WPS6409 Policy Research Working Paper 6409 Subnational Fiscal Policy in Large Developing Countries Some Lessons from the 2008–09 Crisis for Brazil, China and India Shahrokh Fardoust V.J. Ravishankar The World Bank Development Economics Vice Presidency Operations and Strategy Unit April 2013 Policy Research Working Paper 6409 Abstract In response to the Great Recession of 2008, many than landlocked states; revenue moved in opposite national governments implemented fiscal stimuli directions in the two types of state in 2009. Where packages in 2009 and 2010 to prevent further declines fiscal stress varies widely across subnational entities, in aggregate demand and to jump start their economic central transfers alone cannot prevent pro-cyclicality recovery. Where subnational governments responded of subnational fiscal response to a recession. There is with fiscal contraction, as in the United States, the need for flexibility in subnational borrowing within a impact was muted; where states/provinces also expanded sustainable fiscal framework. Many Indian states were expenditures, as in China and India, the impact was able to maintain or accelerate their spending thanks to magnified. Increases in recurrent expenditure, which the additional borrowing permitted in 2009 and 2010. were made in Brazil and India, acted as short-term In comparison, limited borrowing capacity and lack of stimulants; additional public investment, as in China, flexibility in federal grants restricted the contribution appears to have had a more lasting impact on growth. of Brazilian states to fiscal stimulus. Legal prohibition Large developing countries typically exhibit high of subnational borrowing induced China’s provinces to interregional inequality in levels of development and finance additional investments through extra-budgetary global integration, resulting in differential magnitude borrowing by nongovernment entities, with significant and timing of the crisis impact. For example, coastal fiscal risks on account of contingent liabilities. states in India were affected more severely and quickly This paper is a product of the Operations and Strategy Unit, Development Economics Vice Presidnecy. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at sfardoust1@gmail.com and vjravishankar@gmail.com. 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 Subnational Fiscal Policy in Large Developing Countries: Some Lessons from the 2008-09 Crisis for Brazil, China and India Shahrokh Fardoust and V.J. Ravishankar1 March 2013 JEL codes: E22, E62, H5, H7, H12. Key words: Brazil, China, India, fiscal stimulus, infrastructure, federalism, economic crisis. Sector Board: Economic Policy. 1 Shahrokh Fardoust is former Director, Strategy and Operations, Development Economics, the World Bank, sfardoust1@gmail.com. V.J. Ravishankar is former lead economist, South Asia Region, the World Bank, vjravishankar@gmail.com. The authors would like to thank David Rosenblatt for useful comments on the initial outline of this paper as well as on its various versions. Rafael Chelles Barroso, Fabio Solar Bittar, Pablo Fajnzylber, Lili Liu, Dennis Medvedev, Klaus Rohland and Onno Ruhl reviewed an earlier draft and provided very useful comments. The views expressed here are the authors’ and do not reflect those of the World Bank, its Executive Directors, or the countries they represent. 2 Introduction The global economic crisis of 2008–09, with its epicenter in the United States, has been characterized as “one of the broadest, deepest, and most complex crises afflicting the world since the Great Depression� (Didier, Helvia, and Schmukler 2010). It resulted in a sharp deceleration of economic growth in both developed and developing countries. Between 2008 and 2010, emerging economies suffered declines in gross domestic product (GDP) growth rates that were as steep as in advanced economies; their recovery was quicker and stronger, however. Most low-income countries were less affected (World Bank and IMF 2010). The crisis fundamentally changed the landscape in the global economy. The immediate outlook is affected by the downside risk that the euro area will fall into recession. Recently released forecasts by the International Monetary Fund (IMF) and the World Bank warn that global recovery remains fragile and that many advanced economies are likely to face a long of period of slow growth, high unemployment, and significant excess capacity in key economic sectors as they face large fiscal deficits and high debt levels (IMF 2012; World Bank 2012). At the same time, the world economy is undergoing massive shifts, underpinned by the rapid rise of emerging economies such as Brazil, China, and India. In 2011, emerging and developing economies as a group accounted for nearly 50 percent of world GDP (in purchasing power parity [PPP] terms), up from 35 percent in the mid-1980s. They accounted for nearly 70 percent of world economic growth. Going forward, their weight in global production, trade, investment, and finance is likely to continue to rise. These economies, including their subnational governments, will play an increasingly important role in the global economy. National governments in most developed and developing countries responded to the global crisis with countercyclical fiscal and monetary policies. They adopted measures to expand aggregate demand for goods and services by increasing net government expenditure, at the cost of higher fiscal deficits and public debt, and kept interest rates in check through accommodative monetary policy, at the cost of increase in inflation in some countries. 3 However, national efforts at stimulating economic growth were partly annulled by expenditure contraction, tax increases, or both at the subnational level in some federal countries. The negative impact of the crisis on state and local government finances in the United States was large and potentially long lasting (Box 1). In general, subnational fiscal outcomes in large federal countries varied, depending on inherited levels of debt, historically evolved fiscal behavior, and the rules governing borrowing by subnational governments. Box 1 Working at Cross-Purposes: Intergovernmental Disharmony in the United States Subnational fiscal policy in the immediate aftermath of the 2008 crisis was pro-cyclical in many advanced economies. The problem was especially pronounced in the United States. While the federal government cut taxes, increased spending, and expanded its deficit to stimulate aggregate demand, states cut a broad range of spending items. Balanced budget rules limited their options when revenues fell, as a result of weaker economic growth, rising unemployment rates, and a sharp decline in housing prices. Total nominal spending by state governments fell by almost 4 percent in fiscal year 2008/09 and by more than 6 percent in 2009/10—the most rapid declines since World War II. California cut spending by about $31 billion in 2009/10, despite federal stimulus funds providing an additional $8 billion. There have been some cases of financial distress and subnational government bankruptcy (examples include Alabama; Harrisburg, Pennsylvania; and Vallejo, California), though average default rates for states and municipalities have remained low. Growing political pressure on the federal government to reduce its fiscal deficit means that states and municipalities will receive less federal financial aid in the future. Moreover, despite some recovery in revenues following moderate economic recovery, state and local tax collections are projected to be weak to moderate due to the lagged response to falling house prices. These trends imply additional discretionary expenditure cuts, particularly as entitlement programs (e.g. Medicare/Medicaid) continue to rise rapidly. Hence the pro- cyclical behavior of sub-national finances is likely to continue. Source: Jonas 2012 and authors. Recent research (Canuto and Liu 2010a,b) discusses the pressures on subnational finance arising from the global crisis and from the debt created by subnational units through special- 4 purpose vehicles (SPVs) to finance infrastructure investments, which carry “inherent� risks as they often circumvent borrowing limits and create contingent liabilities. Regression analysis using time series data through the 1990s for seven large federal countries (Argentina, Australia, Brazil, Canada, Germany, India, and the United States) reveals that subnational government finance is overwhelmingly pro-cyclical, implying that the behavior of state governments in the United States is not an exception but the general rule (Rodden and Wibbels 2010). Only Canada was found to be engaged in harmonized expenditure smoothing in that decade. In the recent crisis, other countries also harmonized their policies. China, with its combination of high savings and large fiscal space, provides an important example of a countercyclical subnational response (Ter-Minassian and Fedelino 2010; Fardoust, Lin, and Luo 2012). Expansionary subnational government spending played a key role in strengthening the overall impact of the stimulus and sustaining economic growth. A less dramatic example is India, where the central government eased borrowing constraints on state governments in 2009 and 2010. Global economic growth recovered in 2010 but slowed in 2011. Serious concerns have emerged that the relatively fast-growing economies are also beginning to experience a downturn. In view of these trends, the IMF’s 2012 assessments recommend a gradual fiscal adjustment, cautioning governments against drastic cuts in their deficits and spending, lest they exacerbate the contraction. This advice is based on the conclusion drawn from empirical studies that the “fiscal multiplier� is especially large during times of crisis. Over the past decade, subnational spending as a share of government spending increased in many federal countries (Ter-Minassian and Fedelino 2010). The change has important implications for the size of fiscal multipliers and their impact on the national economy and for policy/institutional reform for improved coordination across government levels in implementing a harmonized crisis response. This paper is a follow-up to a research project carried out at the World Bank on the macroeconomic impact of the Great Recession in developing countries. The results of that project, which focused on the macroeconomic events of the 2008–09 crisis and national 5 policy responses in 10 developing countries, are reported in Nabli (2010). The focus here is on the policy responses by subnational governments in India and a comparative analysis of the three largest developing economies (Brazil, China, and India). This paper reviews recent research and draws on the lessons learned from practical experience in advising policy makers in subnational governments in developing countries to examine the question of subnational fiscal response to the recent crisis. Conceptual Framework The widening of government deficits during a period of economic downturn is generally the result of two sets of factors: (a) automatic effects on revenues and expenditures (known as automatic stabilizers, because they counteract the decline in demand) and (b) discretionary countercyclical policy measures including tax cuts and public spending increases to stimulate aggregate demand. Revenues automatically slow down or decline when economic activity slows; some expenditures, such as unemployment benefits and other demand-driven social programs, automatically rise when employment slows or declines. The second set of factors is the outcome of deliberate fiscal policy response on the part of governments at the central, state/provincial, and local levels. The existence of fiscal rules and the stance of monetary policy (high or low interest rates) also affect the size of national and subnational deficits. The size of fiscal multipliers depends on the composition of fiscal expansion and the initial conditions of the economy (box 2). There is a broad consensus in the literature that fiscal expansion through higher spending has larger multiplier effects than expansion through a tax cut, because additional disposable income from tax cuts may not be fully spent. Some portion is saved, depending on the income level of the beneficiary of the tax cut; the additional saving may not create additional investment demand in a poor investment climate. An increase in government capital spending tends to have a larger multiplier effect than increases in recurring expenditures (Ducanes and others 2006). Within recurrent government spending, the multiplier effect may be smaller for salary expenditures than for the purchase of goods and services (because employees may save part of their additional income). If a large share of the government wage bill goes to clerical or midlevel staff with a high propensity to spend additional income on consumer durables—as, for example, in India 6 during 2009—an upward revision of salary scales could have a significant impact on private consumption demand in the short run. Box 2 How Large Is the Fiscal Multiplier? The effect of fiscal expansion on output is referred to as the fiscal multiplier. It is measured by the unit change in output or GDP per unit change in government expenditure, other determinants of output remaining unchanged: F = ∂Y / ∂G Where F = fiscal multiplier, Y = income or GDP, and G = government expenditure. Large standard errors of estimation have prevented researchers from pinning down precise estimates of the value of F. Recent estimates range from negative to more than one (Spilimbergo and others 2008). Calculating the fiscal multiplier is difficult because there is no simple way to control the “fiscal experiment� (Barro and Redlick 2011; Christiano, Eichenbaum, and Rebelo 2009)—that is, to separate the impact of changes in government spending from changes in other variables that simultaneously affect changes in income, through various channels. Estimates of fiscal multipliers depend on the methodology, period, initial conditions, and controls applied. Study of the experience in the United States suggests that deficit-financed state government spending has a lower multiplier than windfall-financed spending, because deficit-financed state spending tends to crowd out private consumption and investment (referred to as Ricardian equivalence). However, in periods of stagnant or declining consumption and investment demand, deficit-financed government spending does not crowd out private demand. The multiplier is therefore larger than during boom times. A survey of empirical evidence suggests that for the 2008–09 stimuli by the G20, the low set of multipliers was 0.3 for revenue, 0.5 for capital spending, and 0.3 for other spending; the high set of multipliers was 0.6 for revenue, 1.8 for capital spending, and 1.0 for other spending (Spilimbergo, Symansky, and Schindler 2009). For a sample of 102 developing countries, the one-year government spending multiplier resulting from borrowing from official creditors is estimated to be about 0.4 (Kraay 2012). In emerging markets, where fiscal space is large, the safety net shallow, and infrastructure deficits great, the effects of 7 fiscal policy are greater than in advanced economies (Aizeman and Jinjarak 2010). However, the cumulative (multi-period) multipliers tend to be smaller for emerging markets than for advanced economies as the positive impact of an increase in government expenditures on GDP tends to fall off quickly (Ilzetki, Mendoza, and Vegh, 2011). Also, the size of public spending multipliers tends to be much smaller (zero or even negative) in highly indebted countries due to crowding out of private investment. Source: Fardoust, Lin and Luo (2012), and review of literature by authors. Some researchers find that expenditure on infrastructure investment and transfers targeted at the poor have the largest fiscal multipliers among all categories of government spending (Horton, Kumar, and Mauro 2009). Other researchers point to the inherently medium- to longer- term nature of infrastructure investments. Implementation takes time, and due processes need to be followed in appraising and selecting investment projects. Government revenues at all levels generally follow a procyclical pattern: they automatically grow faster during economic upswings and slower during downturns. For a particular subnational government, revenue growth depends on both (a) its own tax revenue base, related to GDP within its region, and (b) the transfer of resources from the central government, in the form of a share of central taxes, central grants, or both. The distribution and timing of the impact of the economic slowdown on revenues depends on the division of taxing powers between the levels of government. In countries where states/provinces rely mainly on their own revenues (for example, the United States), subnational government revenues could vary because of differences in regional economic growth. Where subnational governments rely largely on central transfers (for example, Mexico), the revenue impact is more uniform across regions. In this case, subnational government revenues are driven mainly by GDP at the national level and its impact on central revenues. The situation is more complicated when there is large variation across subnational governments within a country in terms of their dependence on central transfers (for example, Brazil, India, and to a lesser extent, China). In the countries of the Organization for Economic Co-operation and Development (OECD), subnational government revenues react 8 later to economic cycles than do the revenues of the central government, with a lag of about one year (Blochliger and others 2010). Total resources available to be spent by a state/provincial government are the sum of its total revenues and net borrowing capacity. Whether a government spends to its full capacity, however, depends on its constitutional mandate to spend on specified goods and services and the incentives governing its spending decisions, including fiscal responsibility laws, borrowing rules, and rewards for fiscal prudence. Although the subnational government Box 3 What Determines Subnational Government Spending? The structure and institutional framework of subnational finances differ markedly from national government finance in at least two important ways. First, because expenditures are more decentralized than revenues in many countries, most subnational governments rely on intergovernmental transfers, revenue sharing, or both for a significant part of their total revenues. Second, many subnational governments operate under balanced budget rules and face strict limits on borrowing from domestic or external sources (external sources generally come through the central government). In general, a subnational government spends in pursuit of developmental outcomes in its own jurisdiction and to contribute to national economic growth. It may or may not link its own spending with output growth in its own jurisdiction. Because subnational economies are open to trade within the country, a state/province’s own multiplier (∂Yi/∂Gi) will not be large, and cross-multipliers (∂Yi/∂Gj, where i is different from j) will be significant. Irrespective of the size of its own multiplier, a second-tier government in a large federal country is interested in achieving its targeted ∂Gi for developmental reasons. Let R denote government revenue, G government expenditure, and FD the fiscal deficit. For a subnational government that faces a limit FD* on its deficit because of its borrowing constraint and seeks to spend a desired or budgeted G*, G = G*, if G* < R + FD* (additional borrowing compensates revenue shortfall)a G = R + FD*, if G* > R + FD* (expenditure constrained by available financing). 9 The fiscal policy response of subnational governments will thus be procyclical during a downturn as long as government revenues are procyclical, unless (a) there are additional compensating transfers from the central government, (b) withdrawals from contingency (rainy day) funds, or (c) adequate flexibility in the borrowing rules for subnational governments to respond to cycles within sustainable limits. a. The condition is that R*-R, the revenue shortfall, is less than FD*-(E*-R*), the additional borrowing capacity, which translates into R-R*>E*-FD*-R*. Dropping R* and reversing the order yields E* Growth a/ -> Growth a/ National Defence 409.9 482.5 6% 8.0 12.6 42% Public Security 64.9 84.6 17% 341.1 389.8 3% Education 4.9 5.7 4% 851.9 987.0 4% Science & Technology 107.7 143.4 20% 105.2 131.1 12% Culture, Sports & Media 14.1 15.5 -1% 95.5 123.8 17% Safety Net & Employment 34.4 45.4 19% 646.0 715.2 0% Medical & Health Care 4.7 6.4 22% 271.0 393.1 30% Environment 6.6 3.8 -49% 138.5 189.6 23% Community Affairs 1.4 0.4 -76% 419.2 510.4 10% Ag, Water & Forestry 30.8 31.9 -7% 423.6 640.2 36% Transportation 91.3 106.9 5% 144.1 357.8 123% All Others 563.6 599.2 -4% 1480.8 1653.8 0% Total Gov. Expenditure 1334.4 1525.6 3% 4924.8 6104.4 11% a/ using implicit GDP deflator Source: National Bureau of Statistics, China. Note: Real growth rates were calculated using an implicit GDP deflator. They included investments of about Y1.2 trillion by the central government and Y2.8 trillion of supporting investment projects and programs by provincial and county governments and nongovernmental entities, financed by loans from domestic banks. Infrastructure 20 development, post disaster reconstruction, and housing guarantees made up almost 75 percent of the stimulus package. The fiscal balances reflect only transactions on budget; they omit public investments financed by the domestic banking system off budget, a large component in China. The increase in expenditure on budget was heavily concentrated at the subnational level (table 5). Among subnational expenditures, the steepest increase in 2009 was for transport infrastructure, followed by “agriculture, water, and forestry.� The two categories accounted for 36 percent of the increase in subnational expenditures. Health and education spending accounted for 22 percent. Table 6 Real Annual Changes in Capital Spending in Indian States(percent) Percent of total spending by all states in State 2010 2003–07 2009a 2010 2008–10 Andhra Pradesh 7.8 24.3 5.7 –15.2 –5.3 Bihar 7.1 17.9 9.6 23.1 16.1 Gujarat 5.4 17.5 –25.6 14.2 –7.8 Haryana 2.9 25.2 21.1 –32.9 –9.8 Karnataka 8.2 21.7 2.7 5.2 3.9 Kerala 2.3 14.4 2.0 23.4 12.2 Madhya Pradesh 8.8 15.5 27.9 3.9 15.3 Maharashtra 13.6 15.0 –13.7 –3.2 –8.6 Odisha (Orissa) 2.4 12.3 –15.6 18.2 –0.2 Punjab 1.0 30.9 –70.8 255.9 2.0 Rajasthan 2.7 14.5 –15.9 9.1 –4.2 Tamil Nadu 10.7 27.3 –42.2 111.8 10.7 Uttar Pradesh 11.4 27.3 7.0 –14.0 –4.1 West Bengal 4.6 8.3 –27.8 3.4 –13.6 All states 100.0 19.4 0.0 5.0 2.5 Sources: Reserve Bank of India state finances; individual state accounts; Central Statistical Organization. a/ 2009 refers to fiscal year beginning April 1, 2009. In India, pre-crisis fiscal stimulation measures—including increases in the food and fertilizer subsidy, a central pay hike, a farm debt waiver, and a national rural employment guarantee 21 program—amounted to 1.9 percent of GDP. This stimulus was complemented by three packages in three consecutive months starting in December 2008, each valued at about 0.6 percent of GDP. The aggregate fiscal stimulus amounted to 3.6 percent of GDP, consisting largely of expenditure measures (2.9 percent), including additional borrowing permitted for the state governments. In addition, there was the automatic impact of the decline in government revenues by 1.9 percent of GDP in 2009 and another 0.5 percent in 2010. Current expenditures accounted for more than 85 percent of expenditure measures in India. Only a minority of the major states also increased capital spending. In all states taken together, real capital expenditure remained unchanged in 2009 and rose by 5 percent in 2010, less rapidly than national GDP (table 6). In Brazil, fiscal expansion at the national level was partly counteracted by fiscal contraction by subnational governments in 2009. The fiscal deficit of the central government expanded 3 percentage points of GDP, reflecting both the operation of automatic stabilizers and discretionary stimulus measures, including temporary tax reductions and spending increases. However, in 2009 the fiscal deficit of subnational governments (states and municipalities combined) contracted 1.3 percentage points of GDP, turning into a small surplus. The net effect was a fiscal expansion of only 1.7 percentage points. Table 7 Real Annual Changes in Subnational Government Expenditures in Brazil, 2003–09 (percent) Expenditure 2003–07 2008 2009 Education and culture 4.2 10.2 3.6 Health and nutrition 8.7 8.5 9.0 Social programs 8.1 20.0 14.1 Transport 8.1 21.7 15.0 Housing and urban 5.8 22.2 –7.9 Capital investment 7.5 33.8 0.1 Sources: National Treasury Secretariat, Government of Brazil; and authors’ calculations. As part of its policy response to the global crisis, the federal government raised the permissible borrowing limits for those states that met agreed fiscal targets; taken together, the 22 ceiling for states’ borrowing was raised by Real 11 billion in 2008 and by an additional 17 billion in 2009. However, the net outstanding debt of all states combined remained unchanged in 2009, indicating that those states that were permitted to borrow more did not do so. Although spending on social programs continued to grow, housing and capital projects suffered (table 7). The federal fiscal deficit narrowed in 2010 by 2 percentage points of GDP, while the deficit of subnational governments expanded by only 0.4 percentage point. States’ outstanding debt rose but less than the permitted limit due to the time interval needed to contract and disburse a loan after the increase in borrowing room is allowed. The composition of fiscal stimulus was similar in India and Brazil, dominated by current expenditure, including salary payments. In contrast, infrastructure investment dominated in China. Economic growth recovered in 2010 in all three countries after having declined in 2009. In Brazil, the uptick was caused largely by the prompt monetary policy response to extend credit through public financial institutions to make up for the collapse in private credit flows. However, the recovery has been short lived, as economic growth decelerated during 2011 and 2012, more steeply in India and Brazil than in China (see figure 2). The multiplier effect of fiscal expansion was largest in China, where the investment share was highest. This is consistent with the empirical finding that public investment expenditure has stronger and longer-lasting impact on economic growth than recurrent expenditure, as shown by recent estimates of fiscal multipliers for China, India, the G7 and G20 groups of countries (table 8). It should be noted that size of multiplier tends to be sensitive to the level of debt, with fiscal multiplier being close to zero or even negative in highly indebted countries (Ilzetzki, Mendoza, and Vegh, 2012). Recent analysis points to two factors that contributed to the unusually large multiplier effect of China’s fiscal stimulus: “(a) the large marginal effects of government spending on investment, consumption, and in medium to long run on net exports and (b) the strong countercyclical nature of subnational government expenditures� (Fardoust, Lin, and Luo 2012, page 12). 23 Table 8 Estimates of the Size of Fiscal Multipliers in Brazil, China, India, and the G7 and G20 Countries Entity Period Size of multiplier Source Brazil Post 2008 (model 1.2 (public spending) Araujo, Azevedo, and simulation) 0.9 (tax revenue) Costa (2012) China 2008 global crisis Short term: 0.84 He, Zhang, and Zhang Medium term: 1.10 (2009) India 1980–2010 Short run (total expenditure): Reserve Bank of India 0.55 (2012) Long run (capital expenditure): 2.41 G7 Since mid-1970s High set: 0.6 for revenue, 1.8 IMF (2012) for capital spending, 1.0 for other spending G20 2008 global crisis Low set: 0.3 for revenue, 0.5 Spilimbergo, for capital spending, 0.3 for Symansky, and other spending Schindler (2009) Source: Authors’ compilation. The overall budget deficit expanded in all three countries during 2009 and 2010 (table 9). The fiscal expansion took place largely at the subnational level in China and India; subnational fiscal response was pro-cyclical in Brazil. 4 China financed its stimulus package largely through expansionary monetary policy. A special program of long-term bank loans at concessional interest rates provided capital for investment projects undertaken by public enterprises with subnational government sponsorship. Some World Bank analysts (Vincelette and others 2010) have noted that failure of these capital projects to generate enough revenue could lead to bank failures or impose a heavy fiscal burden on central or provincial governments. The annual reports from the five largest commercial banks show 17.6 percent annual growth in the loan categories used for this purpose (Wolfe 2012). Substantial financing was also made possible by using local land 4 This is consistent with findings by Sturzenegger and Werneck 2008. Their results imply that spending of subnational governments in Brazil have been markedly procyclical (1992-2002) and that it can be largely attributed to the highly procyclical pattern of tax revenues directly collected by subnational governments. 24 as collateral, with the center’s permission, raising concerns about the fiscal impact of a real estate bubble. Table 9 Central and Subnational Overall Fiscal Balance in China, India, and Brazil 2008–10 (as percent of GDP) (Country/level of government 2008 2009 2010 China Central 6.0 5.2 6.7 Subnational –6.6 –8.3 –8.3 General –0.7 –3.1 –1.5 India Centrala –6.1 –6.6 –6.1 Subnational –2.5 –3.3 –3.2 a General –8.7 –10.0 –9.4 Brazil Central –0.2 –3.1 –1.3 Subnational –1.2 0.1 –0.3 General –1.3 –3.0 –2.7 Sources: Official national data sources; IMF 2012. a. General government balances are as reported by IMF, with central deficit duly adjusted. In sum, the experience of China, India, and Brazil in responding to the global crisis of 2008 presents a mixed picture and raises various country-specific concerns. China scores high in terms of the composition of the fiscal stimulus and its multiplier effect on the economy; it scores low in terms of prudent financial management, with excessive reliance on off-budget domestic bank financing. India and Brazil score low in terms of the composition of fiscal stimulus and its (low) investment content. India scores high in enabling state governments to contribute to the stimulus within the limits of a sustainable fiscal framework. Slack credit demand from the private sector was a factor that enabled Indian states to access domestic bank credit, which was a binding constraint for the states in Brazil. Although Brazil scores high in terms of protecting real spending levels on social programs in 2009, the procyclicality of subnational fiscal response, resulting from procyclical central transfers and limited access to credit, together imply that SNGs faced increasingly tight resource constraints since 2011. A study by the IMF (2012) compares pre-crisis (2005–07) and post-crisis (2008–10) trends in subnational government finance in Australia, Brazil, Canada, China, Germany, Mexico, and 25 the United States. It notes that regional variation is greater in the emerging economies than in the advanced economies of North America and Europe, increasing the challenge of enabling a countercyclical subnational fiscal response. The possibility that different regions experience the growth cycle differently (that is, their troughs and peaks occur at different times) calls into question the conventional wisdom that stabilization, including countercyclical fiscal policy, is best left to the federal government, with central transfers the main instrument for expanding or contracting subnational fiscal space to counter the cycle 5. Central grants may be used with inter-temporal flexibility, but their movement is generally uniform across subnational entities—that is, they are expansive for all or restrictive for all subnational governments in any particular year. This issue of the unevenness of fiscal stress and the inadequacy of central grants to address the problem is explored in the Indian case study below. Case Study of Indian States Annual economic growth in India had accelerated to 8.0 percent during 2002–07, up from 5.5 percent the previous four years. The surge was driven mainly by a boom in private domestic investment (which grew at an average annual rate of 17 percent), accompanied by a sharp rise in corporate profits and a surge in private capital inflows. A cyclical downturn in investment had already begun when the global crisis broke out in 2008 6. Exports, which had grown by more than 25 percent a year during the boom period, fell dramatically. Bank credit decelerated and external financing dried up, although foreign direct investment remained relatively buoyant 7. The pattern of growth experienced by individual states in 2008 and 2009 varied depending on their degree of global integration and differing composition of aggregate demand. Broadly speaking, among the largest 14 states, which account for 87 percent of India’s population and 85 percent of its aggregate GDP, one pattern is evident in the eight major coastal states and 5 This recommendation was made, for instance, by the 13th Finance Commission of India, submitted in 2009/10 and covering the period 2010–15. 6 The Indian fiscal year runs from April to March; 2008 refers to the fiscal year beginning April 1, 2008. 7 This summary is based on the detailed analysis by Dipak Dasgupta and Abhijit Sen Gupta in “India: Rapid Recovery and Stronger Growth after the Crisis�, Chapter 6 of Nabli (2010), 26 another in the six major landlocked states 8. The difference in the timing of the impact reflects the structure of interstate economic transactions in a country of wide regional diversity (figure 3). The decline in exports had an immediate negative impact on the coastal states, leading to a sharp deceleration of growth to 6.5 percent from 10.0 percent and higher in the previous three years. Growth deceleration was less steep at the national level, because growth in the landlocked states accelerated in 2008 to 8 percent, up from 7 percent the previous year. The negative impact of the global crisis caught up with these states only in 2009, when growth declined by a relatively mild 0.5 percent. Figure 3 Economic Growth in Coastal and Landlocked States of India, 2005–11 Real annual growth (percent) 12 Major coastal states 11 10 9 All states 8 7 6 Major landlocked states 5 4 2005 2006 2007 2008 2009 2010 2011 Source: Central Statistical Organization, Government of India. Note: Figures are based on gross state domestic product at constant 2004/05 prices. The landlocked states have a disproportionately large share of agriculture and production of nontradable commodities in their output (where “nontradable� refers to India’s external trade). Growth in these states is driven more by national demand than by export demand. 8 The eight major coastal states are Andhra Pradesh, Gujarat, Karnataka, Kerala, Maharashtra, Odisha (Orissa), Tamil Nadu, and West Bengal. They account for 47 percent of the population and 58 percent of national GDP. The six major landlocked states are Bihar, Haryana, Madhya Pradesh, Punjab, Rajasthan, and Uttar Pradesh. They account for 40 percent of the population and 27 percent of national GDP. 27 They could also be driven more by consumption than by investment demand, a hypothesis that is difficult to test with the available data. In states with large shares of agriculture in output, rainfall and floods also affect growth. The immediate impact of the global crisis was an absolute decline in labor-intensive exports in the coastal states, leading to significant loss of jobs. The cyclical downturn in domestic private investment further accentuated the impact. Drops in incomes in the coastal states dampened domestic demand and retarded output growth in the landlocked states, to a lesser degree and with a time lag. Although agricultural growth decelerated during 2009 and 2010, rising food prices improved the terms of trade for farmers, muting the impact on farm incomes. Table 10 Real Annual Changes in Gross State Domestic Product and Own Revenue in Major Coastal and Landlocked States in India, 2007–10 (Percent) Gross state domestic product (GSDP) Own revenues States 2007 2008 2009 2010 2008 2009 2010 Major coastal 10.0 6.5 8.1 6.8 9.3 -0.8 16.2 Major landlocked 7.1 8.0 7.5 9.0 -1.8 13.3 13.6 Alla 9.3 7.0 8.1 7.7 5.0 4.1 14.5 Sources: (i) Central Statistical Organization; (ii) Reserve Bank of India; (iii) Individual state budgets and finance accounts. a. GSDP growth rate is different from national GDP growth rate in table 1 because defense and central public administration are included there and excluded here. The movement in states’ own revenues reflected the difference in impact in coastal and landlocked states. In the coastal states, economic growth deceleration in 2008 resulted in the deceleration of own revenue growth in 2009. In the landlocked states, growth accelerated in 2008, and revenues rose in 2009. In 2008 and in 2009, states’ own revenue moved in opposite directions for these two groups of states (table 10). Comparing the postcrisis years (2009 and 2010) with the precrisis period (2003–07), annual growth in states’ own revenue decelerated only slightly, from 10.2 percent to 9.2 percent. This aggregate picture hides significant variation, however: own revenue growth declined by more than 5 percentage points in some coastal states (Andhra Pradesh, Odisha, Tamil Nadu, 28 and West Bengal) and rose by more than 7 percentage points in some landlocked states (Bihar, Haryana, and Madhya Pradesh). Central transfers, state expenditures, and fiscal deficits are not influenced by geographical location; hence the distinction between coastal and landlocked states is not relevant in analyzing the movement of these variables. Central transfers are made up of mandated shares in central taxes and central grants. Table 11 Real Annual Changes in Central Transfers in India, by State, 2009–10 (percent) Total central transfers Shared central taxes Central grants Precrisis Postcrisis State 2009 2010 2009 2010 2003–07 2009 and 2010 Andhra Pradesh –4.8 –3.7 10.4 –20.5 12.8 –5.1 Bihar –3.3 23.7 –10.7 20.4 15.0 7.6 Gujarat –2.9 6.4 –21.1 6.4 8.7 –2.5 Haryana –4.3 21.0 65.1 –4.0 10.7 17.4 Karnataka –4.6 20.5 37.1 –18.7 16.4 6.5 Kerala –1.5 0.8 –20.4 –15.2 12.5 –6.7 Madhya Pradesh –4.0 16.0 6.2 17.4 15.9 7.7 Maharashtra –6.7 33.9 –11.1 –3.4 32.6 1.0 Odisha (Orissa) 0.0 13.6 7.7 9.8 13.0 7.4 Punjab –2.2 34.5 30.1 –2.3 23.2 13.8 Rajasthan –5.3 28.2 –15.9 7.8 12.6 4.7 Tamil Nadu –4.6 20.5 –28.3 19.9 13.9 0.9 Uttar Pradesh –6.0 29.5 36.2 –14.2 17.3 9.7 West Bengal –4.3 18.1 –10.9 13.3 10.9 4.3 Small and special category states –4.5 30.4 20.9 –2.0 16.5 9.8 All states –4.7 21.2 8.0 –2.2 16.1 5.4 Major coastal states –4.5 15.6 –4.6 –2.3 14.5 1.4 Major landlocked states –5.1 25.9 12.7 1.9 17.9 8.6 Sources: Reserve Bank of India state finances; individual state accounts; Central Statistical Organization. The share of states in aggregate central tax collections remains fixed for five years at a time, as recommended by the Finance Commission. As 2009 was the last year of the award period of the 12th Finance Commission, all states experienced a similar movement in shared central taxes. In 2010, the first year of the new award period, when the formula for interstate distribution is determined afresh, growth over the previous year varied widely across states, ranging from close to zero to 30 percent (table 11). 29 Central grants are a combination of discretionary grants (for both general and specific purposes) and a small number of mandatory grants. For all states taken together, a real decline of about 4 percent in shared taxes was roughly compensated for by an 8 percent real increase in central grants in 2009, so that total central transfers remained almost constant in real terms. In 2010, the first year of a new award period, shared central taxes rose 20 percent in real terms, reflecting the fact that the 13th Finance Commission raised the states’ share. Central grants declined 3 percent; the combined effect was a real increase in central transfers of 9 percent, a larger increase than the change in real GDP. However, the real rate of growth in central transfers slowed, falling sharply from 14.5 percent before the crisis to 1.4 percent after the crisis in the major coastal states (table 11), which are also the states that suffered more from the declines in own revenues. Eight of India’s 14 major states, including both coastal and landlocked states, managed to maintain or accelerate their expenditure growth (table 12). An important factor underlying the growth in states’ expenditures was the timing of the adjustment in civil service pay scales, which occurs every 10 years. The award of the Sixth Central Pay Commission happened to be announced close to the time the global crisis broke out, raising the salary scale by about 20 percent on average. The impact on central spending was 0.5 percent GDP 9. As always, pressure by employees’ unions made sure that upward adjustment of central pay was followed by similar hikes in all states. The payments were made in 2009 and 2010, backdated to 2007, resulting in a large lump-sum addition to the disposable income of about 20 million central and state-level public employees, amounting to about 1.5 percent GDP. This adjustment contributed to consumption-led economic growth in the short run. In Bihar, Madhya Pradesh, Odisha (formerly Orissa), Tamil Nadu, and West Bengal, accelerated total expenditure growth was achieved through significant expansion of the fiscal deficit. These states used the additional borrowing room granted to all states in 2009 and 2010. The 12th Finance Commission had recommended that states contain their deficits to no 9 The monthly salary of central and state employees in public administration and publicly funded social services consists of two major parts. “Basic pay� is adjusted once every 10 years; the “dearness allowance� is enhanced annually, linked to consumer price movements and absorbed into basic pay once every decade. The decadal adjustments are based on the recommendations of a central pay commission that is constituted anew every decade. It submits its report to Parliament. 30 more than 3 percent of GSDP by 2009. Using its power to set annual borrowing ceilings for the states, the federal government effectively raised the states’ fiscal deficit ceiling to 3.5 percent in 2009 and 4.0 percent in 2010. Table 12 Annual Changes in Real Expenditures and Fiscal Deficits in India, by State, 2003–11 (Percent) Expenditure and net loans (real annual change) Fiscal deficit Fiscal correctiona (percent gross state domestic product State Precrisis Postcrisis [GSDP]) Precrisis Postcrisis (2003–07) (2009–11) (2009–10) (2010–11) (2003–07) (2009–11) Andhra Pradesh 7.9 –6.5 2.9 2.3 1.7 0.6 Bihar 6.1 12.8 3.0 4.0 4.3 –2.4 Gujarat 5.2 5.5 3.6 3.1 2.0 –0.3 Haryana 7.3 6.8 4.5 3.3 3.0 0.3 Karnataka 12.1 6.9 2.9 2.9 2.1 –0.1 Kerala 3.1 1.8 3.4 2.8 2.9 0.4 Madhya Pradesh 6.9 13.6 2.7 3.3 2.5 –1.1 Maharashtra 6.8 9.9 2.9 2.4 2.4 –0.6 Odisha (Orissa) 5.2 7.9 1.4 2.5 6.5 –2.3 Punjab 2.3 6.7 3.1 3.1 1.8 0.7 Rajasthan 5.4 0.4 3.9 2.3 4.1 0.7 Tamil Nadu 6.9 14.1 2.1 3.2 2.7 –1.1 Uttar Pradesh 10.8 6.6 3.6 3.8 1.5 0.8 West Bengal 4.4 7.7 6.2 4.4 1.9 –0.5 Small and special category states 8.1 14.2 All states 6.9 7.6 Major coastal states 6.7 6.3 — — — — Major landlocked 7.2 7.6 — — — — Sources: (i) Reserve Bank of India state finances; (ii) individual state accounts; (iii) Central Statistical Organization. Note: Base year is 2004/05. — not available. a. Figures show change in fiscal deficit to GSDP ratio; negative values indicate deficit expansion. A significant number of states made use of the extra borrowing room: 8 of 14 states ran deficits in excess of 3 percent of gross state domestic product in 2009 and 2010 (Table 12). Five states in 2009 and three in 2010 had deficits of at least 3.5 percent. These findings vindicate the policy decision of the central government, in the sense that the states that needed it could and did make use of the additional borrowing room to stimulate aggregate demand and national output. Increases in states’ aggregate expenditure between 2008 and 31 2010 accounted for 44 percent of the increase in consolidated government expenditure at the national level. The aggregate stock of states’ debt is less than half the stock of central debt (figure 4). Despite higher levels of fiscal deficit in 2009 and 2010, the ratio of states’ debt to GDP did not rise, thanks to the high nominal growth of GDP; deceleration in real output growth was more than compensated for by acceleration in inflation, which crossed 10 percent in 2010. The debt relief mechanism prescribed by the 12th Finance Commission, conditional on adherence to rule-based fiscal consolidation, helped states contain the ratio of their aggregate debt to national GDP well below the target of 30.8 percent recommended by the 12th Finance Commission for 2009. This ratio peaked in 2005 at 26.8 percent. It has declined to 23.5 percent in 2010. Figure 4 Outstanding Liabilities of Central and State Governments in India relative to National Gross Domestic Product, 2000–10 Percent of national GDP 70 60 50 40 30 20 10 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 State Governments Central Government Source: Indian Public Finance Statistics, 2011-12; Government of India, Ministry of Finance. Based on information available for 18 states, the Reserve Bank of India estimates the outstanding guarantees of state governments, the main form of contingent liabilities, at 2.8 per cent of GDP by end-March 2010, only marginally higher than the 2.7 per cent at end- March 2009 (Reserve Bank of India 2012). The 12th Finance Commission had recommended that all states impose a limit on these contingent liabilities through their Fiscal Responsibility 32 and Budget Management Acts and set up guarantee redemption funds (GRFs). Fourteen states have set up such funds, in which the outstanding balance was Rs 37 billion ($800 million) by end-March 2011, less than 10 percent higher than at the end of the previous year (that is, almost constant in real terms). Real growth in the mining and manufacturing sectors turned negative in April–June 2012. Public investment has not offset the decline in the rate of private corporate investment to any significant degree. The fiscal stimulus succeeded mainly in boosting public consumption and contributing to private disposable incomes through tax benefits. Growth in output of consumer durables, which fell from 14 percent in 2010 to 3 percent in 2011, recovered to 7 percent in April–June 2012; the production of capital goods declined by 20 percent that quarter (Reserve Bank of India 2012). The key findings of this case study can be summarized as follows: • Coastal and landlocked states in India experienced differential impact of the global crisis, with revenue moving in opposite directions in 2008 and 2009. • In the coastal states, central transfers did not compensate for the significant deceleration of revenues. Nevertheless, the majority of states were able to maintain or accelerate their spending growth, thanks to the additional borrowing room permitted by the center in 2009 and 2010. • States contributed 44 percent of the increase in the level of general government spending between 2008 and 2010. • The increase in public spending, at both the central and state level, was predominantly on current expenditures (salaries, subsidies, and social programs). • Public investment rose by less than a quarter of the steep decline in private corporate investment. • The impact of fiscal stimulus on the economy was significant but short-lived. • National economic growth is estimated to have fallen below 5 percent in 2012, for the first time in a decade. 33 • Despite a temporary rise in the fiscal deficit of both the center and the states in 2009 and 2010, debt and debt-servicing ratios are falling, thanks to high inflation and negative real interest rates. Policy Lessons China and India demonstrate that subnational governments can make a positive contribution to countercyclical fiscal policy where there is room to temporarily loosen their borrowing constraints, on or off budget. Brazil demonstrates the converse—that where subnational governments have limited borrowing room or access to credit, the fiscal response of subnational governments is likely to be pro-cyclical, dampening national attempts to stimulate aggregate demand and economic growth. An important lesson from the experience of China is that a large fiscal stimulus implemented in a coordinated manner, with all levels of government aligned to a single plan, can have a significant positive impact on both subnational and national economic growth. On the negative side, the financing of subnational spending through off-budget borrowing has raised the fiscal risk associated with inadequately accounted contingent liabilities. The risk would be manageable as long as economic growth remains high but not otherwise. With lending and investment driven mainly by government policy directives, there is also a concern that the economic rationale underpinning investment decisions may be compromised. An important lesson for China is that it is more prudent to rely on budget financing than on off budget loans should another round of stimulus become necessary. There is a case for introducing some room for provincial governments to borrow directly on budget, along with improving financial management and accountability at that level, including adequate monitoring of contingent liabilities. In India, the fiscal stimulus that was implemented in 2008–10 was less successful than China’s, with its impact not extending beyond 2010. The increase in public spending, at both the federal and state level, was predominantly on the current account; public investment rose by less than a quarter of the steep decline in private corporate investment. The multiplier effect was short-lived as a result. 34 An important policy issue for India is the need to allow greater flexibility in controlling the growth of the public salary bill, in order to create fiscal space for stepping up public investment. There is a need to move away from the archaic system of adjusting civil service salaries once every 10 years, which acts like a periodic exogenous shock that squeezes fiscal space for capital spending and non-salary recurring expenditures. A deficit-financed fiscal stimulus is not easily implementable in India if such a necessity were to emerge in the near future. The national government is already on a tight fiscal consolidation path, as are the majority of states, in line with the recommendations of the 13th Finance Commission concerning the fiscal framework during 2010–15. Looking forward, the focus has to be on the major pending reforms to expand fiscal space, such as by (i) introducing the proposed goods and services tax (GST), which would expand the tax net, (ii) rationalizing user fees in selected sectors, and (iii) rationalizing expenditures such as subsidies and centrally sponsored programs that have outlived their utility. In Brazil, the resilience of domestic consumption demand has been a distinct advantage in weathering the global crisis. Fortunately, there was not strong pressure for a fiscal stimulus in 2008–10. However, if conditions are less favorable in the future and the need for countercyclical policy measures greater, Brazil would be hard pressed to implement such a policy, given the present structure of fiscal federalism arrangements, with limited flexibility in central grants to subnational governments and limited capacity of subnational governments to undertake additional borrowing even if permitted. This paper focuses on the fiscal policy response to the global recession. Subnational governments also have important responsibilities toward medium-term development goals, which need to be protected at all times. China’s provinces need to enhance spending on social programs in order to reduce the need for rural households to save for unforeseen shocks, thereby increasing their propensity to consume. Indian states are responsible for a large part of the deficit in irrigation, road transport infrastructure, public health, and education. In addition to infrastructure, human capital, and environmental needs, subnational governments in most developing countries also need to consider demographic factors. In both China and Brazil, the rapid rise in the average age of the population over the coming decades implies additional fiscal pressures over the medium to longer term. 35 Given the expected increase in subnational spending requirements, it is essential for federal countries to establish mechanisms for periodic review and revision of their systems of intergovernmental resource-sharing and transfers. India’s Finance Commission, which is constituted afresh once every five years to assess the situation and recommend the formula for tax-sharing and transfers over the next five years, has some useful lessons for China, where ad hoc earmarked transfers and inadequate fiscal equalization remain major problems. In Brazil the pro-cyclicality of transfers and the constraints on states’ access to the credit market together call for a review of fiscal federalism arrangements, so as to refine them to enable SNGs meet their expenditure obligations in the future. 36 References Aizenman, J., and Y. Jinjarak. 2010. “The Role of Fiscal Policy in Response to the Financial Crisis.� United Nations, Development Policy and Analysis Division, New York. Almunia, M., Benetrix, A., Eichengreen, B., O’Rourke, K.H., and Rua, G. 2009. “From Great Depression to Great Credit Crisis: Similarities, Differences and Lessons.� NBER Working Paper 15524, November. Araujo, Carlos Hamilton, Azevedo, Cynthia, Costa, Silva. 2012. “Fiscal consolidation and macroeconomic challenges in Brazil,� BIS Papers No 67. Barro, R. J., and C. J. Redlick. 2009. “Macroeconomic Effects from Government Purchases and Taxes.� Department of Economics, Harvard University. Blochliger, H., Charbit, C., Campos, J.M.P., and Vammalle, C. 2010. “Sub-Central Governments and the Economic Crisis: Impact and Policy Responses.� OECD Economics Department Working Papers, No.752. http://dx.doi.org/10.1787/5kml6xq5bgwc-en Canuto, Otaviano. and Liu, Lili. 2010a. “Subnational Debt Finance and the Global Financial Crisis.� Economic Premise, The World Bank, May. Number 13. www.worldbank.org/economicpremise Canuto, Otaviano, and Lili Liu. 2010b. “Subnational Debt Finance: Make it Sustainable� in The Day After Tomorrow: A Handbook on the Future of Economic Policy in the Developing World, ed. Otaviano Canuto and Marcelo Giugale. Washington, Christiano, L., Eichenbaum, M., and Rebelo, S. 2009. “When is the Government Spending Multiplier Large?� Northwestern University, August. Clemens, J. and Miran, S. 2012. “Fiscal Policy Multipliers on Subnational Government Spending�. American Economic Journal: Economic Policy 2012, 4(2): 46–68. Didier, T., Helvia, C., and Schmukler, S. 2011. “How Resilient were Developing Economies to the Global Economic Crisis?� Policy Research Working Paper 5637, World Bank, Washington, D.C. Ducanes G., M. A. Cagas, D. Qin, P. Quising, and M. A. Razzaquel. 2006. “Macroeconomic Effects of Fiscal Policies: Empirical Evidence from Bangladesh, China, Indonesia and the Philippines.� http://www.ecomod.org/files/papers/1530.pdf. 37 Fardoust, Shahrokh, Lin, Justin. Yifu, and Luo, Xubei. 2012. “Demystifying China’s Fiscal Stimulus.� Policy Research Working Paper 6221, The World Bank. Horton, M., Kumar, M., and Mauro, P. 2009. “The State of Public Finances: A Cross-Country Fiscal Monitor.� IMF Staff Position Note. Jul IMF Staff Position Note. Jul. SPN/09/21. Ilzetzki, Ethan, Mendoza, Enrique and Vegh, Carlos. 2011, “How Big (Small?) Are Fiscal Multipliers?� IMF Working Papers, WB/11/52. International Monetary Fund 2009, “Macro Policy Lessons for a Sound Design of Fiscal Decentralization— Background Studies�. July 2009. International Monetary Fund. 2012. Fiscal Monitor, April. Appendix 3: The Impact of the Global Financial Crisis on Subnational Government Finances. Jonas, J. 2012, “Great Recession and Fiscal Squeeze at U.S. Subnational Government Level�, IMF Working Papers, WP/12/184. Kraay, A. 2012, “Government spending multipliers in developing countries: evidence from lending by official creditors�, Policy Research Working Paper 6099, The World Bank. Liu, L., and S. Webb. 2010. “Laws for Fiscal Responsibility for Subnational Discipline: International Experience�. World Bank Policy Research Working Paper Series. No.5409.Washington, DC: World Bank. Liu, L., and Pradelli, J. 2012. “Financing Infrastructure and Monitoring Fiscal Risks at the Subnational Level�, Policy Research Working Paper 6069, the World Bank. Nabli, M. (Editor). 2010. The Great Recession and Developing Countries, the World Bank. OECD 2012, Economic Outlook, Volume 2012/1. Reserve Bank of India, Annual Report 2011-12. Spilimbergo, A., Symansky, S., Blanchard, O., and Cottarelli, C. 2008. “Fiscal Policy for the Crisis.� IMF Staff Position Note. December. SPN/08/01. Spilimbergo, A., Symansky, S., Schlinder, S. 2009. “Fiscal Multipliers.� IMF Staff Position Note. May. SPN/09/11. 38 Sturzenegger, Federico, and Werneck, Rogerio. 2008. “Fiscal Federalism and Procyclical Spending: The Cases of Argentina and Brazil,� in Guillermo E. Perry, Luis Serven, and Rodrigo Suescun. 2008. Fiscal Policy, Stabilization, and Growth: Prudence or Abstinence? The World Bank. Ter-Minassian, T., and Fedelino, A. 2009. “Impact of the Global Crisis on Sub-National Governments’ Finances�, in World Report on Fiscal Federalism, Institut d�Economia de Barcelona. Ter-Minassian, T. and Jiménez, J. P. 2011. “Macroeconomic challenges of fiscal decentralization in Latin America in the aftermath of the global financial crisis.� 39