unsustainable land unnecessary and undesirable. Taking the above argument as a maintained position, this paper will attempt to address two inter-related issues critical to th.; role of agriculture in the context of economic development 'Narain (1965), Krishna (1963), Behrman (1968), Schultz (1978), among others, provide some of the early evidence on the price responsiveness of the third world farmers. A more recent evidence on this as well as on the effects of agricultural technology on productivity and rural income is provided in Eicher and Staatz (1986) for example, in the context of the broader debate regarding the role of agriculture in economic development. - 2 - in Sub-Saharan Africa. The first issue is the extent to which policy2 induced distortions influence the structure of incentives for agriculture in Sub-Saharan Africa; and second, the consequences for the growth of agriculture of these distortions given other relevant growth fundamentals. In my view, these are important issues for the current debate on adjustment and growth in Sub-Saharan Africa because of the large evidence on the extent of government intervention in SSA both directly at the micro and sectcral levels as well as through indirect policies directed at the macroeconomic management of the economy. The above issues are also important in light of the recent findings in the endogenous growth literature which show that such policy induced distortions can have significant and deleterious effects on growth. Economic distortions may be emanating from micro and sectoral policy interventions directly deployed on agriculture. This direct set of distortions should be distinguished from distortions that influence the sector indirectly and are induced by policy directed at the macroeconom,ic management of the economy. Direct distorticns are both relatively easier to measure and their impact on agriculture is well understood. It is now a generally accepted proposition that economy wide distortions, such as the overvaluation of the real exchange rate, can have deleterious effects on tradable sectors of the economy such as agriculture. An assessment of real exchange rate overvaluation can be problematic, however, since it must be measured relative to its unknown equilibrium. The concept of 'Equilibrium RER' is, therefore, critical to this analysis and hence modelling and estimation of this 'unobserved' index will be emphasized in this paper. 2 These are policies that either directly or indirectly affect relative prices and resource allocation - such as tariffs and import quotas, controls on prices and interest rates, taxes, and expansive fiscal and monetary policy. -3- Contrary to the predictions of the standard neo,lassical growth models3, empirical evidence have, shown that growth is significantly related to economic distortions (e.g. Easterly and Wetzel (1989)).4 Formal models of endogenous growthS so1ve for the rate of grov.th as dependent (negatively) on economic distortions along with other more conventional growth furdamentals such as the stocks of physical and human capital. The model predicts .hat small changes in distortions do not affect growth much if the initial conditions are ones of very high or very low distortions, but have a strong effect in between those extremes Optimal saving behavior would make growth sensitive even to low distortions, however, Easterly (1990). Using data from the Sudan - an SSA country with a sizable agricultural economy, we test the implications of the endogenous growth model in this paper. Other factors relevant to economic growth in SSA that are also considered in the empirical model of this paper are the impacts on agricultural growth of exogenous worsening of terms of trade and import compression induced declines in capacity utilization, Ndulu (1990). These two factors which reflect the pote-.;al impact on economic performance in SSA of unfavorable external environment, provide some balance to the mainly 'intemalistic' view of the causes of economic crisis which centered around domestic policy-induced real overvaluation.6 3Given the existence of non-reproducible factors, constant retums to scale, and diminishing returns to each factor, steady-state growth-as predicted by these models - can only take place through exogenous technological change. 'Also these models fail to eAplain the observed nonconverging per capita income growth rates across regions (Easterly (1990)). se.g. Easterly (1990) which also contains a list of recent works in this literature. 6For a discussion of the "interiialist versus externalist" interpretation of the source of economic crisis in Africa see Mkandawire (1989) and references cited therein. - 4 - 1.1 Government Intervention and Agriculture in Sub-Saharan Africa: An Overview Governments' interventiors in agriculture are wide spread and common to all countries developed as well as less developed alike. SSA has not been an exception as can be seen from the record of agricultural sector pricing policy in table (1.1) below. This form of intervention is explicit and operates along the supply curve and is executed through the instruments of forced procurement, administered pricing, export and local taxation, among others. Tne practical motivation for such practice on the part of governments fould be to raise revenue or to keep down the prices of food.7 Such squeeze on agriculture is argued to be a main cause of rural poverty8, and a major obstacle p,-venting agriculture from accelerating thc process of growth and structural transformation.9 The potential deleterious impact on agriculture of the above direct form of government policy, may be reversed, however, by another set of direct interventions which tend to shift tile agricultural supply curve. In this case government interventions throagh the 'shifters of supply' such as agricultural technology, credit, investment, and price stabilization schemes are helpful to agriculture. Indeed in the case of the countries of SSA such role for government is critical for the development of a viable agriculture. Micro and institutional considerations aside, the ability of governments in SSA to undertake such a role depends critically on the prevailing international economic environment and domestic macroeconomic conditions. For example after 1982 a 7There is now a vast body of literature on the underlining political economy of these policies. An interesting analysis for the case of African agriculture is p,ovided in Bates (1981,82), and Commins, Loftchies, and Payne (1986). 8Garcia (1981), for example, found agriculture prices to be a major determinants of rural wages. 'Me arguments regarding the implications of such sector-specific policies for the growth of agriculture are well understood and has occupied a central position in the literature. e.g. Eicher and Staatz (1986) and the literature cited therein. combination of worsening terms of trade, reduced concessional foreign aid, and the required need for adjustment led to a further squeeze on investment (see table 1.2)). More recently it is becoming widely accepted tha. the structure of incentives facing agricu!ture is influenced, as well, by indirect interventions in the form of exchange rate, foreign trade, and other macroeconomic policies. The impact on agriculture, of these policies is 'implicit' or 'indirect' and it operates through the real exchange rate."0 Given the high degree of tradability in agriculture, policies that lead to real overvaluation will create a structure of incentives biased against agriculture vis-a-vis other sectors in the economy, especially the non- traaed service sector. The evidence in the literature suggest that these implicit policy induced effects can be quite strong, in fact they can be so strong to overwhelm possibly favorable di:ect sector-specific agricultural price policies and other direct forms of interventions (e.g. Krueger, Schiff, and Vald6s (1988)). The average performance of the economies of SSA has worsened over the last two decades, and by the turn of the 1980's decade economic conditions in the continent assumed crisis proportions. Real GDP growth rate declined from an annual average of 3.7% in 1970-81 to only 1.4% for 1982-85. With the high and steady rate of population growth in Africa, this translated into substantial declines in the standard of living with per capita income declining at an average rate of 0.9% in 1970-81 and 2.5% in the following period 1982-85. Other concomitant aspects of Africa's economic crisis are reflected by the sharp declines in foreign sector indicators. Thus, between the above two sub-periods the average rate of growth of exports declined from 3.6% per annum to 1.1%. The worsening export performance is closely related to the declining share of '0This strand of the literature is an outgrowth of the Dornbusch (1974) two sectors model of tradables and nontradables. This model was subsequently extended by Sajaastad (1980) to a three sector model of importables, exportables, and home goods. Applications of this model to agriculture include Vald6s (1985,86), Mundlak, Cavallo and Domench (1987), Garcia (1981), Bautista (1985), Elbadawi (1987), Oyejide (1986,87) and Tshibaka (1986), to mention a few. -6- agriculture in the domestic economies of SSA and the expansion of the nontraded service sector between the two periods (see table (1.3)). Similarly, the average ratio of external debt service to exports increased sharply from an average of 9.6% in 1973-81 to average 16.7% for the 1982-85 period (table (1.2)), the stock of external debt io GDP ratio also rose from 39% in 1980 to 69% in 1987. In short the story of economic performance in Africa is summed up by Oyejide (1990), There is very little debate regarding Africa's poor economic performance and the Long-term nature of the decline in living standards, particularly during the 1980's. But controversy continues to surround the issue of which factors are responsible for the crisis." One important interpretation of the causes behind the declining shares of agriculture and the general worsening positions of the economies of SSA as described above, emphasizes the role of domestic economic policies, especially real appreciation and real overvaluation.1' The evidence from Sub-Saharan Africa (see table (1.2)) points to episodes of dramatic real appreciation over the seventies and the first half of the 80s. The real appreciation is clearly related to the high rate of fiscal expansion and the increaced domestic absorption (negative of the resource balance) where both indicators respectively rose by 2.1 and 1.4 percentage points of GDP between the first two periods (table (1.2)). On the face of this expansive macroeconomic policies, SSA experienced severe negative external shock, where the value of this aggregate shock turned from a small but positive average of 0.1 to a negative at -5.3. The terms of trade component of this shock also reflect the same effect, with the TOT index declining from 106.3 in 1973-81 to 91.6 in 1982-85. Also external finance available to SSA dropped by more than 50% over the two periods. Therefore, as the RER is appreciated, the major fundamentals call for an equilibrium depreciation of the RER. Hence, it is likely that the economy wide-structure of t'A frequently cited analysis of the economic crisis of SSA representing this tradition is the World Bank's Berg Report (World Bank (1981)). -7- incentives has been unfavorable for agriculture in most of the economies of SSA during the period considered. Another evidence in support of this view is the tremendous expansion of parallel markets in SSA and the rising black market exchange rate premium which is d.rectly related to real overvaluation and economic distortions in general (tab': (1.2>.2 Another possible explanation of Africa's economic performance emphasizes the exogeniety of the determinants of economic growth in Africa, especially the effect due to the observed sustained worsening TOT for SSA and the declining external finance available to it. Regardless of the extent of domestic policy accommodation to external shocks, "The results of these unfavorable TOT have been increased indebtedness which, in its turn, has given birth to crippling debt repayments that are starving all sectors of the economy of the essential imported inputs. Faced with limited resources, African governments have reduced investments in infrastructure and in most cases even existing infrastructure is in disrepair for lack of necessary inputs for maintenance. This further contributes to the structural rigidities that blunt supply responsiveness of African agriculture," Mkandawire (1990). Also Ndulu (1990) found strong evidence in support of the exogeniety of growth thesis in the context of SSA.'3 Cross sectional evidence from SSA suggests that while negative externai shock has been the trigger faltor that led to economic decline in SSA; it does not, however, explain economic performance in isolation of initial conditions or policy stance, nor does it explain variations in economic performances across countries in SSA. (Elbadawi (1991)), "2Using an (. noIrrtric model, Ghura and Grennes (1991) estimated RER misalignment in SSA in excess of 28% per annum over the (1972-87) period. '3The official African view at the time also emphasized the role of external factors such as world recession, falling commodity price, rising inerest rates and debt burden, as well as drought, as the major factors responsible for Africa's economic crisis (Lagos Plan of Action (1980), O.A.U). -8- In actuality both exogenous external shocks and domestic policy mistakes have been at work to undermine economic performance in SSA over the past two decades. Also it "should be noted, of course, that African governments ackrowledge that poor domestic policies in the early 1980's have played a role in the economic crises. In the same way, external commentators certainly acknovledge the impact of external and climatic constraints on the P--onomies of SSA countries," Oyejide (1990). At any rate, with regard to our specific interest in this paper concerning the structural shifts in the economies of SSA and the role of agriculture, it is clear that both of the ibove two approaches necessarily rule out any possibility that such change in the structure of the African economies could have been brought by the diversification effects of a growing a ,.iculture at the initial stage of economic development. Faced with severe macroeconomic problems such as falling export earnings, worsening balance of payments, mounting debts, and declining economic growth; many African countries undertook economic reform programs, almost .4ll 0 them assisted by multilateral and bilateral donors. In 1979 the World Bank for example, introduced the concept of structural adjustment lending (SAL) in order to lielp countries experiencing difficulties in adopting to external shocks, to phase out the initial cost of the stabilization part of adjusiment while implementing appropriate policy and institutional reforms aimed at making the economy more flexible and strengthening its capacity for adjusting relatively more efficiently and easily to future shoks (World Bank (1985)). The participation on the SALs programs on the part of the countries of SSA over the 1980s has been quite substantial."4 The SALs and SECALs policy prescriptions strongly emphasize the adoption of outward-oriented development strategy, specially export expansion as "By FY 1988, the share of SALs and the broadly similar Sectoral Adjustment Loans (SECALs) in Bank lending was almost 25%. Between 1979 and 1987, some 25 SSA countries received World Bank adjustment loans almost half of total SAL and SECAL lending (Oyejide (1990), table (2)). - 9 - the primary channel for eliminating the balance of payments and debt problems. Considerable and recurrent nominal devaluations, macroeconomic retrenchment, and foreign trade and institutional reforms were the main vehicles for eliminating real overvaluation and creating a structure of incentives consistent with this strategy. Given the intensity and the nature of tl ese reforms, in addition to the dominant role of the highly tr4. dble agricultural sector in the economies of SSA, agriculture became the main focus of these programs an the success of the later depended heavily on the supply response of agriculture. As can be seen from table (1.2), some real depreciation has been achieved over 1986- 89.'5 Also the deteriorating performance of agriculture have been reversed."' This however, does not imply that this improvement can be solely or even mostly attributed to improved stricture of incentives for agriculture, since other nonprogram factors such as the weather conditions may have influenced the outcome. Even if the structure of incentives was sufficiently favorable to agriculture, the supply response may be blunted by other components of the program, especially the cut in government expenditure and import compression. The observed decline in investment and import ratios over the reform period (table (1.2)) can reduce both capacity growth as well as capacity utilization'7 in the economy. '`1986-89 is argued to be the appropriate period to assess the effect of structural adjustment programs on economic performance (see World Bank (1990)). "Total agricultural production grew at an average rate of 4% per annum during 1984-88 (FAO, 1989). Also per capita agricultural output recovcred from earlier declines and was rising or stable between 1985 and 1988. "'This is because imported intermediate goods are imperfect substitutes to domestically produced goods in most of the economies of SSA (Ndulu (1990)). - 10 - Table (1.1) Average Ratios of Producers Prices to Border Prices for selected African countries (converted at nominal exchange rates), 1970-86. 1970-81 1982-86 KENYA Smaliholder coffee 0.93 0.84 Smallholder tea 0.64 0.73 MAL AWl Smaliholder Dark-fired 0.23 0.23 Estate Burley 0.51 0.36 Estate Flue-cured 0.67 0.41 TANZANIA Smallholder tobacco 0.42 0.33 Smaliholder cotton 0.51 0.85 Smaliholder coffee 0.41 0.46 CAMEROON Arabica Coffee 0.52 0.37 Robusta Coffee 0.44 0.38 Cocoa 0.44 0.46 Cotton 0.48 0.73 NIGERIA Cotton 0.70 0.90 Palm Kernel 0.73 1.27 SENEGAL Groundnuts 0.32 0.52 Cotton 0.13 0.15 Source: Computed from Uma Lele (1990) Tables (12) and (14). -ll- Table (1.2) Selected Macroeconomic Indicators for Sub-Saharan Africa INDICATOR 1973-81 1982-85 1986-89 Investment to GDP ratio 21.5 18.7 17.1 Domestic Savings to GDP ratio 13.0 8.8 9.1 Resource Balance to GDP ratio -8.3 -9.7 -8.0 Imports to GDP ratio 37.2 37.0 34.5 Debt Service to Exports ratio 9.6 18.4 26.4 REER (1980=100) 11 95.5 113.5 89.4 Terms of Trade Index 106.3 91.6 80.8 Rate of change of CPI (inflation) 16.5 17.7 20.5 Black market exchange rate premium (%) 2/ 128.9 221.9 90.9 Fiscai Deficit to GDP ratio 5.3 7.4 7.8 External Shock 3/ 0.1 -5.3 -2.2 External Financing (net flows in 1980 US$mn) Total 4/ 7830 (29%) 3839 (-28%) 4635 (250k) Public 7136 (28%) 3357 (-3o%) 4272 (310k) Private 694 (55%) 482 (-13%) 363 (.03%) Notes: 1/ Index of the period average exchange rate of the currency to a weighted geometric average of exchange rates for the currencies of selected partner countries and adjusted for relative price movements in national price of the home country and its partners. An increase in the index reflects an appreciation. 21 Includes only Ethiopia, Ghana, Kenya. Malawi. Sudan. Tanzania, Zaire and Zambia. 3/ The total effect of external shocks as % of GDP is computed as the sum of the real interest rate effect and the terms of trade effect. The interest rate effect Is calculated as -(r-rbase)1(debt/GDP)beg, where r is the real interest rate computed as (i-dp/p)/(1+dp/p); rbase is the average real interest rate of base period; it is the ratio of Interest payments to total debt; interest payments are calculated by adding public interest payments to private interest payments; private interest payments are proxied by multiplying private debt by L (L equals three-month annualized LIBOR plus one percent); the private debt is estimated by subtracting public and publicly guaranteed debt from total debt; dp/p is 'world inflation (proxied by the percentage change of the GNP deflator of the US), and (debt/GDP)beg is the ratio of debt to GDP of the year preceding the beginning the end period. Debt data correspond to total disbursed guaranteed and non guaranteed debt. The effect of terms of trade is computed as (((PX/PXbase)-1) (X/GOP)begl-[((PM/PMbase)-l) (M/GDP)begj. where PX and PM are the average export and import price indices deflated by US GNP deflator. respectively: PXbase and PMbase are the average price indices of the base period: X and M are exports of GNFS and imports of GNFS respectively: and (XJGDP)beg and (M/GDP)beg are the ratios of X and M to GOP respectively at the year preceding the begining of the end period. All the va-iables are denominated in current US dollars. 4/ The periods used are 1970-1980, 1983-1985 and 1986-1989 respectively. The figures in parentheses refer to average annual growth rates. Source: World Bank (BESD), OECD 1990 Report, Pick's Currency Yearbooks. Table (1.3) Economic Structure and Performance in Sub-Saharan Africa 1970-81 1982-85 1986-89 Real arowth rates GDP 3.7 1.4 2.0 Exports 3.6 1.1 3.3 Share in GDP Agriculture 39.3 36.7 36.3 Manufacturing 20.1 22.0 21.7 Services 37.9 41.3 41.2 Share in Labor Force Agriculture 78.5 Manufacturing 7.8 .. Services 13.7 Rural population (% of total population) 79.6 74.4 71.2 Source: World Bank Data Bank (BESD) - 13 - 2. THE EQUILIBRIUM RER AND ITS DETERMINANTS Measuring the degree of misalignment is difficult, since it requires measuring an unobserved variable, the "equilibrium" real exchange rate. An common approach is based on the purchasing power parity doctrine: a base period is chosen in which the economy is thought to be in equilibrium, and then the real exchange rate for this year is dubbed the equilibrium for the remainder of the sample period. A fundamental problem with this, however, is that economic theory tells us that the real exchange rate moves over lime in an economy in equilibrium. The PPP approach therefore runs the risk of identifying as a misalignment what may in fact be an equilibrium movement in the real exchange rate. We follow Edwards (1989) in defining the "equilibrium real exchange rate" (ERER) as "the relative price of tradables to nontradables which, for given sustainable values of other relevant variables such as taxes, international terms of trade, commercial policy, capital and aid flows and technology, results in the simultaneous attainment of internal and external equilibrium", (Edwards (1989), pp. 16). Internal equilibrium is achieved when the market for nontradable goods clears in the present and is expected to clear in the future; external equilibrium hoids when present and future current account balances are compatible with long-run sustainable capital flows. As pointed out by Edwards, this definition of the ERER differs from the traditional PPP definition in treating the ERER as a function of other real variables (the "fundamentals") rather than as a fixed number. Furthermore, since the above notion of equilibrium is necessarily intertemporal in nature, the path of the ERER will not only be affected by the current values of the fundamentals, but also by anticipations regarding the future evolution of these variables.8 '"Edwards (1986b) and chapter 2 of Edwards (1989) formalizes this concept of ERER in the context of 2n intertemporal optimizing model; see also Lizondo (1989). - 14 - The ERER therefore experiences movements in response to exogenous and policy-induced shifts in its real fundamentals. In addition to such movements, however, the observed RER is also influenced in the short to medium run by macroeconomic and exchange rate policies that are not part of the fundamentals. RER misalignments can occur (as in the standard PPP theory) when those policies are inconsistent with the fundamentals. In a system of pegged nominal exchange rates, expansionary fiscal and monetary policy can be a cause of persistent real overvaluation; Edwards (1989) and Elbadawi (1989) provide strong empirical evidence on this. The remainder of this section is devoted to formulating a simple and parsimonious real exchange rate model. As indicated above, a successful modelling strategy should have at least three elements: (i) it should specify the ERER as a forward-looking function of the fundamentals; (ii) it should allow for flexible dynamic adjustment of the RER toward the ERER, and (iii) it should allow for the influence of short to medium run macroeconomic policy on RER. 2.1 The Model The RER model accounts for the traditional long run real determinants such as the terms of trade and commercial policy. The model also incorporates the effect of domestic absorption which reflects the impact of excess aggregate demand in the economy. Another attractive aspect of this model is that it (;xplicitly considers the impact on RER when conditions of excess domestic demand for foreign exchange under currency inconvertibility give rise to the emergence of a black market for foreign exchange where domestic residents can acquire foreign currency at a premium rate; and where a large number of imports are transacted according to this depreciated rate while most of exports, mainly agricultural or mineral in SSA, were required to be surrendered at the officially sanctioned rate. -15 - Let the (dollar) denominated international price of exportables and importables be given by PI and P respectively. By invoking the small cou.ntry assumption, P. and P can be considered as exogenous variables. Therefore for a given set of exchange rate and commercial policy, P, and P, the domestic prices of exportables and importables will be determined respectively by PX and P . Let E, and E, be respectively the official and the black market rates of exchange where the exchange rate is given in terms of units of domestic currency per unit of the foreign currency (dollar). Also let t, be the net export tax rate, and t, be the net tax rate on imports which may also include non-tariff implicit taxes such as quota rationing. The domestic price of exportables and importables can then be defined as in equation (2.1) and (2.2) below: (2.1) P, =E0(1-t,)P7 (2.2) Pm -E4El-a.(l+t,,,)P*,OO ; then since the coefficient is positive, the level of the RER in the following period will increase. This estimate of automatic adjustment for the Sudan is much larger than the estimate of 0.19 obtained by Edwards(1989) for a group of developing countries using a partial adjustment model. Even though Edwards(1989) method is different from ours, the comparisons suggests that automatic adjustment in the Sudan is more effective than in other developing countries. The coefficients due to AF,. reflect the short run impacts of transitory changes in the fundamentals. In the short run the effects due to foreign prices have the same signs as in the long run and both are statistically significant; the transitory influence of the foreign price of imports however, is much stronger than the one due to the foreign price of exports. Commercial policy is also shown to have negative and large transitory effects on RER, the effect due to export taxes however, is only marginally significant. Though only slightly significant as well, the short run effect of a temporary rise in domestic absorption is predicted, as expected, to lead to RER appreciation. Finally, the effect of the rate of devaluation of the official exchange rate is both, unexpectedly, negative and highly insignificant. This result which suggest that starting from a position of overvaluation, nominal devaluation in Sudan has not been effective in accelerating the process of real depreciation; does not bode well with the evidence obtained for other developing countries (for example Edwards ogpcit.). The full effect of the devaluation, however, may be embedded in the effect due to the term Alog(DC ),.,, since we could write Alog(DC),., = Alog(DC),. - AlogEo where y ~~ ~~~y** y - 28 - the first term of the right hand side reflects excess aggregate demand as measured by the excess of the rate of growth of domestic credit over that of real output. In this interpretation of the short run effect of domestic absorption a 100% devaluation will lead to a 26% real depreciation in the short- run. - 29 - Table (3.1) Key Indicators of the Agricultural Sector Role in the Sudan Economy INDICATOR YEAR (%) Share of agriculture in GDP 1988/89 36 Contribution to foreign exchange earnings 1986/87 93 Share to GDP of Govt. investment in agriculture 1984/85 8 Share of total invesment in GDP 1984/85 17 Share of total employment 1982/83 44 Source: An updated version of Table 1, Elbadawi (1988) - 30 - Table (3.2) Sectoral Distribution of GDP, Exports and Labor in the Sudan Indicator Year Agriculture Manufacturing Services 1970 43 15 42 Percent share in GDP 1980 34 14 52 1988 36 15 49 1970/71 89 11 .. Contribution to value of export (/) 1977/78 95 53 1982/83 90 10 1970/71 70 3 27 Distribution of labor force (%) 1977/78 69 4 27 1982/83 70 5 31 Source: An updated version of Table 3, Elbadawi (1988) - 31 - Table (3.3) Sudan's Real Exchange and its Determinants RER 1/ TOT tx tm DC/GDP H/GDP (M-X)/GDP 1970 100.00 100.00 0.08 0.52 0.29 0.12 -0.02 1971 94.43 100.06 0.09 0.52 0.31 0.11 0.02 1972 89.36 100.59 0.08 0.49 0.34 0.12 0.02 1973 80.21 100.53 0.07 0.39 0.35 0.14 -0.00 1974 64.62 103.18 0.08 0.42 0.32 0.13 0.04 1975 55.46 98.69 0.08 0.30 0.42 0.12 0.13 1976 53.62 99.31 0.09 0.38 0.45 0.12 0.11 1977 40.83 101.82 0.16 0.68 0.46 0.15 0.07 1978 43.00 102.87 0.12 0.51 0.46 0.18 0.09 1979 39.89 97.83 0.13 0.46 0.50 0.21 0.07 1980 40.14 90.42 0.09 0.38 0.49 0.19 0.13 1981 46.30 82.46 0.16 0.41 0.53 0.23 0.15 1982 52.17 79.54 0.17 0.22 0.42 0.19 0.16 1983 44.04 90.70 0.02 0.22 0.40 0.21 0.16 1984 40.79 99.30 0.05 0.48 0.39 0.20 0.08 1985 40.79 93.12 0.03 0.54 0.41 1.53 0.09 1986 41.45 95.74 0.05 0.37 0.37 0.25 0.08 1987 38.57 90.57 0.05 0.36 0.39 0.24 0.03 1988 28.48 85.38 0.04 0.36 0.39 0.23 0.09 Notes: 1/ RER=exchange rate for exports/price of nontradables. (decrease indicates a real appreciation) Source: Elbadawi (1988) - 32 - Table (3.4) Rates of change in some of Sudan's key macroeconomic indicators (percent) Money Official Domestic Inflation Real Official Depr. in Supply(M1) Reserve Credit %\ACPI GDP Deval. B.M. rate RER 1i 1971 7.09 4.08 16.83 2.85 6.94 0.00 0.00 -5.57 1972 16.64 109.80 19.81 6.32 -2.14 0.00 2.86 -5.36 1973 21.61 72.90 11.76 9.67 -8.07 0.00 2.78 -10.24 1974 35.08 32.43 29.87 12.20 10.09 0.00 0.00 -19.44 1975 18.59 40.41 58.52 6.65 12.67 0.00 0.00 -14.17 1976 24.49 28.49 28.15 1.42 18.41 11.90 2.70 -3.33 1977 41.92 161.31 31.66 1.12 15.19 36.17 -2.63 -23.84 1978 27.46 52.12 24.02 12.15 -1.56 32.81 8.11 5.30 1979 32.04 36.60 22.71 10.84 -10.37 15.29 25.00 -7.22 1980 31.19 -18.21 21.63 26.00 0.95 -11.22 26.00 0.63 1981 39.50 99.34 34.40 17.99 2.09 17.24 42.86 15.34 1 1982 36.59 4.37 16.08 10.76 12.68 69.61 44.44 12.67 1983 11.72 102.30 29.47 14.04 2.06 21.39 7.69 -15.58 1984 18.33 13.37 19.53 16.57 -5.01 16.19 21.43 -7.37 1985 49.94 114.52 36.45 1.52 -6.28 164.34 47.06 -0.00 1986 41.12 34.10 39.14 8.00 9.73 -7.75 24.00 1.61 1987 32.82 21.82 38.99 7.69 1.13 12.77 19.35 -6.94 1988 44.41 -2.43 33.26 14.02 -1.90 50.82 21.62 -26.16 Notes: 1/ RER = exchange rate for exports/price of nontradables (decrease indicates a real appreciation) Sources IMF International Financial Statistics, Picks World Currency Yearbook, Elbadawi (1988) Table (3.5) VARIANCE-RATIO STATISTICS OF k-DIFFERENCES FOR RER AND ITS FUNDAMENTALS (Equation 3.12') RER PX PM TX TM DCY RESIDUAL k statistic std error statistic std error statistic std error statistic std error statistic std error statistic std error statistic std error 1 1.000 0.530 1.000 0.281 1.000 0.149 1.000 0.079 1.000 0.042 1.000 0.022 1.000 0.012 2 1.187 1.292 1.086 0.764 1.438 0.598 0.727 0.237 0.980 0.126 1.027 0.071 0.905 0.035 3 1.184 1.990 0.909 1.013 1.543 0.875 0.584 0.286 0.675 0.108 0.831 0.050 0.542 0.015 4 1.222 2.822 0.868 1.414 1.240 1 012 0.493 0.288 0.485 0.081 0.844 0.039 0.307 0.007 5 1.275 3.801 0.778 1.763 0.859 0.903 0.426 0.229 0.413 0.057 0.763 0.026 0.162 0.002 delta RER delta PX delta PM delta TX delta TM delta DCY k statistic std error statistic std error statistic std error statistic std error statistic std error statistic std error 1 1.000 0.530 1.000 0.281 1.000 0.149 1.000 0.079 1.000 0.042 1.000 0.022 2 0.538 0.586 0.728 0.232 0.904 0.114 0.538 0.033 0.728 0.013 0.904 0.007 3 0.370 0.622 0.361 0.126 0.781 0.055 0.370 0.011 0.361 0.002 0.781 0.001 4 0.288 0.665 0.294 0.113 0.456 0.030 0.288 0.005 0.294 0.001 0.456 0.000 5 _0.257 0.766 0.226 0.103 0.316 0.019 0.257 0.003 0.226 0.000 0.316 0.000 Notes Standard errors are computed according to Bartlett asymptotic procedure. RER = (Exchange rate/Price of non tradables) Pm = Foreign price of imports Px = Foreign price of exports tx = export tax rate tm = import tax rate DCY = Domestic CreditlGDP Figure (3.1) VARIANCE RATIO STATISTICS OF k-DIFFERENCES OF RER AND ITS FUNDAMENTALS (Equation 3.12') RER PX RER 1.40 C20 1.20 1.00 0.60~~~~~~~~~~~~~~~~~~~0 0.40~~~~~~~~~~~~~4 0.900''oo 0.00 0.20~~~~~~~~~~~~~~~~04 0.a a 4* 1 3 4 6 3 4 a 4 RER dRERA- PX dPX PM + PM TX TM DCY 1.20 20 { {20 *0.0 O" OA.00 OOJ ' \OJ0\ 0.40 020, 0.20 _ 020 0A|,.0.00 O." 0.00 a 2 a 4 1 3 2 a 4 S 1 2 3 4 6 -TX +TX TM - TM - oCY + aDC Figure (3.1) contd. VARIANCE RATIO STATISTICS OF k-DIFFERENCES OF THE RESIDUAL 1.20 1.00 0.80 N 0.60- 0.40- 0.20- 0.00 ' 1 ~~~~2 3 4 5 RESIDUAL Based on equation 3.12' Figure 3.2 RATE OF CHANGE OF RER IN SUDAN (Actual, Fitted and Forecast) 0.101 0.05 F 0.00 -0.05 -0.10 -0.15 , , , , I , I I , 1972 1974 1976 1978 1980 1982 1984 1986 1988 -Actual + Fltted Forecast Note: Based on Eqn.(3.13') in the text. - 37 - 4. REAL OVERVALUATION AND THE TAXATION OF AGRICULTURE IN SUDAN 4.1 Equilibrium RER and Overvaluation Now we can proceed to compute an index for ERER using the estimates of equation (3.8') above for given 'sustainable' or 'permanent' values of the fundamentals. The permanent components of the fundamentals are obtained by using the time series technique introduced by Beveridge and Nclson(1981) and further elaborated upon by Nelson and Plosscr (1982), Cuddington and Winters (1987), and Cuddington and Urzua (1989). Beveridge and Nelson (1981) show that any variable x, with an integrated process (ie. a unit root) can be decomposed into a random walk with drift and stationary component, ie. (4.1) x, = x,-l + pt + C(L) et This technique is desirable for the problem at hand because unlike the trend stationary model based decomposition it allows the steady state growth path of the time series to shift upwards or downwards over time. Fluctuations around the (shifting) permanent path reflect cyclical effects. The stochastic and growing nature of the permanent fundamentals predicted by this procedure is indeed a minimal identification condition for the derived ERER to be consistent with the concept of RER equilibrium as outlined in section 2 above. The decomposition procedure is based on the following steps. The first step is to take logarithms of the individual variables and then first difference to obtain stationary series. Sccond, the Box-Jenkins techr 'que is used to identify and estimate an ARIMA model. The results of the estimated models are shown in the appendix table (A.4.1 ). Third, using the steady state "gain function" approach (Cuddington and Winter (1987)), a First order difference equation in the permanent componcnt of the series can be obtained as a linear function of the corresponding - 38 - residuals. Finally, using the computational technique suggested in Beveridge and Nelson, and Cuddington and Winters, the difference equations can be solved for the levels of the permanent components. The transitory component then obtains as the permanent minus the observed series. Figures (A.4.1) - (A4.6) (of the appendix) show the permanent and transitory components of the variables in question. The derived ERER series is shown in Figure (4.1) together with the actual RER series, where the two series are set to equal to 100.00 in 1970. This anchoring of the two rates is needed for the estimation of the degree of overvaluation. Despite the arbitrariness in this choice, given the development of the economy, 1970 is an appropriate base period to assume equality for the two rates. An important stylized fact emerges from the above graph. The equilibrium real exchange rate, being a functional of its changi-,g fundamentals, can in fact undergo considerable variability. It follows that at least some portion of observed RER variability may be related to equilibrium behavior, and that analyses of real exchange rate misalignment based on historical comparisons of observed RER levels (i.e., the PPP approach) may lead to erroneous conclusions. A closer look at the extent of real misalignment is provided by Table (4.1), which gives the average annual percentage excess of the RER over its equilibrium level. The results of the table show steadily increasing real overvaluation over the second half of the 1970's, with overvaluation increasing from 8% in 1974 to 16% in 1979. As described in section (3) above this period was one of expansive macroeconomic policy and exogenous terms of trade shocks triggered by the post 1973 oil price hike. This later exogenous effect is predicted successfully by the model as the index of real overvaluation reached 8% in 1974 which is more than 250% of the level in 1973. Despite the failure of the IMF-assisted reform program of the 1980's, it appears that real depreciation was achieved and consequently real overvaluation declined steadily from 16% in 1979 (when the program started in September) to 11% in the following year. Overvaluation then declined sharply to only 3% in 1981, - 39 - and in 1982 there was an undervaluation of more than 5%; at this stage of the program, substantial liberalization of exports has been effected. The following two years witnessed a rise in overvaluation to 5% and 7% in 1983 and 1984, respectively, as economic reforms slowed or temporarily stalled. Overvaluation once again came down in 1985-86, where it declined to 3% and 5% respectively. This is made possible basically by the adoption of a more depreciated nominal exchange rates for exports, even though no integrated macroeconomic reform was in effect. During the last three years 1987-89, overvaluation increased steadily from 7% in 1987 to 17% in 1988, to reach 19% in 1989. 1988 marked the beginnings of a move away from reforms which developed into a major policy reversals since the second half of 1989. 4.2 Direct and Indirect Taxation of Agriculture As explained by Krueger, Schiff, and Valdes (1988), Hereafter (KSV), there are four well- known stylized facts about the agricultural policies of developing countries. First most countries have attempted to promote industrial development through trade protectionism, second the exchange rate has been kept overvalued as a result of restrictive trade and payment regimes. Third, agricultural prices are generally suppressed through marketing boards, forced procurement, export taxation, etc. Fourth, some governments, have attempted to affect or partly neutralize the effects on agriculture of the above interventions by investing in agricultural supply such as irrigation, research and extension; by subsidizing input prices or extending cheap credit to farmers. Also as noted by (KSV); the consequence of and interactions among, these interventions have not been fully appreciated. We focus here on assessing the extent of the impact on Sudanese agriculture due to direct interventions (the last two set of policies) and its indirect counterpart provided by the first two types of interventions. The methodology used to derive these indices is given in (KSV). A slightly adopted version of (KSV) formula is given below: -40 - We start by introducing some definitions. Let P, be the domestic producer price of the aggregate tradeable agricultural product (and adjusted for transports, storage, and other marketing costs). Let P, Z P.EA7 be the border price P evaluated at the official nominal exchange rate applied for agriculture tradables. Let PvA = aPNAr + (1-a) PN,,A, be the nonagricultural sector price index which is a weighted average of the nonagricultural tradable price index PNAT and the nonagricultural home good price index PNAH. Let PB = P;.E' be the border price evaluated at the equilibrium nominal exchange rate E'. And letPN4 = aP;A7E' + (1-a)PN,M, where oPAT = PNA?ENA(1 +tNAT), ENAT iS the nominal exchange rate applicable for NAT, and tNAT is the rate of taxes on nonagricultural tradables. P;L therefore is the nonagricultural price index where the price index of the tradeable part is evaluated at the equilibrium nominal exchange rate, EC, and in the absence of trade policy affecting nonagricultural tradables. Then the direct nominal protection rate, which measures the difference between the relative domestic price and the relative border prices as a ratio to the relative price at equilibrium, is (4.2) NPRD = (XNA - PB/PMA) (PP/PNA) This index measures the effect of price controls, export taxes or quotas and the other policies directly affecting P,. The indirect nominal protection rate which measures the effect of the disparities between the official EAT from its equilibrium, and the inter-tradeable effect of trade policy on PNAT, is - 41 - (4.3) NPRI = D P8 /PNVA Note that the total nominal protection rate is given as: (4.4) NPRT = NPRD + NPRI = (P1 IPNA) I (PB /P,A) 1. In our case we use an equilibrium real exchange rate, c = (E/Pv,4,)' , instead of the nominal equilitrium rate E'. This requires adjusting the formula for PB /PvA in equations (4.2) and (4.3) as below: (4.5) P/(PNA = CCPAT2e + (1-a) where e is the equilibrium RER. Using annual Sudanese data (see appendix table (A4.2) and our ERER and overvaluation derived estimates in equations (4.2) - (4.5) above, I calculated the direct, indirect, and total indexes of taxation imposed on Sudanese agriculture. The results are shown in table (4.2) below. The numbers on direct intervention provide estimates of the percentage by which doliiestic producers' prices diverge from those that would have prevailed in an environment with no taxes on agricultural tradables (given the actual exchange rate and the degree of industrial protection). This measure is equivalent to the rate of nominal protection, KSV (1988). The results reveal that direct interventions have been rather substantial over the first half of 1970s with an annual average in excess of 20%. For the following period however, direct taxation of agriculture declined considerably to average less than 10% per year and in 1987 and 1988 it even dropped to less than 4%. This evidence is partially explained by the fact that the - 42 - Sudanese government is apparently pursuing a policy of self-sufficiency in food grains based on extending sector specific price protection. The table also contains the estimates of indirect interventions; which includes both the effects of trade and macroeconomic policies on the real exchange rates, and the extent of protection afforded to nonagricultural commodities. As the numbers show, the most devastating impact on agriculture is provided by this economy wide interventions. Since 1975, the cost of indirect intervention to the Sudanese agriculture have accounted for about five times that of direct taxation. The extent of indirect taxation increased steadily from an annual average of less than 20% for 1970 - 1973, to 40% for 1974 - 76, before jumping to an annual average of 55% during 1977-80. The cost of indirect intervention only partially declined to an average of 42% per annual for the two years 1981-82, before rising to an annual average of more than 55% for the following years. This brief decline in NPR, could not match the reductions in RER overvaluation over the first half of the 1980s. Actually as can be seen from Figure (4.2), real overvaluation is transmitted into agriculture with a magnification effect. Among other determinants, we will study in the next section the extent to which such severe taxation (especially the indirect) has impacted the growth of agriculture in Sudan. 43 Table (4.1) Equilibrium Real Exchange Rate and Overvaluation in Sudan Real Exchange Rate Equilibrium (EatVPn) Exchange Rate Overvaluation 1971 100.00 100.00 0.00 1972 98.99 99.27 0.28 1973 97.01 99.77 2.85 1974 93.05 100.46 7.96 1975 90.26 98.12 8.71 1976 89.63 97.01 8.22 1977 84.65 98.82 16.75 1978 85.59 98.41 14.97 1979 84.22 98.02 16.38 1980 84.33 93.54 10.91 1981 86.95 89.53 2.97 1982 89.13 84.28 -5.44 1983 86.03 90.03 4.65 1984 84.63 90.40 6.82 1985 84.63 86.94 2.73 1986 84.92 89.06 4.87 1987 83.60 89.48 7.02 1988 78.05 90.95 16.52 1989 75.89 90.55 19.31 EatIPn = (Exchange rate for agricultural tradables/Price of non-tradables) 44 Table (4.2) Taxation of Sudanese Agriculture (Percent) NPRI NPRD NPRT 1970 1.54 20.94 22.48 1971 6.15 21.03 27.18 1972 ,1.53 '8.40 29.92 1973 20.72 16.95 37.67 1974 34.17 15.85 50.02 1975 42.24 13.14 55.39 1976 45.17 11.3? 56.49 1977 57.91 8.99 66.90 1978 55.49 8.95 64.43 1979 56.51 8.58 65.09 1980 54.17 10.08 64.25 1981 43.55 11.92 55.47 1982 40.22 12.07 52.29 1983 50.37 9.22 59.59 1984 52.51 9.67 62.18 1985 53.83 8.68 62.51 1986 52.87 5.51 58.38 1987 56.86 3.37 60.23 1988 68.37 3.76 72.12 Notes: NPRI = index of indirect rate of taxation on agriculture NPRD = index of direct rate of taxation on agriculture NPRT = NPRI + NPRD Figure (4.1) Equilibrium Real Exchange Rate and Overvaluation in Sudan 800o 75U 258 1972 1974 1976 1978 1980 1982 19a4 1986 1988 -Real Exchange Rate -+-Eqlb Exchange Rate Figure (4.2) Real Overvaluation and Indirect Taxation of Agriculture in Sudan 80.0 60.0 - 20 00 - < 40.0- -20.0 , , , i , , , , , , , , , , , , , 1970 1972 1974 1176 1978 1980 1982 1984 19i6 1988 --Ovsrvaluatlon 4-NPRI - 47 - 5. AN EMPIRICAL MODEL OF ENDOGENOUS AGRICULTURAL GROWTH IN SUDAN Using the derived indexes of direct and indirect taxation on agriculture obtained in the previous section, I now test the implications of the endogenous growth model ala Easterly (1990) on Sudanese agriculture. The endogenous growth model allows for economic distortions to influence economic growth along with other traditional determinants such as the stock of physical and human capital, and the model predicts this distortion induced effect to lead to deceleration in real output. I estimated the model for two measures of agricultural growth: the rate of growth in real agricultural output (GDP), and the share of agriculture in real aggregate GDP.' In addition to the two measures of distortions (NPRD and NPR,), the specification includes the rate of population growth (age 15-64) as proxy for labor force, and the investment rate (I/GDP.1)26; the last two factors being representative of the more conventional sources of growth. Two further factors are included in the estimation as well, in order to reflect stylized facts of the growth process in SSA. In the light of the observation made by Ndulu (1990) - and referred to in section (1) above - regarding the role of intermediate goods imports in capacity utilization and the evidence that the deceleration of growth in SSA over the last decade may at least partially be explained by the decline in capacity utilization and compression of imports; I included the rate of growth of real intermediate imports for agriculture as a potential determinant. 'The second factor included is the growth rate in the permanent level of TOT. This factor provides an opportunity for testing the hypothesis of the exogeniety of growth in SSA. According to this hypothesis, permanent 'Strictly speaking, this is not a direct, measure of growth but since agricuiture is the most dominant sector. of the economy in most of SSA, a decline in its share of the economy is likely to be associated by a slower or 'distorted' aggregate growth. 26Perhaps a better measure in this case would be the share of agriculture in aggregate investment. Data limitations however precluded this option. - 48- shocks in TOT still affect growth even when it was acco'nmodated through real depreciation (as represer' 'ed by NPRI). The two equations were estimated using Sudanese annual data from 1971-1988. The results are reported in equations (5.1) and (5.2) below. (5.1) (a GDPg) = 1.79 -9.24 NPRD -2.12 NPRJ GDPag (2.35) (-2.63) ' (-2.19) + 2.12( I )t + 0.05 (ATOT) (2.95) GDP-l (0.03) TOT 0.01 (AIMP)t - 0.40 (AGDPag - Aveiage GDPag (-0.36) IMP (-1.50) GDPag GDPagt R' = 0.56, R2 = 0.29, DW = 2.02 (5.2) (GDPag) = 0.44 -0.22 NPR, -.22 ( ) f;DP (14.91) (-3.98) '(2.03) GDPX + 0.57 T) -0.05 ( AG.CON. (1.87) TOT (-1.19) G.CON. A IMP - 0.004 ( mP (-1.02) IMP RI = 0.73, R2 = 0.62, D.W. = 1.74 t- statistics are in parenthesis. The results show both measures of distortions to have - as expected - negative influences on the growth of real agricultural output (equation (5.1)); the influence of direct taxation is much stronger in terms of magnitude, however. Only indirect taxation affects the share of agriculture in -.49^ aggregate output, however (equation (5.2)). Direct taxation was not found to be significant and therefore subsequ'ently dropped from equation (5.2). In any case the stropg results obtained for the influence of aggregate economic distortions on agricultural growth as well as its share in the economy, in addition to the fact that the distortions have increased quite substantially from intermediate levels, corroborate the predictions of the Easterly (1990) model which shows that the impact on growth is likely to be larger, the larger the increase in distortions starting from intermediate initial levels. For both measures of agricultural growth, investment has appreciable and statistically significant effect. The factor reflecting the role of labor was not found to be significant in each of the two specification and was therefore dropped from both equations. The factor representing the effect of capacity utilization as reflected by the rate of growth of intermediate goods' imports, has a perverse but insignificant effect in each of the two specifications. On the other hand, there is some weak evidence that permanent deterioration in the TOT tend to reduce the share of agriculture in aggregate output. Real growth rate in ag.iculture, however, does not seem to be influenced by the exogenous factor (TOT shock) beyond its effect through overvaluation. The evidence - at least for the share of agriculture in the economy - supports the widely held view regarding the exogeniety of growth in SSA (e.g. Ndulu (1990)).'7 This also shows the vulnerability of agriculture in Sudan (and SSA) to permanent worsening of TOT even when domestic corrective measures were taker to accommodate the TOT shock. The evidence on the non-relevance of the effect due to import compression - induced decline in capacity utilization noted for SSA (e.g. Ndulu (1990)), is contrary 27"n fact permanent TOT deterioration would lower .-.quilibrium share of agriculture in the economy. This does not say that growth is determined by external TOT. But here I am implicitly assuming that at this early stage of the development of SSA; the global growth of its economy requires a rising share for agriculture-especially export-oriented agriculture, so that the process of structural transformation towards a more diversified and fast growing economy can be achieved. - 50 - to the widely held view regarding the importance of this effect in SSA.' Finally the negative sign of the coefficient of the deviation term in equation (5.1) shows that growth in agriculture tend to revert to the average where previous period higher than average growth rates get corrected for in the following period. 'Given the very few observations available for the estimation of this effect and others, further and more robust application- of the model is needed, perhaps in the context of a broadly based panel data from SSA. - 51 - 6. CONCLUSIONS Starting from the premise that agriculture should play a pivotal role in the process of structural transformation and economic development in Sub-Saharan Africa, this paper addressed two inter-related issues in this regard. The first issue is the extent to which policy induced distortions tax agriculture by causing the structure of incentives to be biased against the sector in Sub-Saharan Africa. Here the paper distinguished between two types of policy induced distortions: the direct one which is induced by micro and sectoral policies directly deployed on agriculture, and the implicit or indirect set of distortions caused by policies directed at the macroeconomic management of the economy. The second issue considered in the paper is the consequence for the growth of agriculture of these distortions given other relevant growth fundamentals. The potential deleterious effects on economic growth of such distortions have been the subject of study in the recent endogenous growth literature (e.g. Easterly (1990) and Easterly and Wetzel (1989). In the introduction section of this paper, preliminary evidence on the extent of economic distortions in Sub-Saharan Africa is reviewed. The evidence on agricultural prices in six major Sub- Saharan African countries show that despite variations across countries, farm gate prices remain generally lower compared to their equivalent border prices. Furthermore, worsening macroeconomic conditions in Sub-Saharan Africa, such as high and unsustainable fiscal deficits, worsening terms of trade, and dwindling foreign aid, have meant that the ability of Sub-Saharan African countries to invest in agricuitural supply and therefore partially mitigate the potential negative consequence of such distortions on agriculture, is substantially reduced. With regard to indirect taxation of agriculture, the analysis of Sub-Saharan Africa wide real exchange rate and its fundamentals shows sustained real appreciation over most of the 1970's and the first half of the 1980's while major fundamentals such as the terms of trade and capital flows call for equilibrium real depreciation. It is likely, therefore, that real exchange rates in SSA have been overvalued over the period. This - 52 - evidence on the direct and indirect distortions is consistent with the declining shares of agriculture and the general worsening positions of the economies of SSA over the period. The second section of the paper developed a forward-looking real exchange rate model which allows the derivation of an equilibrium real exchange rate that is consistent with both home goods and current account equilibriums. In sections 3 and using data from the Sudan - an African country with sizable agriculture - this model is estimated and used to generate an index for the equilibrium rate for sustainable levels of the fundamentals, where sustainability is given by the permanent components of the data on fundamentals as obtained by time series decomposition. In section 4, the derived equilibrium real exchange rate index is used to derive indexes of direct and indirect taxation of agriculture in the Sudan following the procedure outlined in KSV (1988). The results corroborate evidence obtained from other studies (e.g KSV) in that direct taxation of agriculture tend to decline over the years perhaps as a partial requirement to enhance self-sufficiency in foodgrains for example. Direct taxation, however, has generally been dominated by indirect taxation which remains high and in the case of the Sudan it averaged more than five times the former form of taxation since 1975. Finally in section 5, the two above indexes along with other growth fundamentals were used to test the implications of the endogenous growth model (ala Easterly (1990)) on sudanese agriculture. The estimation results show strong, negative and statistically significant effects on Sidanese agriculture for distortions as measured by the indexes of direct and indirect taxation. These results coupled with the fact that the distortions have increased quite substantially from relatively low levels, corroborate the predictions of the Easterly (1990) model which shows that the impact on growth is likely to be larger, the larger the increase in distortions and the smaller their initial levels. The results also as expected, provide strong support for the role of investment in the growth of agriculture, we could not, however, obtain significant influence for human capital, the other conventional growth fundamental. Lastly the slightly significant and negative effect due to permanent - 53 - change in the terms of trade indicates that agriculture in the Sudan could be vulnerable to permanent worsening of TOT even when domestic policy measures were taken to accommodate the shock. - 54 - REFERENCES Bates, R. (1981), Markets and States in Tropical Africa, University of California Press. Bates, Robert H., (1982), "A note on Patterns of Market Intervention in Agrarian Africa," CIT, Social Science Working Paper 456. Bautista, Romeo M., (1985), "Effects of Trade and Exchange Rate Policies on Export Production Incentives in Philippines Agriculture, "Washington, D.C.: IFPR (mimeograph). December. Behrman, J. R. (1988) Supply Response in Underdeveloped agriculture North Holland, Amsterdam. Beveridge, Stephen and Charles Nelson (1981), "A New Approach to Decompositions of Economic Time Series into Permanent and Transitory Components with Particular Attention to Measurement of the Business Cycle, "Journal of Monetary Economies 7, 151-174. Cochrane, J. (1988), "How Big is the Random Walk in GNP?," JPE, 96, 893-920, October. Commins, S., M. Loftchies, and Thys Payne (1986), Africa's Agrarian Crisis: The Roots of Famine, Boulder, col., line Rienner Publishers. Cuddington, John and L. Alan Winters (1987), "The Beveridge-Nelson Decomposition of Economic Times Series: A Quick Computational Method," Journal of Monetary Economics 19, 125 - 127. Cuddington, John and Carlos Urzdia (1989), "Trends and Cycles in Columbia's Real GDP and Fiscal Deficit," Journal of Development Economics 30, 325 - 343. Dornbusch, R., (1974), "Tariffs and Non-Traded Goods," Journal of International Economics, 4: 177- 185. Dornbusch, Rudiger (1980) Open Economy Macroeconomics, New York: Basic Book. (1986b), "Tariffs, Terms of Trade and Real Exchange Rate in Intertemporal Models of the Current Account," NBER Working paper. (1989), Real Exchange Rates. Devaluation and Adjustment: Exchange Rate Policy in Developing Countries (Cambridge, massachusetts, MIT Press) Easterly, William and Deborah Wetzel (1989), "Policy Determinants of Growth: Survey of Theory and Evidence." PPR Working Paper Series WPS 243, The World Bank. Easterly, William (1990), "Endogenous Growth in Developing Countries with Government Induced Distortions," Unpublished mimeo, The World Bank, August. - 55 - Edwards, Sebastian (1986b), "Tariffs, Terms of Trade and Real Exchange Rate in Intertemporal Models of the Current Account," NBER Working Paper Edwards, Sebastian (1986c), "Commodity Export Prices and the Real Exchange Rate in Developing Countries: Coffee in Columbia," in S. Edwards and L. Ahamed (eds.) Economic Adjustment one Exchange Rates in Developing countries, University of Chicago Press. Edwards, Sebastian (1989), Real Exchange Rates. Devaluation and Adjustment: Exchange Rate Policy in Developing Countries, Cambridge, Massachusetts, MIT Press. Eicher, Carl E., and John Staatz (1986), Agricultural Development in the Third World, John Hopkins University Press. Elbadawi, Ibrahim A., (1987). "Foreign Trade and Exchange Rate Policy and Their Long Term Impact on Sudanese Agriculture," prepared for the USAID/SUDAN. Elbadawi, Ibrahim (1989), "Terms of Trade, Commercial Policy and the Black Market for Foreign Exchange: An Empirical Model of Real Exchange Rate Determination, "Economic Growth Center Discussion Paper No. 570, Yale University. Elbadawi, Ibrahim (1990), "The Black Market for Foreign Exchange and Macroeconomic Management in Sudan," Unpublished mimeo, The World Bank, September. Elbadawi, Ibrahim (1991), "The Effectiveness of World Bank-Supported Adjustment Programs in Sub- Saharan Africa," Unpublished mimeo, The World Bank. Engle, Robert and Clive Granger (1987), "Co-Integration and Error-Correction: Representation, Estimation, and Testing," Econometrica, 35, 251 - 276. Garcia, Jorge G., (1981), "The Effects of Exchange Rates and Commercial Policy on Agricultural Incentives in Columbia: 1953-1978" Research Report No.24, Washington, D.C., International Food Policy Research Institute, June. Ghura, D. and T. Grennes (1991), "Impact of the Real Exchange Rate Misalignment and Instability on Macroeconomic performance in Sub-Saharan Africa,", memo, NC state, Raleigh, NC. Hussein, M. N., (1986) "Remittances: A synthesis of information and Analyses," in Employment and Economic Reform. Kaminsky, G. (1988), "The Real Exchange Rate Since Floating: Market Fundamentals or Bubbles?," Unpublished Manuscript, U. of California, San Diego. Kaminsky, G. (1990), "Dual Exchange Rates: The Mexican Experience 1982-1987,", mimeo, Dept. of Economics, U. of California, San Diego. Krishna, Raj (1963) "Farm Supply Response in India-Pakistan," Economic Journal, September. - 56 - Krueger, Anne O., M. Schiff and Alberto Valdes (1988), "Agricultural Incentives in Developing Countries: Measuring the Effects of Sectoral and Economy wide Policies," The World Bank Economic Review, Vol 2, No.3. Lele, Uma (1989), "Agricultural Growth, Domestic Policies, the External Environment, and Assistance to Africa: Lessons of a Quarter Century," MADIA Discussion Papers, The World Bank. Lizondo, J. Saul (1989), "Overvalued and Undervalued Exchange Rates in an Equilibrium Optimizing Model, "World Bank PPR Working Paper Series #223, August. Mellor, John (1990), "Emphasizing Agriculture in Economic Development - Is it a Risky Business?" Agriculture in Economic Development Series, IFPRI. Mkandawire, Thandika (1989), "Structural Adjustment and Agrarian Crisis in Africa: A Research Agenda," CODESRIA Working Paper 2/89, Dakar, Senegal. Mundlak, Yair, D. Cavallo and R. Domench, (1987), "Effects of Trade and Macroeconomic Policies on Agriculture and Economic Growth: Argentina 1913-1984," IFPRI Workshop on Trade and Macr .economic Policy's Impact on Agriculture, May 27-29, 1987, Annapolis, Maryland. Narain, Dharm (1965) The Impact of Price Movements on Areas Under Selected Crops in India. 1900- 32, Cambridge. Ndulu, Benno (1990), "Growth and Adjustment in Sub-Saharan Africa," A paper presented at the World Bank Africa Economic Issues Conference, Nairobi, June. Neary, J. Peter (1988) "Determinants of the Equilibrium Real Exchange Rate," American Economic Review, 78, 210-15. Nelson, Charles and Charles Plosser (1982), "Trends and Random Walks in Macroeconomic Time Series: "Some Evidence and Implications," Journal of Monetarv Economies 10, 139-162. Oyejide, T. Ademola, (1986), "The Effect of Trade and Exchange Rate Policies on Agriculture in Nigeria," Research 55, IFPRI, October. Oyejide, T. Ademola, (1990), "Supply Response in the Context of Structural Adjustment in Sub-Saharan Africa," AERC, special paper 1, Nairobi --, (1987), "the Oil Boom, Macroeconomic Policies and Nigerian Agriculture: Analysis of 'Dutch Disease' Phenomenon," IFPRI Workshop on Trade and Macroeconomic Policies' Impact on Agriculture, May 27-29, Annapolis, Maryland. Roldos, Jorge (1990), "The Terms of Trade and the Real Exchange Rate: Theory and Evidence," unpublished, University of Rochester and CERES, Uruguay, September. Schultz, T. (1978) Distortions in Agricultural Incentives. India (ed.) - 57 - Sjastad, Larry A., (1980), "Commercial Policy, True Tariffs, and Relative Prices, "Chapter 3 in John Black and Brian Hindley (eds.) Current Issues in Commercial Policy and Diplomacy, New York: St. Martin Press. Tshibaka, Tshikala B., (1986), 'The Effects of Trade and Exchange Rate Policies on Agriculture in Zaire," Research #56, IFPRI, November. Valdes, Alberto, (1985), "Exchange Rates and Trade Policy: Help or Hindrance to Agricultural Growth" Proceedings of the XIX International Conference of Agricultural Economists in Malaga, Spain, September. -, (1986), "Impact of Trade and Macroeconomic Policies and Agricultural Growth: The South American Experiment." in Economic and Social Progress in Latin America. 1986 Report. World Bank (1981) Accelerated Development in Sub-Saharan Africa: An aeenda for Action., Washington, D.C.: World Bank World Bank (1985) Annual Re, washington D.C.; World Bank. World Bank (1990) Adjustment Lending Policies for Sustainable Growth, Washington D.C.; World Bank. APPPENDIX Table (A.4.1) Series Decomposition: ARIMA Estimation SERIES MODEL R2 DW Log (DC/Y) (1-0.615L2) Alog(DC) = 0.854L2e& 0.31 1.65 y Log PX (1-0.480L2) AIogP = -0.974L2 et 0.36 2.09 Log P' (1-0.814L3) ALogP, = 0.029 + 0.567L1e, 0.56 1.94 Log (1-tx) ALog(1-tx), = -0.408L',& 0.14 1.98 Log (1 +tm) (1 +0.507L'+0.614L2) ALog (I +tm), = -0.961L3 e, 0.62 1.68 Log RER (1-0.699L2) ARER, = -0.862L2 e, 0.09 1.71 -60- Table (A.4.2) Basic Data used in the computation of Taxation of Agriculture in Sudan Pnat tx tm Pat(*) Pat(f) Pnat(*) Eat Enat 1970 100.0 0.08 0.52 100.0 82.8 188.0 0.35 0.35 1971 102.2 0.09 0.52 102.5 79.2 192.1 0.35 0.35 1972 107.5 0.08 0.49 104.1 96.6 189.9 0.36 0.38 1973 123.9 0.07 0.39 121.8 104.5 222.8 0.37 0.40 1974 150.8 0.08 0.42 166.1 114.2 265.5 0.37 0.40 1975 176.3 0.08 0.30 168.7 143.1 339.0 0.37 0.40 1976 178.1 0.09 0.38 165.7 164.5 322.6 0.38 0.40 1977 252.1 0.16 0.68 219.0 187.6 375.1 0.37 0.40 1978 255.5 0.12 0.51 231.5 259.3 376.0 0.40 0.45 1979 322.4 0.13 0.46 194.8 278.2 408.9 0.50 0.54 1980 387.9 0.09 0.38 141.2 256.1 432.4 0.63 .0.65 1981 479.5 0.16 0.41 108.9 305.3 500.1 0.90 0.68 1982 622.0 0.17 0.22 93.8 364.3 499.8 1.30 1.02 1983 813.3 0.02 0.22 154.0 616.2 414.1 1.40 1.61 1984 1036.9 0.05 0.48 192.3 809.0 374.7 1.70 1.87 1985 1486.8 0.03 0.54 123.5 1086.0 329.5 2.50 2.93 1986 1861.7 0.05 0.37 172.2 2057.4 402.0 3.10 3.38 1987 2454.4 0.05 0.36 195.4 2876.9 593.7 3.70 4.50 1988 3214.9 0.04 0.36 196.5 3548.9 777.6 4.50 9.00 Notes: Pnat = indexed price of non agricultural tradables tx = tax on exports tm = tax on imports Pat(') = foreign price of agricultural exports (US$) Pat(f) = farmgate prices for agricultural tradables (LS/MT) Pnat(^) = [Pnat/(Enat(1+tm)] Eat = exchange rate for exports (Ls/US$) Enat = exchange rate for non agricultural tradables (Ls/US$) Source Elbadawi 1988 Figure A.1.1 REAL EXCHANGE RATE Transitory Permanent 2.00 0.20 - 0.00 \ 1.50 -0.20 -l -0.40 - \ / -0.60 - 0.50 -0.80 - - 1.00 , , , , , , , , , , , I , , , I I 0.00 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 -TRANSITORY -PERMANENT Figure A.1.2 FOREIGN PRICE OF EXPORTS Transitory Permanent 1.00 2.50 - 2.00 0.60 0 1.50 0.00 1.00 0.60 -0.50 L 1972 1974 1976 1978 1980 1982 1984 1986 1088 -TRANSITORY -PERMANENT Figure A.1.3 FOREIGN PRICE OF IMPORTS Transitory Permanent 0.70 2.50 0.60 0.50 2.00 0.40 - 0.30 - 1.50 0.20 - 0.10 , , , I , , I I I , , , I , , , , 1.00 1972 1974 1974 1978 1080 1982 1914 1988 loss -TRANSITORY -PERMANENT Figure A.1.4 TAX ON EXPORTS Transitory Permanent 2.50 1.20 - 1.00 _ 0.80 -2.00 0.60- 7 0.40 - 0.20 1 - 1.50 0.00 / -0.20- , , , , , , , , , , ,. , , I , , I , 1.00 1972 1974 1978 1978 1980 1982 1984 1988 19B8 -TRANSITORY -PERMANENT Figure A.1.5 TAX ON IMPORTS Transitory Permanent 0.60 2.50 0.40 2.00 0.20- 0.00 v 1.50 -0.20 - 1.00 1972 1974 1976 1975 1950 1952 1984 1986 1988 -TRAN8ITORY -PERMANENT Figure A.1.6 DOMESTIC CREDIT Transitory Permanent 0.30 2.00 0.20 - _ 1.90 0.10 ' 1.80 N 0.00 - 1.70 -0.10 L 1.60 -0.20 ' ' ' ' ' ' ' ' ' ' ' i 1.50 1970 1972 1974 1976 1978 1980 1982 1984 196 1988 -TRANSITORY -PERMANENT Policy Research Working Pager Series Contact TM Autho r DQ for paper WPS814 Finance, Growth, and Public Policy Mark Gertier December 1991 W. Pitayatonakarn Andrew Rose 37666 WPS815 Governance and Economy: A Review Deborah Brautigam December 1991 Z. Kranzer 37494 WPS816 Economic Consequences of German Gerhard Pohl December 1991 CECSE Reunification: 12 Months After the Big 37188 Bang WPS817 How Does Brady-Type Commercial Mohua Mukherjee December 1991 Y. Arellano Debt Restructuring Work? 31379 WPS818 Do Rules Control Power? GATT J. Michael Finger January 1992 N. Artis Articles and Arrangements in the Sumana Dhar 37947 Uruguay Round WPS819 Financial Indicators and Growth in a Robert G. King January 1992 W. Pitayatonakarn Cross Section of Countries Ross Levine 37666 WPS820 Taxation in Decentralizing Socialist Christopher Heady January 1992 D. Sebastian Economies: The Case of China Pradeep K. Mitra 80423 WPS821 Wages and Unemployment in Poland: Fabrizio Coricelli January 1992 V. Berthelmes Recent Developments and Policy Ana Revenga 39175 Issues WPS822 Paternalism and the Alleviation of Nancy Jesurun-Clements January 1992 F. Betancourt Poverty 18-126 WPS823 How Private Enterprise Organized Steven M. Jaffee January 1992 C. Spooner Agricultural Markets in Kenya 30464 WPS824 Back-of-the-Envelope Estimates Sergio Margulis January 1992 J. Arevalo of Environmental Damage Costs in 30745 Mexico WPS825 The Empty Opportunity: Local Control Marlaine E. Lockheed January 1992 D. Eugene of Secondary Schools and Student Oinghua Zhao 33678 Achievement in the Philippines WPS826 Do Workers in the Informal Sector Ariel Fiszbein January 1992 N. Perez Benefit from Cuts in the Minimum 31947 Wage? WPS827 Free Trade Agreements with the Refik Erzan January 1992 J. Jacobson United States: What's In It for Lati:, Alexander Yeats 33710 America? Policy Research Working Paper Series Contact lit Aufth Dile for Q=r WPS828 How the Macroeconomic Environment Ar%'i Van Adams January 1992 V. Charles Affects Human Resource Robert Goldfarb 33651 Development Terence Kl