Session II. Presentation by Andy Berg PRELIMINARY NOT FOR CITATION March 2005 HIGH AID INFLOWS CASE STUDY: GHANA1 I. INTRODUCTION 1. Increases in aid inflows allow recipients to increase consumption and investment. Aid presents an opportunity to reduce poverty, increase the standard of living, and generate sustained growth. However, the effective use of increased aid also presents challenges. Good projects must be found and managed; conditions for budgetary support must be agreed and implemented. The imperative to use the funds well can strain the administrative capacity of recipient governments. In addition, aid flows can weaken ownership, fragment and impair budgetary procedures, encourage rent-seeking behavior, and undermine the accountability of domestic institutions.2 2. Related to but distinct from these microeconomic and institutional issues are the macroeconomic challenges of managing aid inflows. Aid inflows can cause upward pressure on the real exchange rate to the detriment of the exporting industries that may be critical to long-run growth. In the long run, this is a microeconomic phenomenon, in other words one rooted in the real effects of aid. But in the short run, macroeconomic policies can determine how aid is absorbed in the domestic economy. Aid inflows can also create problems of fiscal management and debt sustainability, particularly when they are volatile and when they come in the form of debt. 3. This case study is one of five that will form part of the background for a forthcoming study on the macroeconomics of managing high aid inflows.3 The broader study aims to draw lessons from recent country experiences about important issues in the macroeconomic management of high aid inflows. The questions the broader study will address are: 1 This paper is part of a larger study, the authors of which are Shekhar Aiyar, Andrew Berg, Mumtaz Hussain, Amber Mahone, and Shaun Roache. The principal author of this Ghana case study is Shaun Roache. This paper should not be reported as representing the views of the IMF. The views expressed are those of the author and do not necessarily reflect the views of the IMF or IMF policy. 2 See for example Brautigam and Botchway (1998). 3 The other countries covered in the broader study are Ethiopia, Mozambique, Tanzania, and Uganda. - 2 - • Do we observe macro absorptive capacity constraints? Is aid ultimately used to finance imports? • Is Dutch Disease a concern? Are the supply-side effects of increased aid and policy reforms (or excess capacity) able to offset the demand-side pressures on real exchange rate? • How should exchange rates be managed in the face of high aid inflows? • How should fiscal policy be adapted to the aid inflows? • Are aid inflows inflationary, and what is the appropriate policy response? Is there a role for sterilization? What instruments should be used (reserves, government paper, banking system)? • Did PRGF programs adequately manage the macroeconomic impact of surging aid inflows? How can program design be improved in this regard? 4. While the benefits of higher aid and challenges of scaling up are frequently discussed, there is limited systematic analysis of country experiences on these issues. This paper examines the experience of one low income country that has dealt with these questions over the past decade or so. 5. The approach taken in this study limits the range of topics that can be covered. In particular, it is difficult in this framework to address the equally important long-run implications of increases in aid flows. To understand long-run implications for resource allocation, policies, institutions, and growth, it would be necessary to pursue either much longer-run cases or a broader cross-country analysis.4 II. ISSUES IN THE MACROECONOMIC MANAGEMENT OF AID SURGES A. The Macroeconomic Impact of Aid 6. Aid works ultimately by financing net imports. We thus define the rate of aid absorption as the increase in net imports (or more generally the non-aid current account deficit) as a share of the increase in aid flows.5 This point is central and deserves belaboring. Some aid directly finances government purchases of imports. In addition, the aid serves by 4 Devarajan and Dollar (1998) is example of the former, while the large literature on aid and growth attempts to address the latter. 5 See Appendix 1 for a definition and discussion of absorption and its relationship with balance of payments accounting and reserve accumulation. Our usage not be confused with the idea of “absorptive capacity,” which is related but which also involves questions about the rate of return to investments that might be financed by aid. - 3 - financing second-round increases in net imports resulting from aid-driven increases in government or private expenditures. 6 7. Aid also serves as a way for the government to finance its domestic expenditures, as an alternative to domestic tax revenue or borrowing, either from the public or from the central bank.7 It may seem, therefore, that the financing of expenditures such as the hiring of nurses is an alternative use for aid, in addition to imports. But this approach to the function of aid is misleading; after all, the purpose of aid is not to allow the domestic expenditure. After all, the government could always simply borrow from the central bank (i.e. print money) to finance these expenditures. Rather, the purpose of the aid is to provide the foreign exchange that may eventually be required, either to sterilize the money injection associated with the domestic expenditure, or to satisfy the increased demand for foreign currency resulting from higher import demands. 8. One way to understand the issue is to consider a thought experiment in which the government first decides on the appropriate level of government expenditure and its financing, for a given level of aid. This set of decisions in principle takes into account the scope for seigniorage, the supply response to increased fiscal expenditures, the productivity of the resulting investment and the generation of higher exports that may result, and other such factors. Then, aid increases. The thing that has changed is not that the government could now productively hire, say, more nurses to fight HIV/AIDS. The difference is that, whereas before such additional expenditures would have caused too much inflation or an unfinancable deterioration of the current account through second-round increases in import demand, now the aid can be sold to sterilize the increase in money or to pay for the higher imports. 9. Of course, reserves accumulation can have important benefits. Some minimum level is appropriate as a resource for buffering shocks of various sorts. Moreover, where aid itself is highly volatile, some savings of aid in the form of reserve accumulation may be optimal. We return to this point below. 10. A short-run macroeconomic perspective on different policy responses to high aid inflows can be gained from looking at the behavior of output, net exports (or the non-aid current account), and private investment, in the context of the familiar definition of aggregate 6 The mechanisms whereby this happens, and the resulting issues, are known as “the transfer problem”, as in Keynes (1929). Keynes, however, was concerned with the problems involved for Germany in generating current account surpluses to pay reparations after World War I, rather than the problems involved absorbing these transfers. See Milesi-Ferreti and Lane (2000) for a recent general treatment. 7 To keep things simple, but also reflecting reality for most low-income countries, the possibility of foreign private financing of government deficits is ignored. - 4 - demand as Y = CP + IP + G + (X-M), where Y is aggregate demand or output, CP is private consumption, IP is private investment, G is government consumption and investment, and X- M is the trade balance. 11. Increases in aid flows can lead to increases in government expenditure (G) or private consumption (CP) or investment (IP). This implies an increase in the resources devoted to providing these goods and services. To the extent that these resources were previously unemployed, output (Y) can go up with G. However, it is likely that at least to some extent, and perhaps increasingly as fewer unemployed resources remain, the resources will have to come from elsewhere, in particular from those that had been producing exports and/or import competing goods, implying a fall in X-M. 12. Some combination of exchange rate appreciation and (if there is excess capacity) increased aggregate demand is necessary to generate the increased net imports that aid allows. Thus, some real exchange rate appreciation may be necessary and indeed appropriate, in response to a permanently higher level of aid. In the long run macroeconomic policy can only determine the channel—the domestic price level or the nominal exchange rate—through which this real appreciation occurs. 13. The degree of exchange rate appreciation required to absorb the aid will in general depend on the structural response of the economy and the uses to which the aid is put. For example, real appreciation would be higher if aid inflows mostly finance expenditures on non-traded goods, and lower if higher incomes feed strongly into higher import demand and if non-traded goods supply responds strongly to the increases in their relative price. In economies with significant unemployment and potential for a quick supply response, the additional demand for nontradeables could induce additional employment and production, with little increase in price level and limited real appreciation. In the longer run, investments that increase productivity in the non-tradable sector could reduce or even eliminate the real exchange rate appreciation. 14. The real exchange rate appreciation and resulting discouragement of exports may hurt long-run growth (Appendix 3). The mechanism, known “Dutch disease”, is that the decline in the export sector could impair productivity growth because such growth is particularly high in the non-traditional export sector. This in turn is attributed to particularly high rates of learning-by-doing in these non-traditional export sectors. For this argument to hold, dynamic externalities in the export sector would have to outweigh the benefits of capital accumulation associated with public investment (as well as any related productivity growth). B. Policy Responses 15. How much aid is absorbed domestically, and which components of GDP increase the most, depend on three closely related policy decisions: - 5 - • Are the foreign exchange proceeds associated with the aid inflows kept in the central bank or sold? 8 • Are aid inflows spent in the form of higher fiscal deficits (before aid) or saved by the recipient government?9 • How is monetary policy conducted to manage the monetary implications of the decisions with respect to foreign exchange accumulation and fiscal policy? 16. To organize thinking, it may be useful to consider how the answers to the first two questions frame the third. This implies four possibilities, which are discussed in turn.10 Aid neither absorbed nor spent 17. The government could choose to respond to the aid inflow by building international reserves, neither increasing government expenditures nor lowering taxes. In this case there is no expansionary impact on aggregate demand, and no pressure on the exchange rate or prices.11 Aid absorbed and spent 18. This is the “classic” case, in that this is (explicitly or implicitly) the situation assumed in most discussions of the macroeconomic implications of aid inflows. The foreign exchange is sold by the central bank and absorbed into the economy, and the associated resources spent by the government. In this case, the challenge to monetary policy-makers is to manage the increase in the real exchange rate that may result.12 8 The fungibility of aid implies that the relevant question is not whether a particular flow of aid-related foreign exchange remained in the central bank but whether the overall level of foreign exchange rose by the amount of the aid inflow. 9 A similar fungibility point applies here. Even if the particular aid money is spent, if other spending is reduced so that the fiscal deficit does not change, the aid does not in the end result in higher spending. 10 Appendix 3 provides a numerical example showing how the central bank and financial accounting works in each of these four cases. 11 There may be second order effects, e.g. expectations may change as a result of the central bank’s higher international reserve position, but again, this does not alter the main argument. 12 Sahay et al. (2004) present a model in which reserve accumulation considered in terms of its benefits in reallocating a given aid flow intertemporally. - 6 - Aid absorbed and not spent 19. Increased aid inflows can be used to reduce inflation in those countries that have not yet achieved macro stabilization. In this case, the authorities can sell the foreign exchange associated with increased aid flows so as to sterilize the monetary impact of domestically- financed fiscal deficits. The result would typically be a more appreciated exchange rate and lower inflation. Aggregate demand may increase as the inflation tax declines, with the increase in private consumption and investment. The deterioration of the trade balance that often accompanies such a stabilization program is financed by the aid inflow.13 20. In countries that have already achieved inflation stabilization, taxes could be cut or monetary policy eased, perhaps through the government using the proceeds from aid to reduce the stock of local currency government bonds outstanding. Either would tend to result in increased private consumption and investment. This would raise net imports through the indirect effect of higher private after-tax income on import demand. Aid inflows could fill this gap. Aid not absorbed but spent 21. A fourth possibility is that fiscal deficits would increase with the jump in aid, but that the authorities would not sell the foreign exchange required to finance additional net imports. In effect, the macroeconomic effects of the fiscal expansion are similar to increasing government expenditures in the absence of aid, except that reserves are higher. The increased deficits inject money into the economy. Absent foreign exchange sales, the monetary policy options are the same as with a non-aid-related fiscal expansion. 22. There are several monetary policy responses to this situation. One is to attempt to finance the increased deficits through the issuance of Treasury bills. This strategy would tend to “crowd out” private investment (IP in the above equation.) In effect, there is a switch from private investment to government consumption or investment.14 A second response could be to allow the larger fiscal deficits to lead to money supply increases. This would tend to be inflationary. If the authorities were in the meantime pegging the nominal exchange rate, there would be an appreciation of the real exchange rate. This, in combination with the increase in aggregate demand associated with the fiscal expansion, could lead to an increase in net import demand that would, eventually, require sales of foreign exchange in order to defend the peg. In other words, over time the response might become one of absorbing and spending. 13 This is the case emphasized by Buffie et al. (2004). 14 Of course, private investment and government expenditure could have different import intensities, which would modify the details of the argument but not alter the main point. Similarly, the fiscal expansion may increase aggregate demand, so it is not the case that there need be a one-for-one tradeoff between government spending and private investment. But such an aggregate demand expansion could have been engineered without the aid. - 7 - If the first option is rejected because of a concern for private investment, and the second conflicts with objectives with respect to the real exchange rate and reserve accumulation, a final option would be to reduce the fiscal deficit, leading to the “neither absorbed nor spent” case. Conclusion 23. There is no universally optimal approach to managing high aid inflows. • Unless Dutch Disease is a major concern or the return to public expenditure low, to absorb and spend the aid would appear to be the most appropriate response. • To neither absorb nor spend is not an appropriate long-run strategy unless it is felt that Dutch disease concerns outweigh the benefits from the absorption of aid inflows. Even here, the best response might be to work to improve the quality of public expenditures and the “quality” of aid, specifically the ease with which it can be delivered and used well. However, where aid inflows are volatile or international reserves are too low, reserve accumulation may be the most reasonable short-run response.15 • If resources are scarce for private investment, the rate of return on public expenditure is low, or fiscal policy is already too loose and inflation too high, to absorbing and not spend would make sense. • To spend and not absorb would appear to be the least attractive option, particularly when domestic sterilization domestic sterilization is used to avoid pressures for inflation and exchange rate appreciation. This is unlikely to be a sensible long-run strategy—it tends to shift resources from the private to the public sector and does not allow the country to benefit from the aid resources. III. AID INFLOWS TO GHANA: 1996-2003 A. Overview 24. Ghana has been a relatively strong performer compared to its regional peers in recent years. Per capita GDP growth averaged 1.6 percent in the 1990s. Following a period of 15 Recent cross-country evidence (Bulíř and Hamann (2001), Pallage and Robe (2001)) indicates that aid inflows continue to be volatile, that commitments consistently exceed disbursements, and that aid disbursements are generally pro-cyclical—thereby increase volatility of public expenditures rather than lowering it - 8 - economic volatility around the turn of the millennium, Ghana has followed economic policies that have delivered a degree of fiscal consolidation, lower inflation, and steadily increasing real GDP growth. Together with the achievement of HIPC Initiative Decision Point debt relief in 2002, international support has recently jumped by several percent of GDP. Net aid averaged 7.3 percent of GDP in 2001-2003 compared to 2.8 percent of GDP from 1996 through 2000 (Table 1). 25. A challenge for the authorities has been managing the huge jump in aid inflows and their volatility. This case study assesses the years 1996-2003, a period during which the country was almost continuously engaged in ESAF/ PRGF supported programs. The key points that emerge are: 1. Ghana cumulatively saved all of the increase in aid observed in 2001-03 compared to 1996-2000, where we define the increase in aid as the actual aid flow over 2001-03 less the amount of aid that would have been received had flows continued at their 1996-2000 rate. The aid inflow was saved, from the point of view of the country as a whole, insofar as gross reserve accumulation over the period (US$1.2 billion) was almost equal to the incremental aid inflow (US$1.3 billion). The fact that the inflow was saved can be seen from another point of view: that the non-aid current account deficit (investment less non-aid savings) did not grow at all with the higher aid inflows. On the contrary, the non-aid current account deficit averaged around 3½ percent of GDP over 2001-03, down from around 11 percent over the preceding period. 2. Just as Ghana from a national point of view saved the aid inflows, the government also saved the aid inflows to the budget. The fiscal deficit (before grants) averaged 10 percent of GDP over 2001-03, almost equal to the average for the preceding three and five years. 3. Ghana has avoided Dutch disease, or real effective exchange rate appreciation; over 2001-03 the REER has changed by less than 1 percent. 4. This is a tale of three policy responses, reflecting the pattern of aid flows. • In 2001, the aid surge was largely sold into the foreign exchange market, to stabilize the currency (and inflation) after a negative terms-of-trade shock; it was partially spent by the government in the sense that the fiscal deficit before grants widened, though by somewhat less than the increase in aid. • During 2002, a planned fiscal consolidation, aimed at leading to a reduction in domestic public debt outstanding, fell short of compensating for an even larger negative aid shock. Reserves were accumulated, despite the aid shortfall. - 9 - • In 2003, aid surged again and this time the authorities responded cautiously. All of the aid (and more) was accumulated as reserves. The fiscal deficit before grants did not widen. B. The Pattern of Aid Inflows 26. The pattern of aid flows is critical to understanding the policy response and economic outcomes of the aid over 2000-03. Net aid jumped in 2001, collapsed in 2002, and jumped again in 2003. Most of these changes were driven by changes in gross aid flows.16 As Table 2 illustrates, this volatility was unexpected and therefore required large and rapid policy adjustments. HIPC debt relief has also been a contributory factor; the value of HIPC assistance was around half that of loans and grants in 2002-03. Private flows, mostly transfers but also including unidentified items, have also been rising sharply, amplifying the effects of aid shocks. Table 1. Net Aid Inflows and Selected Economic Indicators 1997 – 2004 (percent of GDP) 1997 1998 1999 2000 2001 2002 2003 2004 Gross aid inflows 8.8 8.7 7.5 8.8 14.9 5.9 9.5 7.4 Project aid 7.8 6.9 5.6 5.0 9.3 3.4 4.4 3.3 Program aid 1.0 1.8 1.9 3.8 5.6 2.5 5.1 4.1 Debt service 1/ 5.3 5.6 4.7 9.0 4.2 3.4 2.4 -0.6 Net aid inflows 3.5 3.2 2.8 -0.3 10.6 2.5 7.1 8.0 Private inflows 2/ 14.0 6.5 12.9 14.8 12.7 13.9 13.7 8.5 Memorandum items GDP (real percent change) 4.2 4.7 4.4 3.7 4.2 4.5 5.2 5.2 Inflation (percent change) 18.4 16.3 13.1 39.3 23.5 14.1 24.0 ... Cedi per US dollar (average) 2050 2314 2669 5455 7170 7932 8677 9004 percent change -20.1 -11.4 -13.3 -51.1 -23.9 -9.6 -8.6 -3.6 RER vs US dollar (percent change) -7.6 1.5 -4.1 -34.0 -8.6 1.5 10.8 REER (percent chg) 6.0 8.2 0.5 -35.5 0.6 -0.6 1.4 0.0 1/ Net of arrears and debt relief, including HIPC. 2/ Includes private transfers (largely remittances) that are reported in the current account. 16 Net aid flows are calculated as the sum of gross aid, less debt service adjusted for debt relief and changes in arrears. Gross aid flows are calculated as the sum of project and program loans and grants. - 10 - 27. The policy response to, and the economic outcomes of aid inflows in Ghana in recent years is likely to have been strongly influenced by the volatile pattern of aid. This volatility was largely unexpected and required large policy adjustments. To illustrate this, Table 2 shows aid actually received less aid that was programmed around the start of that year (as reported in the closest IMF staff report). A positive number implies a surprisingly large amount of aid.17 Table 2. Aid Shocks in Ghana 1998-2003 (percent of GDP) 1998 1/ 1999 2/ 2000 3/ 2001 4/ 2002 5/ 2003 6/ Project grants -0.1 -0.5 -0.7 1.8 -1.0 -0.2 Program grants 0.3 0.2 -0.6 2.0 -0.1 0.3 HIPC assistance 0.0 0.0 0.0 0.0 0.0 0.2 Project loans 1.9 1.1 -0.7 -0.1 -1.8 0.2 Program loans 0.2 -0.3 -3.9 1.0 -1.2 0.6 Gross aid shock 2.3 0.6 -6.0 4.7 -4.1 1.8 Net aid shock 0.5 0.0 -6.1 4.9 -4.5 1.3 1/ Program as detailed in April, 1999 staff report EBS/99/57. 2/ Program as detailed in April, 1999 staff report EBS/99/57. 3/ Program as detailed in August, 2000 staff report EBS/00/160. 4/ Program as detailed in August, 2000 staff report EBS/00/160. 5/ Program as detailed in March, 2002 staff report No. 02/38. 6/ Program as detailed in May, 2003 staff report no. 03/133. C. The Use of The Aid Inflows 28. In this section, the macroeconomic response to the increased aid flows will be viewed from three, connected perspectives: (i) macroeconomic absorption; (ii) fiscal policy and; (iii) monetary policy. Absorption 29. Over the entire 2001-03 aid surge period, Ghana’s non-aid current account deficit narrowed, In other words, the aid was not used to increase net imports, or more generally raise investment relative to domestic savings. Overall, aid inflows were fully accumulated as 17 The source of most of these aid shocks is not clear; further investigation would be interesting. Almost half of the 2002 negative shock, however, reflected nondisbursement of a World Bank loan tranche owing to a delay in the divestiture of Ghana Commercial Bank.