Economic developments in Hungary during 1995–97 illustrate one of the most remarkable cases of macroeconomic adjustment in Europe over the last decade. At the end of 1994, macroeconomic conditions in Hungary were worrisome by any standard: the external deficit had exceeded 9 percent of GDP for two years; net external debt had increased rapidly to more than 45 percent of GDP; the deficit of the general government had exceeded 8 percent of GDP; public debt exceeded 85 percent of GDP; and borrowing conditions, for both public and external debt, were deteriorating. Three years later, the external current account deficit had dropped below 3 percent of GDP, reflecting an upsurge in exports, and was financed entirely by foreign direct investment; net external debt in relation to GDP had declined to its lowest level in the 1990s; and major rating agencies had raised their rating of Hungary’s external debt to an investment grade.

Economic developments in Hungary during 1995–97 illustrate one of the most remarkable cases of macroeconomic adjustment in Europe over the last decade. At the end of 1994, macroeconomic conditions in Hungary were worrisome by any standard: the external deficit had exceeded 9 percent of GDP for two years; net external debt had increased rapidly to more than 45 percent of GDP; the deficit of the general government had exceeded 8 percent of GDP; public debt exceeded 85 percent of GDP; and borrowing conditions, for both public and external debt, were deteriorating. Three years later, the external current account deficit had dropped below 3 percent of GDP, reflecting an upsurge in exports, and was financed entirely by foreign direct investment; net external debt in relation to GDP had declined to its lowest level in the 1990s; and major rating agencies had raised their rating of Hungary’s external debt to an investment grade.

Hungary’s record of external adjustment is even more remarkable compared with the performance of other countries in the region, of which almost all experienced a sizable deterioration in their external accounts during the same period (Figure 2.1). Moreover, the external adjustment in Hungary did not involve a major output loss: while output growth remained well below potential (Chapter VI), it was positive both during 1995 and 1996, and is estimated to have accelerated briskly in 1997 (Table 2.1).

Figure 2.1.
Figure 2.1.

Change in External Current Account in Central and Eastern Europe, 1994–97

Source: Eastern European Data Base, European I Department, IMF.1 Change in the ratio between external current account balance and GDP between 1994 and 1997 (staff estimates for 1997).
Table 2.1.

Selected Economic Indicators

article image
Sources: Data provided by the Hungarian authorities; International Financial Statistics; and IMF staff estimates.

In 1993, includes imports of military equipment from Russia in lieu of outstanding claims by Hungary.

The official estimates of nominal GDP for years prior to 1995 have been adjusted upward by 0.94 percent to render them consistent with the official data for the following years.

The figures on public finance in this table refer to the “consolidated government” for which a consistent time series exists for the whole period (see Box 5.1 in Chapter V for a definition of consolidated government). Deficit figures for a more comprehensive definition, which includes also the profit position of the central bank and the expenditure of the state privatization agency, are presented in Box 2.1 in this chapter.

Overall balance, excluding net interest expenditure and other debt charges. The figures for 1994–97 also exclude profit transfers from the National Bank of Hungary and transfers to cover National Bank of Hungary losses.

The current account deficit is on a settlements basis and differs from the saving-investment balance, which is on a national accounts basis.

Including intercompany loans.

What is behind this turnaround in economic performance? To answer this question, a brief discussion of the factors underlying the 1993–94 crisis is necessary.

Roots of the 1993–94 Crisis

The external and fiscal imbalances that peaked in 1993–941 can be traced back to the output shock suffered by the country in the early 1990s as a result of the collapse of the Council for Mutual Economic Assistance (CMEA) trade, the breakdown of key economic relationships and institutions that had prevailed under communism, and the recession in Europe. During 1990–92, output dropped by a cumulative 18 percent and the subsequent recovery, while relatively subdued, quickly ran into a balance of payments constraint. Other transition economies in the region suffered from similar, or larger, output losses, with severe effects on export capacity. In Hungary, however, the effect of the shock on the external balance was exacerbated by the policy attempt to shelter the household sector as much as possible from the effect of the economic crisis (Kornai, 1997). This attempt is apparent in virtually all aspects of macroeconomic policymaking before 1995:

  • As the tax base shrank, public expenditure failed to adjust commensurately, and the fiscal balance (excluding privatization receipts) moved from a surplus in 1990 to a large deficit position (Table 2.1 and Box 2.1).

  • Reflecting soft budget constraints in public enterprises and generous increases in minimum wages, real wages declined less than output, and income distribution shifted from employers to households (that is, from a high-saving sector to a low-saving sector).

  • Monetary policy was relaxed during 1992, with real interest rates remaining negative until the fall of 1993.

  • Exchange rate policy was used to contain the increase in the consumer price index (CPI)—and the decline in real disposable income, given the rigidity in nominal wage growth; inflation remained fairly low, compared with other transition economies (Table 2.1), but at the cost of a sizable appreciation of the real exchange rate (Figure 2.2).

Figure 2.2.
Figure 2.2.

Effective Exchange Rates1

(1993 = 100)

Sources: Data provided by the Hungarian authorities; and IMF staff estimates.1 An increase denotes an appreciation.

A Comprehensive Measure of the Fiscal Policy Stance

To get a comprehensive definition of fiscal policy stance, the deficit figures for the general government (see Chapter V for a definition of the latter) are adjusted here for a number of factors.

article image

The general government deficit (a) is first adjusted for the component of the expenditure of APV Rt (the agency in charge of privatization) that has a macroeconomic impact (transfers to enterprises, investment expenditure, subsidies) plus the transfers to the general government not formally regarded as privatization receipts under (a). Second, the deficit is adjusted for miscellaneous items (c), the largest of which is the difference between cash and accrual position in the profit and loss accounts of the National Bank of Hungary.1 Third, the figures are adjusted for the difference between nominal and real interest payments on the consolidated debt of the general government and the central bank (d). This yields the operational deficit (e).

1 This adjustment is necessary to have a full consolidation of the balance sheet of the two institutions. As the profits of the National Bank of Hungary are transferred to the state budget, and as the budget covers the deficit of the National Bank of Hungary whenever there is a need, in principle the state budget already represents a consolidated balance sheet of the two institutions. However, the profit and loss account of the National Bank of Hungary that is used to compute the transfers to the budget (or the need for transfers from the budget) is compiled on an accrual basis. Thus, to get the consolidated deficit on a cash basis, it is necessary to adjust the government deficit for the difference between cash and accrual position in the National Bank of Hungary profit and loss account. Note that the consolidation between the National Bank of Hungary and the general government is important given the deterioration in the profit and loss of the National Bank of Hungary related to sterilized intervention (Nemenyi, 1996). It is also important because in 1997, the zero-yield “valuation account’—a credit of the National Bank of Hungary toward the government—was converted into an interest-bearing credit (Chapter IV).

The consequences of these policies for the saving-investment balance of the country were harmful. As real private consumption declined by only 1½ percent a year during 1990–94 (against an average output decline of 3½ percent), private saving in percent of GDP edged downward, at the same time that public saving collapsed (Table 2.2). The impact of the fall in domestic saving on the external account was initially cushioned by the fall in enterprise investment. But, as investment recovered in 1993–94, the current account imbalance widened to levels that were unsustainable, despite comparatively large nondebt capital inflows (foreign direct investment (FDI); Table 2.3).

Table 2.2.

Sectoral Saving and Investment Balances1

(In percent of GDP)

article image
Sources: Central Statistical Office; and Ministry of Finance.

Data based on the new System of National Accounts; staff estimates based on the old national accounts for 1990.

Excludes in 1993 imports of military equipment from Russia in lieu of outstanding claims by Hungary.

Gross investment includes net capital transfers to other domestic sectors.

Table 2.3.

Balance of Payments in Convertible Currencies

(In millions of U.S. dollars, unless otherwise specified)

article image
Sources: Hungarian authorities; and IMF staff estimates.

Excludes reinvested profits.

At current exchange rate.

Includes intercompany loans.

The dramatic worsening of the current account highlighted a long-standing weakness of the Hungarian economy: its high external debt ratio. Hungary’s external debt had increased rapidly during the 1970s and the 1980s. Despite a sizable drop in the early 1990s (reflecting a broadly balanced current account position and increasingly inward FDI, including privatization receipts), net external debt stood at about 35 percent of GDP at the end of 1992 (Table 2.3) and had climbed to 45 percent of GDP by the end of 1994 (a high level by international standards).

Against this gloomy macroeconomic background, some structural developments—which were to prove important to the subsequent recovery—had taken place. By the end of the 1980s, the price system had been liberalized to a large extent, with regulated prices accounting for only about 20 percent of the CPI basket in 1989. This is a key difference with respect to other transition economies—a difference that helps to explain why the transition in Hungary was not accompanied by a major inflationary outburst (Chapter VII). Moreover, the tax system was reformed during 1988–90 with the introduction of the value-added tax (VAT), personal income tax, and a new enterprise profit tax (Boote and Somogyi, 1991). A two-tier banking system was introduced in 1987 (National Bank of Hungary, 1994). Finally, important trade liberalization measures had been implemented by the beginning of the 1990s (Chapter IX). The most significant areas of progress during the first half of the 1990s were:

  • the enactment in January 1992 of a bankruptcy law; this law, while excessively restrictive (Chapter VI), led to the rapid elimination of a large number of loss-making public enterprises, releasing resources to other sectors of the economy;

  • the enactment in 1991–92 of modern commercial banking and accounting laws, which forced banks to provision against bad loans (thus highlighting the need for their further restructuring) and contributed to the gradual tightening of the budget constraint on enterprises; and

  • the surge of FDI, partly as a result of the privatization process (Chapter VIII) and incentives (including tax incentives) to the establishment of joint ventures; by the end of 1994 and since 1989, cumulative FDI in Hungary equaled the sum of investment in all other transition economies in eastern Europe, central Europe, and Estonia, Latvia, and Lithuania (the Baltic countries).

These steps helped to establish a basis for a supply response, particularly in the export sector. Nevertheless, in 1993–94, the reform process slowed down and eventually stalled, particularly in the fiscal area (Chapters V, VI, and VIII).

Economic Goals and Policy Strategy During 1995–97

The adjustment program initiated by the authorities in March 1995 had a clear and immediate objective: to bring the external deficit down to a manageable level, so as to avert the risk of an external crisis. But the overall strategy had more wide-ranging ambitions. The authorities intended to prime an exportled growth process capable of raising the growth rate of the economy to its potential. Achieving these goals required reversing some of the phenomena that had characterized the early 1990s: the profit share in income distribution had to be reversed to contain consumption and stimulate investment; the absorption of saving by the public sector had to be reduced drastically; and external competitiveness had to be restored to stimulate export growth. Reducing inflation (which had lingered in the 20–30 percent range during the 1990s) was not regarded as an immediate priority. On the contrary, it was recognized that a price hike was inevitable and was indeed instrumental in redressing the real imbalances that had developed in the previous year (particularly in income distribution; see below). However, the program aimed to ensure that the price acceleration would be temporary, and that the basis would be laid for a reduction in inflation to single digits in the medium term.

The strategy to achieve these objectives had a macroeconomic component and a structural component. The macroeconomic component had strong orthodox features (the up-front fiscal adjustment and depreciation), coupled with some heterodox features (the use of inflation to correct income distribution in the presence of nominal wage rigidity and the attempt to shield domestic monetary policy from external pressures through sterilized intervention). The structural component centered on the revival of the privatization process (to reduce external debt and increase efficiency), the downsizing of the public sector, the increased opening of the economy to foreign trade, and the long-term reform of public finances.

The Macroeconomic Component

A sharp, up-front, fiscal adjustment was inevitable. The fiscal imbalance was not sustainable, as it would have led to a steady increase in the public debt-to-GDP ratio (Chapter IV). But adjustment had to go beyond the mere restoration of fiscal solvency. There was a need to raise domestic saving to reduce the external deficit and finance the expected increase in private investment. The March 1995 adjustment package included several permanent measures (cuts in transfers to households, a reduction in the number of state employees, and a broadening of the social security contribution base), together with the temporary introduction of an 8 percent import surcharge on consumer goods, with an expected annual yield of about 1 percent of GDP (Chapter V). Moreover, the government decided to freeze other expenditures in nominal terms, with the result that the acceleration of inflation (see below) involved a real expenditure cut with respect to the projections underlying the 1995 budget.

The extent of the fiscal adjustment implemented in 1995 can be better appreciated by focusing on a comprehensive definition of the public sector, which includes the National Bank of Hungary (see Box 2.1) and by adjusting the figures for inflation. Indeed, in a year in which inflation was rapidly rising, the nominal deficit was boosted by the increase in nominal interest payments necessary to compensate holders of government paper for the erosion in the real value of principal. However, the operational deficit (that is, the deficit adjusted for the difference between nominal and real interest payments) fell by more than 3½ percentage points, while the primary balance improved by almost 4¼ percentage points.

Fiscal adjustment continued in 1996 and 1997, with the main goal of consolidating the results achieved, rather than of implementing a further major fiscal tightening. The fiscal stance was tightened again in 1996, but it was relaxed somewhat in the subsequent year. Altogether, in 1997, the primary balance had strengthened by 1 ¼ percentage points of GDP with respect to 1995, while the overall operational deficit had declined by almost 1 percentage point of GDP (Box 2.1).2 In addition to a further cut in the operational defecit, fiscal policy in 1996–97 achieved two main goals. First, some of the temporary measures that were part of the March 1995 package (including the import surcharge) were replaced by lasting measures. Second, downsizing of the public budget continued, as highlighted by further sizable declines in the public revenues and expenditures ratios, with the effect of stimulating private sector growth.3

The choice of the exchange rate policy was dictated by the need to recover the competitiveness lost during the first half of the 1990s, and to provide a nominal anchor to make sure that the inevitable price acceleration was only temporary. The forint was devalued by 9 percent on March 13, 1995, and at the same time, a preannounced crawling exchange rate band was introduced, thus abandoning the policy of discretionary revisions of the parities that had characterized the previous period. The band introduced was relatively narrow (+2.25 percent), so as to provide a clear indication of the authorities’ intentions. The rate of crawl was initially sizable (1.9 percent a month through June 30, 1995), so as to guarantee a further improvement in competitiveness, but was then gradually reduced.4

Why was the exchange rate preferred to other nominal anchors, and what were the implications of such a choice? In many respects, the choice of the exchange rate as the nominal anchor was inevitable. Traditionally, the exchange rate in Hungary, as in many other small, open economies, has been highly visible. As such, it represents an important focus for the formation of the private sector’s expectations. Moreover, the natural alternative, targeting some monetary aggregate, was made difficult by the instability of money demand, which was related to wide-ranging structural changes in the Hungarian financial and payments system (Box 2.2). The choice of the exchange rate as the nominal anchor had, however, an important consequence for the shape of economic policies during 1995–97. Given that the primary focus of the adjustment program was external, the rate of crawl could hardly be used aggressively to disinflate the economy: the possibility of faster disinflation was regarded as a poor compensation for the external account risks arising from tighter exchange rate policies. Thus, it is not surprising that the reduction in the rate of crawl between the second half of 1995 and the second half of 1997 was rather modest.

As mentioned, the adjustment program also had important heterodox features. The first relates to the mechanism through which the needed redistribution of income from employees to employers was brought about. Between the end of 1994 and the middle of 1995, inflation increased from about 21 percent to more than 32 percent, reflecting sharp adjustments in administered prices (particularly in the energy sector) and the depreciation of the forint. This inflationary outburst was to a large extent unexpected and had not been incorporated in the wage agreements for 1995, most of which had been concluded before the March stabilization program was announced. The main effort required was to make sure that the agreed wage increases were not revised on account of higher inflation. To this end, the authorities primarily used their influence over wage developments in the public sector and the public enterprise sector. In the public sector, the decision not to revise the budgetary outlays for 1995 discouraged decentralized spending units from granting wages in excess of those provided for in the budget. As to the public enterprise sector, still sizable at that time, instructions were issued to avoid a wage acceleration. There was no formal agreement with trade unions, though. The fact that the authorities’ tough wage policy was accepted without disruption of work activity shows the unions’ sense of responsibility and the fact that an adjustment in real wages was regarded as inevitable.5 Incomes policy agreements were instead reached in 1996 and in 1997 in the context of the Interest Reconciliation Council (IRC).6 In 1996, the IRC recommended an average increase in gross wages of 19½ percent, with a range of 13–24 percent to allow for different productivity growth across sectors. The agreement assumed an inflation rate of 18 percent. In the event, inflation was much higher (see below), while gross wage growth in the public sector (although not in the private sector) increased by less than recommended. Again, this contributed to the strengthening of fiscal accounts. The overshooting of inflation in 1995–96, however, made the 1997 wage agreement more difficult, and led authorities to announce a relatively unambitious inflation target for that year (see below).

How Stable Is Money Demand?

The decision to adopt an exchange rate rule in March 1995 rather than to target some monetary aggregate was based in part on the perceived difficulty of reliably forecasting money demand during a period of heightened macroeconomic uncertainty and structural transformation, in addition to the absence of a timely and accurate indicator of real economic activity.

Monetary developments since then have tended to support that decision. While the real value of M3 (currency, bank deposits, and bank-issued financial securities) remained relatively constant during 1992–94, real M3 declined sharply (by 17 percent) between the end of 1994 and early 1996, before recovering somewhat since then (see figure below). This dramatic reduction in financial intermediation is also apparent in each of the major components of M3. While this is not evidence of instability per se, the evidence lies in the fact that the decline in the monetary aggregate is poorly correlated with the level of real activity and interest rates. As a consequence, had a monetary target based on historical behavior been adopted during this period, the monetary stance would have been different than intended by the authorities.

The implications of the instability in money demand for monetary targeting can be seen using a multivariate error correction model (VECM) of the form:


where Z is a vector of the variables log(M3adj), log(prod), pi, RdL, Rtb3comp; log(M3adj) is the natural log of real M3, adjusted for interest accrued but not yet paid on deposits; log(prod) is the natural log of the volume of industrial production; pi is the 12-month inflation rate; RdL is the interest rate on deposits held for more than 12 months; Rtb3comp is the compound interest yield on 3-month treasury bills; and A indicates the first difference of the variable. An intercept and a linear time trend are included in the long-run model. All variables are seasonally adjusted.


Actual and Forecast M3 1

Sources: National Bank of Hungary; and IMF staff calculations.1 In real terms, seasonally adjusted, and for interest accrued but not yet paid.

Based on the data available through the time, the new exchange regime was introduced (December 1991 to December 1994) and Johansen’s methodology, a cointegrating relationship is found to exist between M3, industrial production, interest rates, and inflation:1


These results imply that a 1 percentage point increase in industrial production raises M3 by 0.3 percentage point; an increase in the interest rate on deposits increases M3, while an increase in inflation (which represents a cost of holding currency) and in the yield paid on treasury bills (which are a substitute for M3) lower M3. However, these relative price effects are found to be quite small, in line with the empirical findings of other researchers. Since the estimation period covers the initial period of banking sector disintermediation, money demand is found to have a negative trend component.

Using the estimates from the VECM formulation, it is possible to forecast the level of money demand during 1995. Had a monetary rule been pursued during this period, such forecasts would have been used to establish monetary targets under the program. One can see from the figure that the model performs poorly out of sample, providing forecasts of M3 during 1995 well above the actual level.2 The actual level of M3 is outside the two-standard deviation band around the forecast throughout the projection period. This would suggest that, had Hungary followed a monetary target at that time, financial policies would have been excessively lax.

The poor predictive power of the money demand model can be related to the difficulty of capturing the process of financial disintermediation through the inclusion in the equation of trend variables. In particular, a number of factors may have accelerated the financial disintermediation process in 1995. The increased presence of foreign banks in the domestic banking system led to substantial modernization of banking and payments technologies. These innovations included the introduction of (1) automated teller machines (ATMs) (which enabled Hungarians to reduce their average holdings of currency by lowering the transaction costs associated with withdrawal); (2) credit cards (which may have reduced households’ average deposit balances since purchases no longer needed to be financed by drawing down accumulated deposits); and (3) a more efficient interbank clearance system (which served to lower the demand for transaction money balances by limiting the time needed to effect payments, thereby reducing the amount of money in transit). Moreover, a wider range of non-bank savings instruments became available—and more accessible—during this period, including inflation-indexed government bonds, investment and pension funds, and equities.

One would hope that, by adding additional information, the predictive power of the model would improve. Unfortunately, based on a reestimation of the model through December 1995,3 actual money demand is substantially more than 2 standard deviations above the forecast level throughout all but the first quarter of 1996. This underestimation may be partially related to the omission of some relevant variables from the model.4 More fundamentally, it reflects the difficulty of predicting the effect of financial innovation and disintermediation through a trend component.

The relative constancy of real M3 in 1996 following several years of trend decline suggests that forecasting could become more reliable in the future.

This box was prepared by Rachel van Elkan.1 Two cointegrating vectors were found to be present. On the basis of theoretical priors regarding the signs and magnitudes of the elements of the vectors, one eigenvector was rejected.2 The inclusion of squared and cubic trend terms did not improve the forecasting performance of the model.3 The long-run cointegrating relationship for the period 1992:1 to 1995:12 is: L(M3adj)=α+0.31L(prod)+0.0026RdL0.0035Rtb3comp0.0061pi0.0044Trend.4 The cost of reserve requirements declined in 1996 (Chapter XI); insofar as this decline led to a spread between lending rates and treasury bill rates (and assuming that money demand is affected negatively by lending rates), the omission from equation (1) of the lending rate may explain an under-prediction in 1996.

The second heterodox aspect of the adjustment program was related to monetary policy. Over 1995–97, monetary policy strove to keep forint interest rates shielded from external pressures through sterilized intervention. The Hungarian experience shows that, under certain conditions, it is possible to maintain some monetary independence for a fairly long period of time even in the presence of a narrow exchange rate band.7 During 1995–97, the interest rate differential with the deutsche mark, adjusted for the announced rate of depreciation, always remained positive, initially at about 5–10 percentage points, and after the spring of 1996 in the 3–4 percentage point range (Figure 2.3). Taking into account domestic inflation, this was enough to keep real interest rates on government paper positive (at about 1–2 percent using ex post inflation rates—Figure 2.3—and probably higher using expected inflation rates). This policy succeeded for several reasons. First, constraints on capital movements, both inward and outward, have been in place, and were more stringent for short-term investment (Chapter XII). Although they may have been partially circumvented (as suggested by sizable errors and omissions in the balance of payments), they nevertheless helped to shield domestic monetary policy. Second, the limited liquidity of Hungarian financial assets reduced the attractiveness of speculative investment for foreigners, although probably not for residents.8 Third, the cost of sterilized intervention was lowered in Hungary by the fact that, in early 1995, the stock of external debt issued at high spreads was quite large: foreign exchange reserves purchased by the National Bank of Hungary in the market could be used for early repayment of external debt, with an implicit high yield, thus reducing the cost of sterilized intervention commensurately.9 Nevertheless, the extent of the foreign exchange intervention was impressive. It amounted to 80 percent of base money in 1995, and to a still sizable estimated 50 percent in 1997.

Figure 2.3.
Figure 2.3.

Interest Rates 1

(In percent)

Sources: National Bank of Hungary (NBH), Monthly Report; and IMF staff calculations.1 On a compounded annual basis.2 Nominal treasury-bill rate adjusted for the ex post twelvemonth CPI growth rate.3 Treasury bill relative to German interbank rate, adjusted for the monthly rate of crawl for the forint.

The Structural Component

The goals of the structural adjustment program in 1995–97 were to reduce Hungary’s external debt through privatization receipts from abroad, strengthen the supply response of the system, and consolidate the progress made in the fiscal balances.

A key step was the revival of the privatization process, which had stalled in 1994 (Chapter VIII). Privatization receipts during 1995–97 amounted to a cumulative 12 percentage points of GDP (against 8 percentage points in the preceding five years), more than 80 percent of which came from abroad. This had three beneficial effects on the economy. First, it facilitated the increase in productivity needed to sustain output growth in the medium term as foreign ownership injected fresh capital and know-how. Second, it contributed critically to the decline in public and external debt ratios, reducing the country’s exposure to external shocks. Third, it involved sizable savings for both the fiscal deficit and the external current account.10

Another important step was the further relaxation of constraints on external transactions. Important measures in the liberalization of capital transactions were undertaken, partly reflecting the requirements for acceding to the OECD (Chapter XII). The trade liberalization process was also continued (Chapter IX). There was a further relaxation of the remaining import quotas (which, already in 1994, covered a relatively limited number of goods). The effective tariff rate (i.e., the ratio between tariff and fees revenues and the total import value) did increase in 1995, reflecting the above-mentioned import surcharge, but fell rapidly in the following two years as the import surcharge faded away and underlying tariffs and fees were reduced or removed in line with Hungary’s international commitments. Indeed, the effective tariff rate in 1998 is expected to be one-third of its 1994 level.

Fiscal reform focused on the reduction in the overall size of the public sector and on social security reform. While official data tend to overstate the magnitude of the phenomenon (Chapter V), primary revenues and expenditure ratios dropped by about 10–15 percentage points of GDP between 1994 and 1997 (albeit from high levels). The revenue drop reflected sizable cuts in enterprise taxes and trade tariffs (but households taxes were also lowered). It may also have been affected by increased tax erosion and evasion, although, in the same period, several measures were undertaken to strengthen tax administration. On the expenditure side, in March 1997, public employment stood at about 11 percent below its level of three years before. Outlays for subsidies, pensions, and investment also declined in relation to GDP. Important structural reforms—with less immediate effects on the deficit but crucial for the medium- to long-term evolution of public finances—were also undertaken in the social security area. The most important reform was the pension reform: enacted in two steps—in 1996 and 1997—it involved an increase in pension age, penalties for early retirement, a revision in the pension indexation mechanism, and the introduction of a fully funded component of the pension system, aimed at providing about 25–30 percent of pension revenues in the long run. Parliament also approved a resolution assigning to the government the mandate to reform the disability pension system, but a new law in this area is not expected until after the May 1998 elections.11 Measures were also introduced to rationalize the health system (reductions in pharmaceutical subsidies, and steps to reduce unused hospital capacity). However, the reform laws enacted in 1997, while providing a useful streamlining of existing legislation, fell short of what was needed to give the health system sounder and more efficient foundations (Chapter V).

Results of the Adjustment Program

External Accounts

During 1995–97, the performance of the Hungarian economy improved in almost all respects. The most impressive result of the program was the dramatic turnaround in the external accounts, which is illustrated by two key statistics: the external current account deficit fell from 9½ percent of GDP in 1994 to a projected 2¾ percent of GDP in 1997; and net external debt was lowered from more than 45 percent of GDP to less than 30 percent of GDP.12

This turnaround benefited from favorable external conditions (the growth rate of imports of Hungary’s partner countries was 7 percent in 1995–97, against 4½ percent during the first half of the 1990s), but should be credited primarily to the policies implemented by the authorities.

To clarify the link between economic policies and external developments, it is useful to break down the external current account into an interest component and a noninterest component. As shown in Figure 2.4, the improvement in the noninterest component is closely correlated with the improvement in the fiscal balance. Of course, the adjustment also required a corresponding adjustment in the real exchange rate, which was overvalued at the end of 1994. However, after the initial adjustment in 1995, the CPI-based real exchange rate gradually recovered and eventually exceeded its end-1994 value. On the other hand, the depreciation of the unit-labor-cost-based real exchange rate persisted (and was about 20 percent; Figure 2.2). The different behavior of the two real exchange rates reflects a number of factors, including increasing profit margin and faster productivity growth in the tradable vis-à-vis the nontradable sector, strengthened by the fact that adherence to the incomes policy guidelines proved to be somewhat stricter in the tradable sector, possibly reflecting the discipline imposed by the exchange rate peg. The improved competitiveness of Hungarian exports is evidenced by the steep increase in the dollar receipts for exports, which outstripped the sharp rise in imports (Figure 2.5).13

Figure 2.4.
Figure 2.4.

Fiscal Balance and External Current Account

(In percent of GDP)

Source: IMF staff calculations.1 Excluding net interest payments.2 Operational deficit (see Box 2.1).
Figure 2.5.
Figure 2.5.

External Current Account and Trade Balance

(Three-month moving average, millions of U.S. dollars)

Sources: National Bank of Hungary; and IMF staff estimates.1 The dotted line is the raw series; the solid line is seasonally adjusted.

The improvement in the current account also reflected a sizable decline in the burden of net interest payments, which dropped by 1¼ percentage points of GDP between 1995 and 1997.14 This drop was due to two factors: the decline in the stock of net external debt and the authorities’ active policy of refinancing high interest loans undertaken before 1995 with lower interest loans, as the risk premium on Hungarian government debt declined sharply.

The reduction in the current account deficit and the increased privatization receipts from abroad were, of course, key factors behind the drop in net external debt. Other factors were at play, some of which may be related more to the overall improvement of Hungary’s economic conditions than to specific policy measures undertaken by the authorities. Table 2.4 provides a breakdown of the factors behind the decline in net external debt. The table highlights the importance of FDI and, in particular, of privatization receipts. It also shows that a large component of the decline in net external debt (about $3.5 billion out of an overall decline of about $5¾ billion) cannot be identified precisely, as it corresponds to the errors and omissions in the balance of payments. This item is likely to include both current account transactions (which would give rise to a genuine decline in net external debt) and capital account transactions (which would largely be matched by unrecorded debt components). It is also possible that the errors and omissions item hides, as a result of a “confidence effect” in the forint, the surfacing of unrecorded financial assets held abroad illegally by residents. It is impossible to quantify the relative size of these components, particularly during a period in which there was a significant improvement in both recorded current account and capital account transactions.

Table 2.4.

Factors Underlying the Change in Net External Debt, 1995–97

(In billions of U.S. dollars)

article image

Including intercompany loans.

Including changes in cross exchange rates.

Output Growth, Aggregate Demand Composition, and Unemployment

A slowing down of economic activity was regarded as the inevitable by-product of the adjustment program. On average during the 1990s, and even after the initial output shock had been absorbed, output growth in Hungary remained well below its potential (Chapter VI). Nevertheless, the output acceleration of 1994 (Table 2.1) was clearly not sustainable, as it relied excessively on domestic demand and soon clashed against the external constraint. The program’s envisaged shift of resources toward the external sector was accompanied by a deceleration of growth in 1995–96. However, it is remarkable that, despite the magnitude of the required external adjustment, output growth remained positive in both 1995 and 1996, and accelerated briskly in 1997 (Table 2.1 and Figure 2.6). This result owes much to the structural changes that had already taken place in the first half of the 1990s. These changes allowed the export sector to react promptly to the macroeconomic stimuli coming from economic policies.

Figure 2.6.
Figure 2.6.

Industrial Production and Domestic and External Components of Aggregate Demand

Source: Hungarian Central Statistical Office.1 Also adjusted for working days.

The extent of the adjustment of aggregate demand from internal and external components is illustrated by Figure 2.6 (lower panel) and Figure 6.1 in Chapter VI. The contribution of external transactions to growth turned from minus 5 percentage points in the average of 1993–94 to plus 3½ percentage points in 1995–96. This exceeded the negative contribution that was coming from domestic consumption and investment. The decline in consumption was expected and was indeed an integral part of the adjustment process. The adjustment was painful: net real wages dropped by a cumulative 16½ percent in 1995–96 (Figure 2.7), and private consumption dropped by 10 percent. By 1997, however, both real wages and private consumption had begun to recover. The response of fixed investment was initially lukewarm: investment accelerated only in the second quarter of 1996. This delayed recovery is understandable. Even though enterprises’ internal financial sources improved rapidly in 1995, it stands to reason that enterprises initially took a wait-and-see attitude, given the uncertainty about the success of the stabilization, before launching new investment projects, particularly in domestically oriented activities (FDI, even excluding privatization, remained high throughout the period). Moreover, housing investment was constrained by the squeeze in households’ cash flow and cuts in housing subsidies.

Figure 2.7.
Figure 2.7.

Real Wages in Industry 1

(1993 = 100)

Source: National Bank of Hungary.1 Adjusted for working days and seasonally adjusted.

During the last three years, the unemployment rate has remained broadly constant at about 10–11 percent, and has failed to decline even during the most recent recovery (Figure 2.8). While in all countries employment lags behind output, a decisive increase in employment may require structural measures—such as a further reduction in labor taxes and a reform of the income support schemes that would discourage the active search for jobs (see Chapter X).

Figure 2.8.
Figure 2.8.

Unemployment Rate

(Seasonally adjusted)

Sources: Hungarian Central Statistical Office; and IMF staff calculations.


As argued above, the main target of the authorities’ 1995–97 adjustment program was external. The establishment of a crawling exchange rate regime aimed at avoiding a permanent acceleration of inflation, which indeed was initially expected to come down quite rapidly in 1996 and 1997. In the event, however, progress on the inflation front was more limited than initially envisaged (Table 2.5 and Figure 2.9): the 12-month inflation rate was 18.4 percent in December 1997, one of the lowest levels in the 1990s, but still not too far from the range in which inflation had remained in 1993–94.

Table 2.5.

Consumer Price Index Inflation Targets and Outcomes

article image

Targets set in December 1995, as part of the IMF-supported adjustment program.

Revised in June 1996.

Figure 2.9.
Figure 2.9.

CPI Inflation

(Twelve-month growth rate)

Source: Hungarian Central Statistical Office.

The resiliency of inflation in 1996–97 reflects in part the authorities’ priorities (the need to avoid excessive risks to the external sector), and in part objective factors raising the cost of disinflation (see Chapter VII). These difficulties included sticky inflation expectations, after years of moderate but steady inflation—a difficulty that had been exacerbated by the peak in inflation in 1995; the still limited credibility of the newly established monetary framework; and continued pressure on prices arising from the transition process (in particular, the needed changes in administered prices). These difficulties are apparent in 1996, a key period, as price developments that year also explain the authorities’ decision to revise the inflation target upward for 1997.

The authorities’ target for 1996 was ambitious: inflation was expected to fall by more than 10 percentage points with respect to 1995. In the event, inflation did drop fairly rapidly, but was about 5½ percentage points above target. This overshooting was due primarily to two factors: first, the limited credibility of the anti-inflationary program and the related failure of the incomes policy guidelines to keep gross wages in line with the target, which was more evident in the private nonmanufacturing sector; and, second, further large increases in administered prices, whose extent had initially been underestimated.15 By the spring of 1996, it had become apparent that inflation would exceed the target and the authorities were faced with the difficult choice of revising the target or tightening policies (including exchange rate policy). As the latter option would have involved significant risks for output and external developments, the inflation target was revised for both 1996 and 1997. A key factor in this decision was the expected difficulty of reaching an incomes policy agreement with trade unions for 1997 based on low nominal wage increases after the overshooting of inflation in both 1995 and 1996.16 The decision resulted in a limited decline in inflation through 1997. However, it had a positive effect: with inflation developments in 1997 broadly in line with the revised target, the authorities’ credibility was enhanced and this strengthened the prospects for a faster disinflation in 1998 and in the medium term (Chapter XIII).


  • Boote, Anthony R., and Janos Somogyi, 1991, Economic Reform in Hungary Since 1968, IMF Occasional Paper No. 83 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Kornai, János, 1997, “The Political Economy of the Hungarian Stabilization and Austerity Program,in Macro-economic Stabilization in Transition Economies, by Mario I. Blejer and Marko Škreb (Cambridge, Massachusetts: Cambridge University Press), pp. 172203.

    • Search Google Scholar
    • Export Citation
  • Krueger, Thomas, and Mark S. Lutz, 1995, “Developments and Challenges in Hungary” in Road Maps of the TransitionThe Baltics, The Czech Republic, Hungary, and Russia, by Biswajit Banerjee and others, IMF Occasional Paper No. 127 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • National Bank of Hungary, 1994, “Consolidation of the Hungarian Banking System,” Monthly Report (October).

  • Neményi, Judith, 1996, “Capital Inflow, Macroeconomic Equilibrium, the Public Debt and the Profit and Loss of the National Bank of Hungary,Acta Oeconomica, Vol. 48 (3–4), pp. 24170.

    • Search Google Scholar
    • Export Citation

Hungary’s economic developments through 1994 have also been discussed, from an international perspective, in Boote and Somogyi (1991) and Krueger and Lutz (1995).


Note that, in the definition reported in Table 2.1, the adjustment in the primary balance between 1995 and 1997 was larger. The figure reported in the text refers to the definition used in Box 2.1, which is more significant from a macroeconomic standpoint. Also note that the fiscal tightening between 1995 and 1996, and the corresponding relaxation between 1996 and 1997, was larger than intended, as the 1996 deficit turned out to be lower than budgeted by about½percent of GDP. Moreover, because of the lengthening in the maturity of government securities in 1996, the deficit on an accrual basis in 1996 was larger than the deficit on a cash basis. Finally of intense sterilization operations (see below). The inflation-related component of these interest payments does not appear in the definition of operational deficit.


See Chapter V for a detailed discussion of the changes in the level and structure of public expenditure and revenues during 1995–97.


The rate of crawl was 1.3 percent a month in the second half of 1995; 1.2 percent a month during January 1996-March 1997; and 1.1 percent a month from April 1997 until mid-August 1997, when it was lowered to 1 percent. Through December 1996, the forint was pegged to a basket of the European currency unit (ECU) (70 percent) and the U.S. dollar (30 percent). As of January 1997, the ECU was replaced by the deutsche mark.


Other factors were also important. In Hungary, the role of trade unions in the emerging private sector branches (particularly in multinational corporations) has been modest (Chapter X). In these branches, wage increases are decided at the beginning of the year, often without consultation with the trade unions. Wages are not revised until the following year. This facilitated the erosion of real wages in the second half of 1995.


The IRC is a tripartite institution comprising representatives of trade unions, the government, and some employers’ associations. It issues recommendations for wage increases in the private sector, which are not binding.


Moreover, for most of 1995–97, the forint remained close to the more appreciated margin of the band, so that the risk of a depreciation within the band may have appeared limited.


A large share of foreign exchange market intervention reflected reverse currency substitution (in particular, domestic banks converting into forint their large stock of foreign exchange reserves deposited at the central bank).


Refinancing the whole stock of old debt through new borrowing would have been problematic (keeping gross borrowing low facilitated the decline in the spread). In any case, the spread on new issues declined only gradually from high levels in early 1995.


As the average interest rate paid on external debt by Hungary at the end of 1994 was fairly large, the cumulative privatization of 1995–97 resulted in annual savings of about ¾ percent of GDP for the external current account and the public sector budget.


The disability pension system is poorly targeted. As a result, people of working age receiving disability pensions represented almost 9 percent of the workforce in 1997.


The reduction is even larger excluding intracompany loans, which increased rapidly in 1995–97.


The figure, as well as the data in Table 2.1, may overstate the actual increase in external trade. Export receipts in 1994 may have been artificially depressed by hidden capital outflows. Moreover, anecdotal evidence indicates that imports for reexport increased rapidly during 1995–96. The further sharp increases in imports and exports recorded in early 1997 may also be due to changes in the recording of some imports for reexport from a net basis to a gross basis.


The comparison with 1995 is more significant than that with 1994 because the increase in interest rates on Hungarian debt observed in 1994 was reflected in interest payments only in 1995. Moreover, to a large extent, the interest savings related to privatization receipts were not felt until 1996, as privatization receipts in 1995 were concentrated in December.


Initially, it had been expected that the increase in administered prices would raise the CPI by about 2 percentage points in January 1996. The actual contribution was twice as large. Thus, taking into account the reaction of other prices to changes in administered prices, this unexpected increase in administered prices explains about one-half of the 1996 price overshooting.


There was also a problem of timing. The 1997 inflation target was announced in June 1996, when the 12-month inflation rate had just started declining. With the benefit of hindsight, that is, taking into account the drop in inflation in the second half of 1996, the announcement appears to have been premature.

Economic Policies for Sustainable Growth