V Case Studies
- Richard Hemming, Axel Schimmelpfennig, and Michael Kell
- Published Date:
- May 2003
This section reports the conclusions from case studies of 11 financial crises in emerging market economies in the 1990s. In chronological order, these are Mexico (1994–95), Argentina (1995), Bulgaria (1996–97), the Czech Republic (1997), Thailand (1997), Korea (1997), Pakistan (1998–99), Russia (1998), Ukraine (1998–2000), Brazil (1998–99), and Ecuador (1999). The case studies can be found in Appendix V. These obviously do not cover all emerging market financial crises of recent years, neither are they a random sample. Rather, the cases have been selected to cover a range of different types of financial crisis, some of which have already been extensively analyzed, but others of which are less well known. Each case study covers the background to the crisis; a description of the way in which the crisis manifested itself—currency depreciation, banking system collapse, sovereign debt restructuring or default, the impact on output and inflation, and the immediate policy responses; the causes of the crisis; the evolution of various fiscal vulnerability indicators before, during, and after the crisis; and, finally, the fiscal impact of the crisis and the response of macrofiscal and structural fiscal policies. The main conclusions are organized under three headings: the fiscal causes of crises, fiscal indicators, and fiscal consequences of crises.
Fiscal Causes of Crises
Fiscal vulnerabilities were central in at least 6 of the 11 case studies. Table 5.1 summarizes the main causes, both fiscal and nonfiscal, of the 11 crises. In 4 of the cases examined, public sector solvency and liquidity problems culminated in sovereign debt default (Russia and Ecuador), or negotiated restructuring under the shadow of default (Pakistan and Ukraine). In the other two cases, Bulgaria and Brazil, persistent and growing fiscal deficits were a central cause of pressure on the currency. In nearly all these cases, real interest rates were high and increasing—reflecting the legacy of high inflation and low policy credibility—that increased the debt-service burden and made it very difficult to raise interest rates to defend the currency in response to capital outflows. The underlying structural fiscal problems varied across these countries, but cover all the usual suspects: a high share of nondiscretionary expenditure (e.g., interest payments, social transfers, and intergovernmental transfers), poor expenditure control, overreliance on oil-related revenues, weak tax administration, and a lack of political will to address these problems.
|Mexico, 1994–95||Argentina, 1995||Bulgaria, 1996–97||Czech Republic, 1997||Thailand, 1997||Korea, 1997||Pakistan, 1998–99||Russia, 1998||Ukraine, 1998–2000||Brazil, 1998–99||Ecuador, 1999|
|Main underlying vulnerabilities||Pegged exchange rate.||Currency board.||Heavy external borrowing.||Capital inflows/current account deficit.||Pegged exchange rate.||Pegged exchange rate.||Buildup of external and public debt.||Structural fiscal problems.||Persistent budget deficits.||Pegged exchange rate.||Pegged exchange rate.|
|Current account deficit.||Banking sector weakness.||Failure to carry out structural reforms and address governance problems.||Pegged exchange rate.||Capital inflows/current account deficit.||Over-leveraged corporate sector.||High interest rates, and growing debt-service burden||Maturity structure of public debt.||Reliance on short-term external financing.||Current account deficit.||Very high public debt, mainly external.|
|Banking sector problems; and weak supervision.||Reliance on short-term external finance.||Banking sector weakness.||Policy mix (loose fiscal policy, tight monetary policy).||Banking/financial sector weakness and weak supervision.||Banking sector weakness.||Pegged exchange rate.||Lack of progress in institutional reform and governance.||Increasing fiscal deficit.||Structural rigidities.|
|Public debt management.||Reliance on short-term external borrowing.||Reliance on short-term external finance.||Narrow production and export base.||Banking sector weakness.||Pegged exchange rate.||Maturity structure of public debt.||Weak macroeconomic policies.|
|How important were fiscal factors?||Contributory factor.||Important contributory factor.||Central.||Important contributory factor.||Marginal.||Governance problems.||Governance problems.||Fractious domestic politics.|
|Main fiscal factors||Switch into short-term government debt.||Growing deficit in 1994 raised sustainability concerns; above-trend growth in early 1990s masked a lax fiscal stance.||Doubts about fiscal sustainability; inability sufficiently compress discretionary spending.||Failure to tighten fiscal policy despite overheating.||Large contingent liabilities.||Weak banking sector.|
|Failure to tighten fiscal policy in 1994.||Bank financing deficits of provincial governments||Bailing out state-owned banks.||Off-budget spending and implicit liabilities.||Tax distortions encouraged corporate debt.||Marginal.||Central.||Central.||Central.||Central.||Important contributory factor.|
|“Hidden” fiscal expansion through lending by stateowned banks.||Intergovernmental fiscal relations: flawed revenue-sharing arrangements; excessive provincial wage bills.||Weak revenue collection and secular revenue decline.||Poor governance of stateowned enterprises; high government wages; and high proportion of transfers in total spending.||Large contingent liabilities.||High level of public debt, around half external.||Weak tax administration.||Poor debt management.||Doubts about commitment to fiscal adjustment led to debt sustainability concerns.||Heavy reliance on oilrelated revenues; extensive earmarking of revenues.|
|Long-term increase in government expenditure added to real exchange rate appreciation.||Expenditure rigidities; pensions and transfers to provinces accounted for two-thirds of federal expenditure.||History of directed lending and government bailouts, contributing to overinvestment.||Low tax ratio because of narrow tax base and weak administration.||Weak expenditure control.||Reliance on netting-out operations.||Expensive public pensions and wages; rigid expenditure structure; and high share of nondiscretionary expenditure.||Low tax yield from non-oil sector (weak administration, extensive exemptions).|
|High share of unproductive and inflexible expenditure in total government spending.||High share of nondiscretionary expenditure.||Widespread tax exemptions.||Inefficient indirect tax system.||Lack of central government control over expenditure.|
|Reliance on short-term financing for public debt.||Slow progress in expenditure rationalization.||Reliance on shortterm financing for public debt.||Fiscal factors contributed to the banking crisis: financial transactions tax; cost and sustainability of deposit guarantee scheme.|
|Parliamentary opposition to measures.|
Several of the crises demonstrate the importance of the structure of public debt as source of vulnerability to crisis. While debt-to-GDP ratios were very high in Bulgaria, Pakistan, and Ecuador, in other cases they were comparable to those in noncrisis countries and did not obviously give rise to solvency concerns. But in Mexico, Russia, Ukraine, and Brazil, heavy recourse to short-term and/or foreign currency borrowing increased vulnerability to interest rate and exchange rate shocks, and hence to liquidity or rollover problems. The cases of Russia and Brazil also show how liquidity problems can reflect long-standing fiscal problems, while the crises in Mexico and Ukraine demonstrate how this kind of vulnerability can develop over a short period of time.
Conventional deficit and debt measures can mask severe underlying fiscal vulnerabilities. In the cases of Mexico, Argentina, and the Czech Republic, conventional measures suggested that macrofiscal policy was sustainable in the years preceding the crises; indeed, this led many commentators to discount the role of fiscal vulnerabilities in these crises. But it is apparent—at least with hindsight—that fiscal policy was too expansionary prior to the crises. In the case of Argentina in 1995 and possibly Thailand, unsustainable growth rates meant that the underlying fiscal stance was significantly looser than it appeared. In Mexico and the Czech Republic, the government failed to tighten fiscal policy in the months leading up to the crises, putting an excessive strain on monetary policy and hence the banking system. In Mexico and Argentina there was also “hidden” fiscal expansion: through lending by state-owned banks in the former and through bank financing of provincial deficits in the latter.
The case studies show the different ways in which fiscal policy can contribute to banking sector problems. In Ecuador, the introduction of a financial transactions tax prompted large-scale withdrawals from the banking sector, and doubts about the credibility of deposit guarantees in the face of a deteriorating fiscal situation exacerbated the run on the banks. The main cause of weakness among provincial banks in Argentina was their persistent financing of the deficits of provinces. In Thailand, the government created incentives for banks and companies to borrow abroad through implicit and explicit guarantees and tax breaks. Finally, in Russia and Ecuador, the banks held large amounts of government paper that was rescheduled; this compounded the deterioration in their balance sheets, exacerbating the effects of the sovereign debt crisis on the financial sector and the wider economy.
However, fiscal vulnerabilities were not central to all crises. This contrasts to two other factors—having some form of pegged exchange rate and weaknesses in the banking and financial system—that are critical in explaining almost all the crises examined (with the possible exceptions of Ukraine and Pakistan). But even in Thailand, where the key vulnerabilities were rooted in the private sector, fiscal policy played a role. At the structural level, tax distortions and government guarantees resulted in borrowing that was too high and insufficiently hedged in terms of currency and maturity; at the macroeconomic level, the emergence of large contingent liabilities suddenly raised concerns about sustainability that increased capital account pressures.
Given the small sample size and difficulties of comparability, the analysis of fiscal indicators in the case studies is more impressionistic than systematic. With that caveat in mind, it is evident that deficit and debt variables gave, at least with hindsight, advance warning of impending problems in five countries. These are Bulgaria, Pakistan, Russia, Brazil, and Ecuador. However, to be useful as warning signals before the event, these indicators need to be compared to a country-specific benchmark or critical threshold, which in turn would need to depend, among other things, on the fiscal history of the country and of similar crises in other countries. However, in other countries, aggregate fiscal variables gave little or no warning of an impending crisis, for a variety of reasons. In Thailand and Korea, the key vulnerabilities were in the private sector; in Mexico and Ukraine, the main fiscal vulnerabilities related to the structure rather than the level of public debt; in several cases—the Czech Republic, Thailand, and Korea—conventional measures failed to capture off-budget activities and contingent liabilities. In other cases, conventional deficits were misleading because they did not adjust for the effects of unsustainably high growth rates (Argentina and Thailand) or because of data quality problems (Pakistan).
Indicators of the maturity and currency structure of public debt are clearly useful for signaling impending problems. This was the case in Mexico, Russia, and Ukraine. But, at present, the coverage and availability of such information varies widely across countries. There is also the problem of defining critical thresholds, and the possibility of apparently perverse developments, such as the lengthening of average maturity of public debt prior to the Russian crisis.
The same observations apply to indicators of structural fiscal vulnerabilities. Indeed, the only useful structural fiscal indicators found for the crises examined here are tax and expenditure arrears in the Czech Republic, Russia, and Ukraine. It is possible that similar indicators could be unearthed for other countries, but these are not easily available. Much work would be needed to construct and maintain a wider set of useful structural fiscal vulnerability indicators, particularly if historical and cross-country comparable series are to be used to determine critical thresholds.
Finally, market indicators typically give little or no advance warning of impending crisis. This confirms the conclusions of other studies that credit ratings in particular seem to be reactive to crises.
Fiscal Consequences of Crises
Crises had wide-ranging effects on public debt. In three cases, there was a large increase in the debt-to-GDP ratio within a year or so of the crisis (by around 20 percentage points of GDP in Mexico and Thailand, and around 15 percentage points of GDP in Korea). In four cases—Argentina, the Czech Republic, Pakistan, and Brazil—the crises resulted in a relatively modest increase in the debt ratio of around 5 percentage points of GDP. In the other four cases—Bulgaria, Russia, Ukraine, and Ecuador—the debt ratio had fallen below precrisis levels within two years of the crisis.
The impact on fiscal sustainability also varied widely. In five cases—Mexico, Bulgaria, Russia, Ukraine, and Ecuador—sustainability improved, albeit for different reasons and to varying extents. In Mexico the stock of debt increased, but became less exposed to currency and interest rate risk; in Bulgaria, Russia, Ukraine, and Ecuador, various combinations of strong fiscal adjustment, higher oil prices, debt restructuring, and an increase in growth improved the medium-term fiscal outlook. In three cases, the crisis had little impact on sustainability in the short run: in the Czech Republic and Korea this was because of the low precrisis level of debt (though subsequently sustainability concerns have emerged in the Czech Republic because of stagnant growth and growing deficits); the crisis in Argentina also had little immediate impact on fiscal sustainability, although it is clear with hindsight that the underlying fiscal problems were not resolved following that crisis. Finally, in only two cases—Thailand and Pakistan—did the crisis result in a deterioration in fiscal sustainability. Table 5.2 summarizes the impact of the crises on the fiscal situation in each country, and the fiscal policy response in terms of the fiscal stance planned and achieved in the first year following the crises and the structural fiscal reforms implemented.
|Mexico, 1994–95||Argentina, 1995||Bulgaria, 1996–97||Czech Republic, 1997||Thailand, 1997||Korea, 1997||Pakistan, 1998–99||Russia, 1998||Ukraine, 1998–2000||Brazil, 1998–99||Ecuador, 1999|
|Impact on debt, in percentage points (pps) of GDP, and on sustainability||Large increase (around 20 pps);sustainability subsequently improved.||Modest increase (around 5 pps); sustainability concerns reemerged later.||Modest reduction in 1997; larger reduction in 1998 due to growth and lower interest rates. Progress maintained since.||Modest increase (loan guarantees made explicit); since 1998, deficits and slow growth have led to sustainability concerns.||Large increase (20 pps); sustainability has become a concern.||Moderate increase (by 12 pps); no sustainability concerns.||Moderate increase (by 7 pps, between 1998–99 to 2000–01); continued sustainability concerns.||Modest reduction in domestic debt; large reduction in external debt (by 2000); strong growth and exchange rate appreciation also improved debt dynamics.||Initial moderate increase (12 pps) due to fall in exchange rate; subsequent reduction in debt ratio due to fiscal adjustment and growth; sustainability not a major concern.||Modest increase (6 pps); structure of debt improved.||Cost of banking crisis around 20 percent of GDP; debt restructuring and strong recovery in fiscal position reduced sustainability concerns.|
|Stance of fiscal policy following crisis||Original program: 1 pp tightening; first review additional 1 pp tightening. Outturn: 2.7 pps tightening.1||Original program: 2 pps tightening. Outturn: ½ pp tightening.1||Improvement in primary balance of around 5 pps.||1997 budget projected a 1 pp tightening. Outturn: unchanged deficit implying modest tightening.||Original program: 3 pps tightening. Outturn: 1½ pps tightening.1||Original program: 1½ pps tightening. Outturn: 1 pp stimulus.1||Little change in 1999–2000; modest tightening in 2000–01.||General government primary balance improved by 3 pps.||Significant consolidation in 2000; primary balance improved by 1½ pps.||Original program: 3 pps tightening. Outturn: 3 pps improvement in primary surplus.1||Original program: 2½ pps improvement in primary surplus. Outturn: 5 pps improvement in primary surplus.|
|How achieved/implemented?||Higher-than-expected oil revenues; delays to investment projects; cuts in social spending and transfers to stateowned enterprises.||Temporary value-added tax (VAT) increase and broad-based central government expenditure cuts.||Rebound in revenues (due to higher wages) and sharp decline in domestic interest payments offset by some recovery of noninterest expenditure.||Weak revenues (due to crisis and floods) offset by large discretionary spending cuts (current and capital).||Large discretionary measures insufficient to offset falling output: increased social spending and public works; tax deferrals.||Mostly discretionary measures: increased expenditure on social safety net, capital projects, and bank restructuring.||Expenditure compression (current and capital) in 2000–01.||Growth in tax base as a result of recovery; discretionary tax changes; improved tax compliance; expenditure compression at subnational level.||Discretionary expenditure curbs; enforcing cash payments and eliminating offsets; higherthan-expected growth.||In 1999, mostly at central government level (improved tax administration and expenditure control). In 2000, also at subnational level.||Higher oil revenues; improved tax administration; rebound in activity.|
|Structural fiscal responses||Tax administration reforms introduced 1996–97; targeted social safety net scheme introduced 1997.||No major reforms implemented; some minor reforms to provincial finances (privatization and transferring pensions to federal government).||Improved tax administration and expenditure management; tax base broadened, rates lowered, and other measures to address inflation-induced distortions.||Legal requirement to disclose hidden debts and liabilities; tighter rules for government guarantees.||Expansion of social safety net.||Expansion of social safety net.||Tax policy reforms; strengthened fiscal reporting and expenditure management.||Higher export duties; a shift of VAT, profit, and income tax revenues to federal level; improved expenditure control; elimination of offsets at federal level.||Enactment of modern budget code; improved indirect tax enforcement procedures; elimination of noncash offsets.||Fiscal Responsibility Law enacted May 2000; other reforms impeded by Congress.||Financial transactions tax eliminated; VAT rate increased; lower subsidies on fuel; proposal for new oil stabilization fund and Fiscal Prudence and Discipline Law.|
In terms of the response of fiscal policy to crises, the six countries where sustainability improved implemented a large fiscal tightening in the year following the crises. In each case this was based on improved tax collections, but not necessarily on expenditure cuts: Bulgaria used some of the room created by lower interest payments to increase social spending; Argentina and the Czech Republic implemented a modest discretionary tightening based mostly on expenditure reductions; and in Thailand and Korea, there was some offset from discretionary fiscal stimulus. In three of these countries—the Czech Republic, Thailand, and Korea—the initial fiscal tightening was reversed. In the Asian countries this followed the much larger-than-expected decline in output, and indications that fiscal tightening was reducing rather than bolstering confidence, both domestically and internationally. In the Czech Republic, the reversal came somewhat later, although also as a response to slower-than-anticipated recovery in growth.
The successful implementation of discretionary tightening in Mexico, Bulgaria, Russia, Ukraine, Brazil, and Ecuador contrasts with the difficulties of implementing an expansionary fiscal policy in Korea and Thailand. The latter was due to a combination of a tradition of fiscal conservatism, capacity constraints, and budget management problems. The experience of Korea, in particular, shows that the capacity for fiscal policy to support recovery following a capital account crisis depends on structural features of the tax and spending systems, as well as the precrisis debt-to-GDP ratio.
In many cases, the crisis proved to be a decisive catalyst for significant structural fiscal reforms. Notable in this regard are improved tax administration and expenditure management (Bulgaria, Russia, Ukraine, Brazil); a significant expansion of the social safety net (Thailand, Korea); and enhanced fiscal transparency (Czech Republic, Brazil). However, crises do not always result in the degree of structural reform necessary to address fundamental fiscal problems; Argentina is a case in point.