WPS4970 Policy Research Working Paper 4970 How Should Fiscal Policy Respond to the Economic Crisis in the Low Income Commonwealth of Independent States? Some pointers from Tajikistan Martin Brownbridge Sudharshan Canagarajah The World Bank Europe and Central Asia Region Poverty Reduction & Economic Management Sector Unit & Central Asia Regional Unit June 2009 Policy Research Working Paper 4970 Abstract The paper analyses how the global economic crisis will recessionary pressures and making it very difficult to affect the economies of the low income Commonwealth protect priority social expenditures from cuts. To avoid of Independent States (CIS) and discusses the fiscal these outcomes, external support from donors, preferably measures which can be taken to help mitigate the adverse in the form of quick disbursing budget support, is impact of the crisis. It focuses on Tajikistan, the poorest required. If additional external budget support can be member of the CIS but also highlights similarities with mobilized, the priorities for fiscal policy should be to the economies of Armenia, the Kyrgyz Republic and protect spending on budgeted social sector programs and, Moldova. The main channels through which the global if sufficient budget resources are available, to implement economic crisis will affect the low income CIS economies a program of labor intensive repair and maintenance is through a sharp reduction in remittances from migrant of public infrastructure to provide employment for workers in Russia and lower export earnings. The returning migrant workers. Tax cuts are unlikely to be adjustment to this external shock will involve a reduction an effective use of scarce budget resources, either to in imports, private consumption, domestic output and stimulate the economy or protect the incomes of the government revenue. Fiscal policy, constrained by very poor. Up scaling existing social assistance programs may limited macroeconomic and fiscal space, faces acute be a feasible way to protect the poor in some low income challenges. Maintaining budget targets for fiscal deficits CIS countries provided they are not as poorly targeted as and domestic borrowing in the face of revenue shortfalls in Tajikistan. will lead to a tightening of the fiscal stance, exacerbating This paper--a product of the Poverty Reduction & Economic Management Sector and Central Asia Regional Unit, Europe and Central Asia Region--is part of a larger effort in the region to share experience and results of Poverty Reduction and Economic Management work at country level. Policy Research Working Papers are also posted on the Web at http://econ. worldbank.org. The author may be contacted at scanagarajah@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team How should fiscal policy respond to the economic crisis in the low income Commonwealth of Independent States? Some pointers from Tajikistan Martin Brownbridge and Sudharshan Canagarajah* *Martin Brownbridge is a freelance consultant, while Sudharshan Canagarajah works at the World Bank in Washington DC. The views are those of the authors and do not reflect the views of the World Bank. 1. Introduction Tajikistan, a low income transition country which is the poorest of the Commonwealth of Independent States (CIS), will be hit hard by the international economic crisis. The economy will suffer both a large fall in national income and earnings of foreign exchange, mainly because of a drop in transfers of remittances from Tajik migrant workers in Russia. The unavoidable economic adjustment to this external shock will involve lower output growth, lower private consumption and lower government revenues. The incidence of poverty and extreme poverty, estimated at 53 percent and 17 percent respectively of the population in 2007, will inevitably rise. 1 The external shock facing Tajikistan is exogenous and beyond the control of public policy. However, public policy can, arguably, mitigate to some extent the fall in domestic output, and hence incomes, given the macroeconomic constraints, of which the most pertinent is the need to ensure external balance. In addition, given the extent of poverty in the country, a strong normative case can be made that public policy should aim to protect the living standards of the poor and most vulnerable, and in particular those of the poor who will suffer a sharp loss in income as a result of the crisis. The question we address in this paper is whether fiscal policy can play a role in achieving these two objectives, and if so, what specific fiscal policies are best suited to these objectives and what sort of external support is required. Although the Tajik economy recovered strongly, with real GDP growth averaging nearly 9 percent per annum in the 2000s, the circumstances under which fiscal policy must be framed in the current crisis are far from favorable. A lack of international reserves to support the balance of payments, a very shallow and heavily dollarized domestic financial system, virtually no access to external commercial finance and large quasi fiscal deficits all constrain the room for fiscal policy to play a countercyclical stabilization role. 1 Poverty estimates are derived from the Tajikistan Living Standards Survey 2007 (Ivaschenko, 2009). 2 Designing an appropriate fiscal policy response to the crisis is constrained by uncertainty. It is uncertain how long the external shock will last. There is also uncertainty about the transmission of the shock to the domestic economy because we lack important data about the economy ­ for example, the national income accounts are very rudimentary and do not include GDP by expenditure ­ and we do not have a clear understanding of important relationships in the macro economy; for example the extent to which prices clear the market, and hence by how much output will fall in response to a fall in aggregate demand, and how dependent is production on imports. There is also uncertainty about the impact of fiscal policy on real variables; the size and sign of fiscal multipliers. The observations in this paper were carried out in the context of advisory services provided to the Tajik authorities as they faced the consequences of global economic crisis over the last few months. The process of engagement with the Tajik authorities, providing them much needed practical and technical guidance at every stage of the way in addressing the crisis, was very much appreciated by all stakeholders. Countries facing crisis need to share experiences with other countries and this was also a hallmark of the support provided to the Tajiks. The strong local leadership, inter ministerial coordination and multi sectoral approach to crisis management is something not always highlighted but the Tajik approach had its challenges of ensuring sustained effort in this area. The close partnership with bilateral and multi lateral donors, with IMF and World Bank leadership, by the government is very important to ensure consistent advice during crisis period. Such consistency in approach should not be under emphasized especially at a time of crisis when conflicting information can lead to inaction and policy inconsistency. Although the paper focuses on Tajikistan, it may have relevance for other low income non fuel exporting CIS economies, notably the Kyrgyz Republic and Moldova, and probably to a lesser extent, Armenia. 2 These four economies have some important 2 We do not include the CIS oil and gas exporters (e.g. Azerbaijan, Kazakhstan and Uzbekistan) in this paper because, although they are also suffering a large external shock, their macroeconomic and fiscal positions differ substantially from that of the CIS non oil and gas exporters. In particular, because of the 3 characteristics in common; they are relatively small economies, they are very open to international trade (with imports ranging from 37 percent of GDP in Armenia to 80 percent in Moldova), they rely on remittances to provide a substantial share of their foreign exchange earnings and they are fuel importers. They all performed relatively well during the 2000s, with real growth in part driven by the impetus supplied by the expansion of the Russian economy which provided a market for their exports, employment for their migrant workers and foreign capital. The Russian economy is now forecast to contract by 6 percent in 2009, with serious repercussions for those smaller CIS economies which are closely integrated with it. The per capita incomes, real GDP growth and forecast growth for 2009 of these four countries, along with Russia, are shown in table 1. As can be seen in the table, the real GDP growth for 2009 is currently forecast to be lower by between 7 and 12 percentage points than in 2008. Table 1 Per capita incomes, real GDP growth and 2009 forecasts for selected low income CIS economies and Russia Per capita GDP Average real Real GDP Real GDP US$ 2007 GDP growth Growth growth 2009 2000-07 percent 2008 percent forecast percent Armenia 2640 11.8 6.8 -5.0 Kyrgyz 610 4.5 7.6 0.9 Republic Moldova 1090 5.7 7.2 -3.4 Tajikistan 440 8.7 7.9 0.0 Russian 7580 7.0 5.6 -6.0 Federation impact of fuel revenues, the fuel exporters entered the crisis with stronger fiscal and current account balances than the non fuel exporters. 4 Sources: Per capita GDP and average real growth 2000-07 are from the World Bank ECA database. Real GDP growth for 2008 and forecast for 2009 is from World Economic Outlook, 2009, except 2009 forecast for Tajikistan which is the latest forecast from the IMF country team. The paper is organized as follows. Section 2 examines the nature of the external shock facing Tajikistan and analyses how this will be transmitted to the domestic economy, to living standards and poverty and to the fiscal position. Section 3 evaluates whether there is both macroeconomic and fiscal space to allow a more expansionary fiscal stance. We argue that external donor support will be necessary to create the macroeconomic and fiscal space for a more expansionary fiscal stance. Section 4 discusses the optimal characteristics of a fiscal program in terms of the objectives to both stimulate output and protect the poorest and most vulnerable. Section 5 makes recommendations for specific fiscal policy measures to be included in the program. Section 6 provides a conclusion. 2. The nature of the shock and economic crisis facing Tajikistan Since the first half of the 2000s, remittance transfers from migrant workers have risen at a rapid rate to become by far the most important source of foreign exchange earnings for the Tajik economy. Gross remittance inflows in 2008 were estimated at $2.2 billion: this amounts to the equivalent of 43 percent of GDP, 78 percent of merchandise imports and more than three and half times the combined total of all goods and services exports. 3 The Kyrgyz Republic and Moldova are also dependent on remittances, which amounted to 31 percent of GDP and about 50 percent of imports in both countries in 2008, as is Armenia to a lesser extent (see table 2). 3 Official data indicate net migration of 388,000 people in 2008, equivalent to 6 percent of the population. Unofficial estimates are 150 percent to 200 percent higher (World Bank, 2008: 5). 5 Table 2 Remittances in 2008 as a share of GDP and imports Remittances Remittances as Remittances as percent of US$ millions percent of GDP merchandise imports Armenia 1140 10 26 Kyrgyz Republic 1440 31 53 Moldova 2300 31 47 Tajikistan 2210 43 78 * For Moldova, remittances include compensation of employees Sources: World Bank and IMF Russia is the destination for 96 percent of Tajik migrant workers, of whom 55 percent work in the construction industry and the rest mainly work in other service industries (World Bank, 2008). There is a strong seasonal element to remittances with the highest inflows during the summer. Remittance inflows to Tajikistan are falling in 2009 for two reasons. First the Russian construction industry, which grew very rapidly up to 2008 on the back of the oil price boom, is contracting as Russian national income falls: a reversal of the Dutch disease induced construction boom is taking place as a consequence of the sharp fall in oil prices. Secondly, the exchange rate of the Russian ruble depreciated by 51 percent against the US dollar between March 2008 and March 2009, reducing the value of remittances in US dollars. This has already begun to affect remittances to Tajikistan. The fall in remittances in the second half of 2008 was steeper than would normally be the case on the basis of seasonal factors alone, and remittances, valued in US dollars, were 32 percent lower in the first four months of 2009 than in the corresponding 6 period in 2008 (see figure 1). It now appears likely that remittances will be at least one third lower in 2009 than in the previous year, which implies a fall in remittance inflows of about $750 million, or about one quarter of the value of imports. Remittances to the Kyrgyz Republic are also projected to fall by 30 percent in 2009 from their 2008 levels. Figure 1 Monthly Remittances to Tajikistan, Millions of US Dollars: Jan 2005 to April 2009 350 300 250 200 150 100 50 0 n v n v n v n v ar ar ar ar p p p p l l l l ay ay ay ay ch 5 06 07 08 M 9 Ju Ju Ju Ju Ja No Ja No Ja No Ja No Se Se Se Se 0 0 M M M M M M M M ar 20 20 20 20 20 n Ja Source: National Bank of Tajikistan Export earnings will be reduced in 2009 because of a fall in world market prices for metals, which are major exports of Armenia, the Kyrgyz Republic and Tajikistan and lower demand, especially in Russia which is the most important destination for exports.4 4 Tajikistan's main export is aluminum. In the 12 months since April 2008, world market aluminum prices have fallen by more than 50 percent. 7 However, unlike the middle income CIS countries, the low income CIS countries do not have large short term external debts, and so are less vulnerable to a capital account crisis than their middle income neighbors. The transmission of the external shock to the domestic economy of Tajikistan will be mediated through two main channels. First, the economy will suffer a major fall in foreign exchange availability, probably of the order of 25 percent of 2008 imports. The central bank ­ the National Bank of Tajikistan (NBT) ­ does not have sufficient international reserves to fund a BOP deficit and the country lacks the creditworthiness to access external commercial finance, hence the economy will have to adjust to ensure external balance by reducing imports. Imports will bear the brunt of the adjustment because they are the only significant component of the BOP which is income or price elastic in the short term. The adjustment to a lower level of imports will involve some degree of expenditure switching through real exchange rate depreciation, but given the magnitude of adjustment required, it appears inevitable that expenditure reduction will bear a large part of the burden of adjustment. The shortage of foreign exchange is likely to tighten liquidity in the heavily dollarized commercial banking system and could force banks to scale back their lending. Secondly, the fall in remittances will reduce private disposable incomes, probably by around 10 percent; although this could be much less in nominal terms if there were substantial real exchange rate depreciation. 5 Lower private disposable income will in turn reduce private consumption and residential construction. Part of this reduction will fall on imported consumer goods, helping to bring about the adjustment required to ensure external balance, but there will also be lower demand for non traded goods. Private business investment is also likely to be scaled back because of the uncertain prospects for the economy and the tightening of liquidity in the banking system. This will further weaken demand for non traded goods such as construction activity. Although real exchange rate depreciation will induce some switch in consumer demand from traded 5 Remittances contributed about a third of private disposable income in 2008, hence if remittances fall by a third, private disposable income will fall by about 10 percent. 8 goods to non traded goods, if the scale in income reduction is large enough it seems unavoidable that there will be excess supply in the market for non traded goods, causing a fall in their production, or at the very least a marked slowdown from the high rates of real growth recorded so far in the 2000s. Price flexibility could dampen the extent of real output reduction in the non traded goods sectors, but this seems unlikely. Many features of the economy, such as the lack of competition in product markets, extensive regulation of business and the large share of public enterprises in the economy would suggest that price flexibility is limited. The economic crisis will affect living standards and poverty. As noted above, the fall in remittances will reduce private disposable incomes by about 10 percent. The growth in remittances contributed at least half of the fall of 19 percentage points between 2002 and 2007 in the incidence of poverty. This fall will be reversed. Because the population is clustered around the poverty line, there is a high elasticity of poverty to changes in real consumption. A 10 percent fall in purchasing power is estimated to increase the poverty headcount by 9 percentage points and the incidence of extreme poverty by 6 percentage points (Ivaschenko, 2009). The rise in poverty will be higher than this if the fall in remittances is accompanied by a contraction in real output, because the fall in real incomes will be even greater than 10 percent. A fall in incomes will also worsen food security, which in 2007 was estimated to affect 24 percent of the population. Budget cutbacks will add further pressure to the living standards of the poor unless pro-poor expenditure programs are protected. In Armenia, poverty is projected to increase by 5 percentage points between 2008 and 2010 as a result of lower remittances, domestic output and employment, especially in the construction industry, and higher prices (World Bank, 2009). A sharp fall in imports and a very large slow down in real growth will also adversely affect tax revenues, which were 20 percent below the budget targets in the first two months of 2009 in Tajikistan. A continuation of this underperformance throughout the rest of the year would lead to a shortfall in budget revenues of approximately 4 percent of GDP. This poses an acute dilemma for fiscal policy which, unless donors provide more 9 budget support, faces a choice between complying with the budget targets for the fiscal balance and domestic borrowing, which would require expenditure cuts of an equivalent amount to the revenue shortfalls, or alternatively protecting the budgeted expenditures (either fully or partially) which would require raising domestic borrowing. The first option would be damaging for two reasons. First, given the magnitude of the required expenditure cuts, it would be very difficult to protect all of the priority expenditures in the budget, including all of the social sector expenditures which comprise 42 percent of the State budget. Secondly, although cutting expenditures in line with the shortfall in revenues would leave the actual fiscal balance unchanged from that in the budget, a more relevant measure of the fiscal stance, from the standpoint of the impact of the budget on aggregate demand, is the structural fiscal balance and the associated fiscal impulse. Fiscal impulse measures, derived from the cyclically neutral budget, are designed to measure whether the impact of the budget on aggregate demand (i.e. the fiscal stance) is expansionary or contractionary. As defined by Chand (1993), this is shown below, where FI denotes the fiscal impulse, Y and YP are actual output and potential output respectively, G and T are government expenditure and revenue respectively, go is the base year ratio of government revenue to potential GDP and to is the base year ratio of revenue to actual GDP. FI = (G ­ goYP) ­ (T - toY) In effect, the fiscal stance is neutral if revenue moves in line with output but expenditure is smoothed over the course of the business cycle. If expenditure is cut in line with the endogenous fall in revenue, the fiscal impulse will be negative. Hence maintaining the budgeted fiscal deficit and borrowing targets in the face of lower output growth will have the effect of tightening the fiscal stance, thereby exacerbating the fall in aggregate demand and the recessionary pressures on output. Therefore, if governments want to avoid tightening the fiscal stance, expenditures should not be cut to match the endogenous reduction in revenues. This will require either higher public borrowing or additional grant aid from donors. Because of constraints on access to external commercial 10 capital, higher public borrowing will mainly comprise domestic borrowing and this will only be possible if there is macroeconomic and fiscal space, an issue which is addressed in section 3. It is also interesting to note how the economic crisis facing the low income CIS economies compares with the crisis facing middle income transition economies in Eastern Europe and Central Asia (ECA). There are both similarities and differences. Both low and middle income economies in the ECA region face a large external shock to which they must adjust by reducing absorption, but the specific nature of the external shock differs, as does the scope for policy responses. 6 The fuel exporting economies, which include two classified as low income (Azerbaijan and Uzbekistan) have suffered a large adverse terms of trade shock as a result of fall in the oil price in 2008. The non fuel exporters, both low income and middle income, have experienced much less volatility in their external terms of trade. The middle income transition economies (including the fuel exporters) are much more vulnerable to a capital account crisis than the low income economies. The former have much higher levels of short term commercial external debt, with the average levels of short term debt exceeding their international reserves. The latter have, on average, both lower external debt to GDP ratios than the middle income economies and very little short term commercial debt, in part because they lack access to international capital markets. Also in contrast to the middle income transition economies, the low income CIS economies have very little external debt owed by their banks and non financial private sectors. Consequently, the external shock facing many of the middle income transition economies arises primarily from a reversal of commercial external capital flows (a "sudden stop"), whereas the main contributor to the external shock facing the low income CIS economies is the fall in remittance incomes, as discussed above. 6 These paragraphs draw on the data and analysis in the economics note prepared by the Office of the Chief Economist, Europe and Central Asia region, World Bank, entitled "The crisis in Europe and Central Asia in perspective", 2009. 11 The economic crisis will worsen the fiscal balances in almost all ECA economies. However, most of the fuel exporters entered the crisis with strong fiscal surpluses because of the impact of high oil prices on oil related revenues and hence they have more fiscal space to use a more expansionary fiscal stance to stimulate their economies. As discussed in section 4, the lack of macroeconomic and fiscal space in the low income CIS economies will constrain the extent to which they can expand their fiscal stance. Some of the middle income non fuels exporting transition economies also lack the macroeconomic space for a more expansionary fiscal stance because their policy priority is to avoid exchange rate devaluation. This means that the burden of economic adjustment to the external shocks is being borne by drastic expenditure reduction in both public and private sectors. For these economies, exchange rate devaluation would have very damaging effects on the balance sheets of the private sector, which has extensive foreign currency denominated debts. For the low income CIS economies, where private sector foreign currency denominated liabilities are much lower as a share of GDP, there is more scope for using exchange rate depreciation to bring about the required adjustment to the external shock, which lessens the burden on expenditure reducing policies. 3. Is there macroeconomic and fiscal space to allow for larger fiscal deficits and domestic borrowing? There is macroeconomic space if resources can be mobilized for the budget without jeopardizing macroeconomic stability and fiscal space if this can be done without undermining fiscal sustainability. 7 3.1 Macroeconomic space 7 The concepts of fiscal space and macroeconomic space were first used to analyze the scope for expanding public provision of growth promoting and/or poverty reducing public goods in the context of longer term strategies to accelerate economic growth and to achieve the Millennium Development Goals (Heller, 2005). Perotti (2007) provides a critique of these concepts, noting that they are not well defined and add little that is new to the existing theoretical analysis of intertemporal fiscal constraints. 12 Two factors constrain the macroeconomic space for a larger fiscal deficit and higher domestic borrowing in Tajikistan: the monetary consequences and the balance of payments constraint. The domestic financial system in Tajikistan is too shallow to allow for non monetary financing of the fiscal deficit. Domestic currency broad money was only 11 percent of GDP at end-December 2008, the commercial banks have very limited domestic currency liquidity with which to fund purchases of government debt and the non bank financial sector barely exists. Consequently any domestic borrowing requirement of the government would have to be financed by the central bank. Monetary financing of the fiscal deficit would be very destabilizing for the macro economy for two reasons. The money supply has already grown very rapidly during the five year period 2003-2008 ­ reserve money grew by an average of 49 percent a year ­ in part because of the quasi fiscal activities of the NBT, and the NBT does not have effective monetary policy instruments to sterilize liquidity. 8 Because of the large fall in remittance transfers, the economy is likely to face a potentially severe balance of payments constraint. To some extent the excess demand for foreign exchange will be choked off by real exchange rate depreciation, but too large a depreciation will itself have adverse macroeconomic consequences because of its impact on the supply costs of firms who use imported inputs and on the debt burden of domestic borrowers whose debts are denominated in foreign exchange and thus also on the financial fragility of the banks who have extended loans to them. A larger fiscal deficit which is not funded with external resources risks further increasing excess demand for foreign currency and heightening the risk of a BOP crisis to which Tajikistan is already vulnerable. Table 3 presents data on some of the indicators which are commonly used for 8 Between 2004 and 2007, the NBT extended resources equivalent to about 16 percent of 2007 GDP to Kredit Invest (KI), a private sector financial institution which had been set up to finance the cotton trade. These resources took the form of credits in domestic currency, which directly increased the money supply, and the guaranteeing by the NBT of KI's external loans, mostly by pledging the NBT's own international reserves as security. KI defaulted on its external loan obligations as a result of which the NBT lost most of these pledged deposits in 2008 (hence the very low level of unencumbered reserves shown in table 3). 13 assessing the vulnerability of an economy to a BOP crisis, each of which point to a high degree of vulnerability. Table 3 Indicators of Vulnerability to a BOP Crisis in Tajikistan 2006 2007 2008 Current account balance as percent of GDP -2.8 -11.2 -8.8 Trade balance as percent of GDP -34.2 -48.2 -51.7 Inflation (percent, period average) 10.0 13.2 20.4 Growth rate of bank lending to the private 76.8 58.7 21.3 sector (percent) Unencumbered international reserves as 36 26 44 percent of reserve money Unencumbered international reserves in 0.7 0.5 1.2 months of following year's imports Foreign currency deposits as percent of total 68 76 64 deposits Sources: IMF and World Bank Tajikistan ran current account deficits of 11 and 9 percent of GDP in 2007 and 2008 respectively, despite the large remittance inflows which are classified as a current account transaction. The trade deficit in goods and services exceeded 50 percent of GDP in 2008. Inflation has risen to double digit levels. Private sector credit growth has been 14 very rapid, averaging 50 percent per annum during 2006-08. The NBT's unencumbered international reserves have been depleted (also because of the NBT's quasi fiscal activities) and stood at only just over one month of import cover and the equivalent of only 44 percent of reserve money at end-2008. Finally, the banking system is very heavy dollarized, with foreign currency deposits comprising 64 percent of total deposits. Table 4 Current Account Balances and International Reserves in Armenia, the Kyrgyz Republic and Moldova, 2008 Current account balance as Gross international reserves percent of GDP in months of imports Armenia -11.4 3.9 Kyrgyz Republic -10.7 3.0 Moldova -16.7 3.2 Source: Current account balances, ECA database; gross international reserves, IMF Macroeconomic space in Armenia, the Kyrgyz Republic and Moldova is limited by BOP constraints, because these economies also face an adverse external shock which exacerbates an external position which was displaying signs of vulnerability even before the shock. Table 4 provides data on their current account balances and gross international reserves. Each of the three countries had a large current account deficit in 2008, in excess of 10 percent of GDP in Armenia and the Kyrgyz Republic and nearly 17 percent of GDP in Moldova. They held a larger buffer of gross international reserves than Tajikistan, between 3 and 4 months of import cover in each country, but this is not sufficient to fund sustained BOP deficits. 15 3.2 Fiscal space Fiscal space can be created from four potential sources: increased domestic revenue, more efficient public spending, increased public borrowing and the mobilization of budget grants from external donors. Increased revenue mobilization, although an important long term objective of fiscal policy in Tajikistan (Brownbridge and Canagarajah, 2008) is unlikely to be feasible in the short term without raising tax rates, which would be counterproductive in the context of the economic crisis because it would entail a negative fiscal impulse. Improving expenditure efficiency is also an important long term objective but requires institutional reforms and capacity building in the public service and so has limited short term relevance. Table 5 Tajikistan Fiscal Data, percent of GDP: average 2003-06, 2007 and 2008 Average 2003-06 2007 2008 Revenues 18.4 20.5 20.5 Grants 1.6 1.1 0.8 Total Expenditure 21.3 27.8 27.5 Current Expenditure 13.8 13.6 12.4 Capital Expenditure 7.4 14.1 13.9 Fiscal Balance -1.3 -6.2 -6.1 Net External Financing 2.8 6.6 6.4 Net Domestic Financing -1.4 -0.7 -0.7 Public and Publicly Guaranteed Debt 31.5 External PPG debt 33.7* 33.4 30.4 Domestic Debt 1.1 NPV of External PPG Debt 30.6* 18.8 16 Electricity Sector Quasi Fiscal Deficit 12.5 7.3 7.8** * End 2006 ** Calculated on the basis of a revised estimated of the long run average incremental cost of supply Sources: MOF and IMF for budget aggregates, IMF and World Bank 2009 for debt stock, World Bank for QFD Fiscal deficits in Tajikistan were low in the 2000s, although they rose in 2007 and 2008 to accommodate higher capital expenditure financed with concessional external loans (table 5). Nevertheless, the fiscal space for increased public borrowing is highly constrained. As discussed above, the domestic financial system does not have the capacity to fund deficits in a non inflationary manner; while Tajikistan's fragile external debt sustainability constrains its access to external finance. Although Tajikistan's external public and publicly guaranteed (PPG) debt indicators currently fall within all the thresholds for sustainability stipulated in the Low Income Countries DSA framework, debt sustainability is highly vulnerable to macroeconomic shocks, such as lower real GDP growth and real exchange rate depreciation. The most recent joint Fund-Bank DSA, completed in March 2009, concluded that "Tajikistan is at high risk of debt distress" (IMF and World Bank, 2009, paragraph 8). Fiscal space is also constrained by the potential long term consequences for the budget of quasi fiscal deficits (QFDs). In addition to those of the NBT already mentioned, the domestic energy sector incurs QFDs, estimated at about 8 percent of GDP in 2008 (table 5) and state owned enterprises also undertake quasi fiscal activities. The pension system will also start to incur deficits (benefits will exceed contributions) from 2010 onwards. The potential for QFDs to undermine fiscal sustainability, by imposing large future claims on the budget, reinforce the importance of the government adopting a very prudent attitude to the accumulation of public debt. 17 Table 6 presents data on fiscal balances and public external debt in Armenia, the Kyrgyz Republic and Moldova. The fiscal position in Armenia is strong, with low fiscal deficits and very low external public debt. Furthermore, their energy sectors QFDs have been largely eliminated. The Kyrgyz Republic and Moldova also had low fiscal deficits in 2007 and 2008 but their external public debt was much higher as a percent of GDP. In the DSA completed in December 2008, the Kyrgyz Republic was assessed to be at moderate risk of external debt distress but vulnerable to exogenous shocks. In all three countries domestic public debt is small. The Kyrgyz Republic also has large QFDs, with that of the electricity sector estimated at 7.9 percent of GDP in 2008. To summarise, Armenia has fiscal space for a more expansionary fiscal stance, but fiscal space in the Kyrgyz Republic and Moldova is constrained by high stocks of public debt and QFDs. Table 6 Fiscal Balances and External Public Debt as percent of GDP in Armenia, the Kyrgyz Republic and Moldova, 2007 and 2008 Fiscal Balance Fiscal Balance External Public Debt 2007 2008 2008 Armenia -1.5 -1.2 13.2 Kyrgyz Republic 0.0 0.3* 46.9 Moldova -0.2 -1.0 43.8 This is a cash budget balance: if unfunded expenditure commitments equivalent to 1.4 percent of GDP were included, the balance would be negative Source: ECA database 3.3 How can macroeconomic and fiscal space be created? Given the BOP constraint facing the Tajik economy, macroeconomic space to avoid cuts in government expenditure can only be created if external resources are mobilized. Given 18 the high risk of external debt distress, external resources should be in the form of donor grants. To allow the government maximum discretion to formulate fiscal policy measures which best meet the objectives of supporting the domestic economy and protecting the poorest and most vulnerable, donor grants should be in the form of budget support. 4. How should additional budget resources be used? If additional budget support can be mobilized for Tajikistan and other low income CIS countries, what is the best way to use these funds? 9 The fiscal policy options are to use these resources to cut tax rates or to mitigate the cuts to expenditure that would otherwise be necessary because of the fall in revenues. If the latter is the best option, government must then decide which components of the existing budget should be protected and what, if any, new expenditures might be added to the budget. This section discusses the criteria which should be used for determining how additional budget support should be allocated to best meet the two policy objectives identified in the introduction: supporting aggregate demand and protecting the poorest and most vulnerable. Supporting aggregate demand To mitigate the recessionary impact of the economic crisis, fiscal policy should boost demand for non traded goods. Therefore, fiscal policy should minimize expenditure on imports, both to prevent leakages of demand out of the economy and to avoid exacerbating the BOP constraint. This is likely to mean that government expenditures should be labor intensive, as this will reduce the use of imported capital goods. Second, 9 The Kyrgyz Republic received a grant of approximately $150 million in 2009 from Russia as part of a $3 billion package of assistance, most of which involves finance for capital investments in hydropower. 19 fiscal measures should act on aggregate demand with a minimum of delay; hence measures which can be implemented quickly are needed. Protecting the poorest and most vulnerable The criteria for fiscal policy to maximize protection of the poorest and most vulnerable are not as straightforward. In principle several different options could meet this objective but whether they would do so in practice depends in part on how they are implemented. Three areas of expenditure can be identified as potential priorities for fiscal policy during the economic crisis. First, existing programs which provide basic social services on which the poor rely (e.g. primary education and primary healthcare) should be protected from any cuts required by revenue shortfalls. Secondly, labor intensive public works could be undertaken to provide temporary employment for people from poor families who would otherwise be unemployed or underemployed. Thirdly, targeted social assistance programs could be set up, or existing programs expanded, to provide transfers in cash or kind to the poor; an example is a school feeding program in a poor area. The difficulty with both the second and third categories is that implementing them quickly and ensuring that they are well targeted might be difficult because of administrative capacity constraints in the public service. Long run fiscal sustainability Short term fiscal measures should not undermine long term fiscal sustainability, which is already fragile in Tajikistan because of the risk of external debt distress and the QFDs discussed in section 3. As any emergency budget support which can be mobilized from donors will be temporary, the fiscal policy measures which are funded with it must not entail long term commitments on the budget, either in terms of a higher wage bill, higher entitlements to transfer payments, lower tax rates or tax exemptions or new public facilities which are expensive to operate. 5. What specific programs should be included in the fiscal policy measures? 20 Based on the criteria discussed above, we can identify some candidates for inclusion in a program of fiscal policy measures which could be funded with budget support from donors. Cuts in direct or indirect taxes are not the most effective measures to boost the economy or protect the poor, although both the Kyrgyz Republic and Tajikistan have cut, or are planning to cut, tax rates. The Kyrgyz Government cut the VAT rate from 20 percent to 12 percent on January 1, 2009 and the Tajik President announced in April 2009 that the VAT rate would be cut from 20 percent to 18 percent, alongside a cut in corporate income tax from 25 percent to 15 percent. Tax cuts will have lower fiscal multipliers than well designed government spending programs, because the latter should minimize the use of imports, whereas consumers will spend some of their additional income on imports, especially if indirect taxes are cut as these are predominantly levied on imports. Three quarters of the VAT collected in Tajikistan is paid on imports, hence the boost to domestic demand from the VAT cut will be small. Indirect tax cuts will benefit the higher income groups disproportionately because their consumption spending, especially on goods subject to indirect tax, is higher than that of the poor. Direct tax cuts will provide few benefits to poor households dependent on remittance transfers because they are unlikely to be in formal employment and thus liable to pay direct taxes. The planned cut in the corporate income tax rate in Tajikistan is also likely to be ineffective. In principle it could give a boost to private investment, but the impact of the economic downturn and heightened uncertainty about the economy's prospects is likely to be much more important in determining investment decisions than changes in tax rates. Finally, tax cuts might prove politically difficult to reverse once the economic crisis is over. Protecting the existing pro-poor basic social services, especially primary health and education, should have first call on budget resources. These expenditure programs are generally labor intensive, use relatively few imports and, because there is no need to prepare new programs, should not entail any administrative delays. This option does not involve new permanent expenditure commitments and so should not threaten long run fiscal sustainability. Hence it meets all of the criteria identified in section 4. It would be 21 useful to establish some sort of mechanism to ensure that these priority expenditures are protected from cuts, possibly through a virtual fund within the budget. A virtual fund could also be used to attract budget support from donors, with additional budget resources from donors linked directly to the magnitude of protected expenditures within the virtual fund. In addition to this, the government should consider scaling up the repair and maintenance of public infrastructure and public buildings, for which there is a huge backlog of work to be done. An expansion of repair and maintenance, which is far more labor intensive than the construction of new buildings, can provide work for returning migrants who have acquired construction industry skills abroad. Scaling up repair and maintenance will not involve long term expenditure commitments and so will not undermine fiscal sustainability. Expanding the existing social assistance programs may prove problematic in Tajikistan because their coverage is poor and they are not well targeted (Heltberg, 2008). Instead it might be more useful to work with community organizations to provide specific goods where targeting is more feasible, such as school feeding programs. In Armenia, the Kyrgyz Republic and Moldova, scaling up the existing relatively well targeted social assistance programs might be more feasible. This would provide a vehicle for increasing direct transfers to the poor, thereby offsetting some of their loss in income. Scaling up of these programs could take the form of expanding their coverage and/or raising the value of benefits. New public infrastructure projects, especially large energy and transport projects, should not be prioritized in the current crisis. Most of these projects are import and capital intensive, and the administrative capacity to implement them expeditiously is lacking. These projects should contribute to long term development, but they have lengthy project implementation schedules and will not yield benefits for several years. Delaying these projects and using the funds for the sort of expenditure priorities discussed above will be 22 a more optimal allocation of budget resources to alleviate the adverse impact of the economic crisis. 6. Conclusions The low income CIS economies face a large adverse external shock in 2009 caused principally by a fall in remittance incomes from migrant workers in the Russian Federation and by lower export earnings. Their economies will be forced to adjust to this external shock by cutting imports, through a combination of expenditure reduction and expenditure switching. The former will unavoidably affect demand for non traded goods with the consequence that output growth will slow sharply and will probably be negative in 2009, while the cut in remittances and reduced output growth will reverse at least partially the recent falls in poverty headcounts. Budget revenue will be reduced sharply because of the fall in imports and economic growth. The economic crisis poses acute dilemmas for fiscal policy. If governments maintain their 2009 budget targets for fiscal balances and public borrowing in the face of steep shortfalls in revenues, expenditures will have to be cut by an equivalent amount to the revenue shortfall. This will tighten the fiscal stance, thereby exacerbating recessionary pressures in the domestic economy, as well as jeopardizing priority public expenditures such as basic social services and social assistance programs. The low income CIS economies lack the macroeconomic space to increase domestic borrowing to finance higher fiscal deficits, because their economies face a BOP constraint as a result of the external shocks. Furthermore, the Kyrgyz Republic, Moldova and Tajikistan also lack fiscal space because of their vulnerability to external debt distress and large QFDs. Consequently, avoiding damaging cuts to government expenditure will require additional budget support from donors. If donor budget support grants can be mobilized to fund fiscal measures to mitigate the adverse effects of the economic crisis, the priorities should be to protect existing 23 government expenditures in basic social services which contribute to the welfare of the poor. In addition, governments should consider implementing a crash program of labor intensive maintenance and repair of public infrastructure to provide jobs for returning migrants (most of whom have construction industry experience) and to clear the backlog of works in this area. Expanding the existing social safety net programs in Tajikistan is problematic because these programs are small and poorly targeted and as such do not provide a feasible vehicle for supporting the living standards of the poor during the crisis, even if funding for them were to be scaled up. In Armenia, the Kyrgyz Republic and Moldova, the existing social assistance programs are better targeted at the poor and could provide an option for using fiscal resources to support the poorest and most vulnerable sections of society. Any package of fiscal measures, even if funded with donor budget support, must avoid undermining long term fiscal sustainability by incurring permanent fiscal commitments which are not covered by permanent increases in budget resources. References Brownbridge, Martin and Sudharshan Canagarajah (2008), "Fiscal Policy for Growth and Development in Tajikistan" Policy Research Working Paper 4532, World Bank. Chand, Sheetal K. (1993) "Fiscal Impulse Measures and their Impact" in Mario I. Blejer and Adrienne Cheasty (eds) How to Measure the Fiscal Deficit International Monetary Fund, Washington DC. Heller, Peter (2005), "Understanding Fiscal Space", Discussion Paper 05/4, International Monetary Fund. Heltberg, Rasmus (2008), "Social Protection in Tajikistan", chapter 5 of Tajikistan Poverty Assessment Draft Report (forthcoming), World Bank. 24 IMF and World Bank (2009), Tajikistan: Debt Sustainability Analysis under the Debt Sustainability Framework for Low Income Countries. Ivaschenko, Alex (2009), "Poverty Trends and Poverty Profile in Tajikistan: 2003-07", power point presentation, World Bank. Perotti, Roberto (2007), "Fiscal Policy in Developing Countries", Policy Research Working Paper 4365, World Bank. World Bank (2007), "Tajikistan: Pension System Review and Directions for Reform", Human Development Unit, Europe and Central Asia region. World Bank (2008), "Migration in Tajikistan" Tajikistan Poverty Assessment Draft Report (forthcoming), World Bank. mimeo. World Bank (2009), "Armenia: Implications of the Global Financial Crisis for Poverty" Report no 47770-AM, Washington DC. 25