Chapter

II Macroeconomic Imbalances and Financial Strains

Author(s):
International Monetary Fund. Research Dept.
Published Date:
January 1992
Share
  • ShareShare
Show Summary Details

The current macroeconomic difficulties of the industrial countries can, to a large extent, be attributed to the failure to achieve key medium-term policy objectives during the past decade. This failure has both contributed to the imbalances that led to recent turbulence in the exchange markets and considerably narrowed the scope for policymakers to respond effectively to the sluggishness of growth. On the positive side, a major accomplishment has been the widespread reduction of inflation to levels not experienced since the 1960s. But large fiscal imbalances continue to overburden monetary policy and are adversely affecting business and consumer confidence and financial markets. In many countries these problems are compounded by disequilibria in private sector balance sheets. This chapter examines three important areas where macroeconomic imbalances are giving rise to serious strains in financial markets and are contributing to the weakness of activity.

Factors Underlying Recent Exchange Market Turbulence in Europe

As discussed in Chapter I, the recent crisis in European foreign exchange markets was triggered by uncertainty about the outcome of the September 20, 1992 referendum in France on the Maastricht treaty on economic and monetary union (EMU) and by a perception that rejection of the treaty by French voters—following rejection by Danish voters on June 2—would jeopardize the prospects for monetary cooperation and exchange rate stability in Europe. The French referendum provided a specific focus for the intensification of speculative pressures, but there were more fundamental economic forces at work, stemming from divergences in economic policies and performances across Europe.

In particular, notwithstanding considerable progress toward economic convergence, several countries still had relatively large inflation differentials vis-à-vis the best inflation performers in the ERM, and in some cases the resulting cumulative losses of competitiveness suggested that a realignment of exchange rates would be necessary at some point. In addition, there was a growing conflict between monetary policy objectives in Germany and those in many other European countries because of cyclical divergences. These intra-European tensions were aggravated by the weakness of the U.S. economy, the marked decline in U.S. interest rates, and the associated shift of funds into assets denominated in the stronger European currencies, especially the deutsche mark.

Inadequate Convergence in the EMS

A high degree of economic convergence has been recognized as a condition for exchange rate stability in the EC since the breakdown of the Bretton Woods system of fixed rates in the early 1970s. Early attempts to peg exchange rates ran into difficulties because the participating economies diverged significantly in terms of growth and inflation and because oil price disturbances and different interpretations of the causes of the stagflation of the 1970s provided a poor basis for effective policy coordination. From the early 1980s, however, the new consensus that emerged on the need to pursue medium-term policy objectives, together with the discipline imposed by the ERM, contributed to greater economic convergence. This led to a reduction in the frequency of currency realignments,7 and all EC currencies with the exception of the Greek drachma eventually joined the ERM.8 These achievements encouraged the Community to put forward ambitious plans for a gradual transition to EMU.

Convergence, however, has been far from complete. To date, using as a benchmark the criteria set out in the Maastricht treaty,9 only three EC member states—France, Denmark, and Luxembourg—have achieved a satisfactory degree of convergence overall, although Germany, Belgium, the Netherlands, and Ireland have satisfied the inflation target for several years (Table 7). An important implication of the lack of convergence of inflation rates has been a progressive appreciation of some EMS currencies in real terms and an associated erosion of competitiveness.

Table 7.European Countries: Convergence Indicators for 1991 and 1992(In percent)
1991 GDP WeightsConsumer

Price

Inflation
General

Government

Balance/GDP
Gross

Government

Debt/GDP1
Long-Term

Interest

Rates
In ECIn world19911992199119921991199219911992
EC countries
France19.75.13.12.8–2.1–2.947.147.79.28.7
Germany25.56.54.85.0–3.2–3.241.742.58.57.8
Italy17.74.56.35.5–10.2–10.4103.5108.513.013.5
United Kingdom216.84.35.93.7–2.7–6.334.435.99.99.0
Largest four countries379.520.44.94.3–4.4–5.455.256.910.09.6
Belgium3.20.83.22.4–7.6–6.8134.4133.49.38.8
Denmark42.20.62.42.1–2.3–2.366.771.39.68.9
Greece51.10.319.516.0–17.4–13.8115.5114.823.321.5
Ireland0.70.23.23.3–2.8–1.898.0100.09.210.0
Luxembourg0.13.12.81.51.06.25.88.27.8
Netherlands4.61.23.93.3–3.9–4.079.680.08.78.0
Portugal0.90.211.49.2–6.7–5.765.369.518.516.5
Spain7.72.05.95.9–4.9–5.144.746.012.612.6
Smallest eight countries320.55.35.55.0–5.5–5.275.577.411.611.1
All EC3100.025.75.14.5–4.6–5.359.461.310.39.9
Maastricht convergence criteria64.44.0-3.0-3.060.060.011.010.7
Non-EC countries
Austria0.73.33.8–2.2–1.956.555.88.67.9
Finland0.64.22.7–6.0–8.821.733.012.212.8
Norway0.53.22.3–1.1–4.343.350.09.99.6
Sweden1.09.32.2–1.5–6.444.850.010.710.3
Switzerland1.05.94.0–1.7–1.732.134.36.46.6
Five non-EC countries3.85.73.1–2.4–4.438.643.79.39.1
Sources: National sources and staff projections.

Debt data are from national sources. They relate to the general government but may not be consistent with the definition agreed at Maastricht.

Debt on fiscal year basis.

Average weighted by 1991 GDP shares.

The debt-GDP ratio would be below 60 percent if adjusted in line with the definition agreed at Maastricht.

Long-term interest rate is twelve-month treasury bill rate.

Unweighted averages.

Sources: National sources and staff projections.

Debt data are from national sources. They relate to the general government but may not be consistent with the definition agreed at Maastricht.

Debt on fiscal year basis.

Average weighted by 1991 GDP shares.

The debt-GDP ratio would be below 60 percent if adjusted in line with the definition agreed at Maastricht.

Long-term interest rate is twelve-month treasury bill rate.

Unweighted averages.

Although the United Kingdom now meets the Maastricht inflation criterion, this has only recently been the case. From the time the pound sterling joined the ERM in 1990—at an exchange rate that was thought to strike a reasonable balance between the goal of reducing inflation and the need to maintain competitiveness—up to the recent crisis, sterling had appreciated by 10 percent in real terms against the deutsche mark. Compared with its low of 1987, the pound’s real appreciation against the mark had been close to 20 percent (Chart 11).10 Similarly, the lira was last realigned in 1987,11 but from that time to mid-1992 the relatively large increase in Italian labor costs had resulted in a real appreciation of some 18 percent against the deutsche mark. The peseta and the escudo had also appreciated in real terms vis-à-vis the deutsche mark in recent years.

Chart 11.Nominal and Real Bilateral Exchange Rates vis-à-vis Germany1

(January 1987 = 100)
(January 1987 = 100)

1Changes in real bilateral exchange rates vis-à-vis the deutsche mark can be inferred as the product of the change in the nominal exchange rate and the cumulative inflation differential (inflation measured either by unit labor costs or by consumer prices). This chart includes estimates through November 1992.

2For Portugal and Greece the cumulative inflation differential is measured by consumer prices through October 1992.

Real effective exchange rates, which take account of patterns of trade, show a similar development (Chart 12). Moreover, based on exchange rates prevailing before the September crisis and on IMF and private sector projections for labor costs and inflation, further real appreciation would have been implied for the lira, the peseta, and the escudo during the period ahead. Similarly, there would have been only a slow correction of the cumulated loss of competitiveness of the pound on the assumption of a further decline in wage inflation in the United Kingdom over the medium term.12 Thus, the pressures for an exchange rate realignment, which had been building for some years as a result of divergent price and cost developments, could reasonably have been expected to persist—or in some cases to intensify—in the future. The deterioration of cost competitiveness was probably smaller for tradable goods industries subject to external market discipline. Nevertheless, there was a clear signal to financial markets of the growing overvaluation of several currencies.

Chart 12.Real Effective Exchange Rates1

(1980 = 100)
(1980 = 100)

1Relative unit labor costs; data extend through November 1992. Data for Portugal and Greece are based on relative consumer prices and extend through September 1992.

It seems likely that inadequate budgetary convergence reinforced perceptions that some currencies were vulnerable, since experience points to the risk that large fiscal imbalances sooner or later will add to inflationary pressures. Moreover, the introduction of restrictive measures to eliminate excessive fiscal deficits may contribute to downward pressure on the exchange rate during the initial stages of the adjustment process. Apart from Greece, where the large fiscal deficit and high inflation rate have precluded membership in the ERM, the lack of fiscal convergence is serious in Portugal, Spain, and especially Italy, which are expected to have general government deficits amounting to 5¼, 5, and 10½ percent of GDP in 1992, respectively, even though the margin of slack, and hence the impact of automatic stabilizers, is small.13 In Italy, the unsustainable nature of the budgetary situation was, and remains, particularly evident. Italy has had deficits close to, or greater than, 10 percent of GDP for some years, and the public debt-GDP ratio is expected to approach 110 percent in early 1993. In the absence of stringent fiscal consolidation, moreover, deficits are projected to rise substantially.

Evidence that market participants did not regard the exchange rates that prevailed in mid-1992 as sustainable over the medium term can be found in the evolution of interest rate differentials vis-à-vis Germany. These had narrowed substantially over the past decade for those countries that achieved a high degree of convergence, but relatively large differentials persisted for Italy, Portugal, and Spain even before the recent turbulence (Chart 13). Although short-term interest differentials have generally narrowed in recent years, long-term differentials remained fairly large for these countries. This suggests that markets saw a significant probability of substantial depreciations of the lira, the peseta, and the escudo over the medium term, even though the perceived risk of realignment in the near term was small; such expectations were consistent with the relatively strong positions of the escudo and the peseta in the exchange rate band during the first half of 1992.14 After the Danish referendum in June, however, short-term differentials for these three countries began to widen, and their currencies weakened significantly in the exchange rate band over the course of the summer (Chart 13).

Chart 13.Interest Rate Differentials vis-à-vis Germany

1 For Portugal, the long-term interest rate is the before-tax yield on indexed government bonds.

As exchange market tensions subsided somewhat after the September crisis, short-term interest differentials generally declined to the levels prevailing before the crisis. Only in the United Kingdom, however, have short-term rates been reduced significantly below those in Germany—a development that has contributed to the relatively large depreciation of the pound sterling. In contrast, long-term rates remain higher in the United Kingdom than in Germany, although the differential has narrowed slightly. The persistence of relatively large long-term interest differentials for Italy, Spain, and Portugal even after the depreciation of their currencies suggests continued uncertainty about the medium-term inflation outlook as a result of large fiscal imbalances, particularly in Italy.

Cyclical Divergences and German Unification

Cyclical divergences have developed since 1990 between Germany, where rising fiscal deficits in the wake of unification led to overheating, and other European countries, where output has fallen or growth has been insufficient to prevent further increases in unemployment. In these circumstances, the need for Germany to pursue a tight monetary policy posed a difficult dilemma for many other European countries, since the high level of interest rates necessary to defend the commitment to stable exchange rates increasingly appeared to be inappropriate in light of domestic macroeconomic conditions.15 This dilemma between external and domestic objectives affected not only the ERM countries but also those European countries that directly or indirectly had been shadowing the deutsche mark through unilateral pegs, including Finland, Norway, and Sweden.

The conflict with domestic requirements was particularly sharp in the United Kingdom and Finland, which have both been in protracted recession since 1990. A similar, although less acute, conflict existed in countries where inflation had been reduced to levels below those in Germany and where activity was sluggish. However, for these countries—including Belgium, Denmark, France, and the Netherlands—gradual improvements in competitiveness vis-à-vis Germany since 1989 helped to strengthen their net export performance, partially offsetting the adverse impact of high interest rates on domestic demand. For the nonconverging countries, the high level of interest rates necessary to defend their parities was consistent with domestic requirements for further disinflation, but it complicated the fiscal adjustment process by raising public debt-servicing costs. Previous issues of the World Economic Outlook have emphasized the need for early measures to reduce the fiscal deficit in Germany in order to alleviate these tensions and the pressures for a realignment.

Outside of Germany, inadequate inflation convergence and cyclical divergences were probably the decisive factors behind the September crisis, especially as it affected the lira, the pound sterling, and the peseta.16 Market assessments of the credibility of exchange rate commitments, however, appear to have been influenced by other considerations as well. Among these are countries’ historical inclination to resort to exchange rate depreciation in times of crisis and the constraints (both in terms of policy instruments and in terms of central bank independence) that might prevent central banks from raising interest rates sufficiently. The actions of trading partners undoubtedly also played a role, as illustrated by pressures against the Swedish krona after the floating of the Finnish markka, against the Irish pound after the depreciation of sterling, and against the Norwegian and Danish kroner after the floating of the Swedish krona.

But the intensity of the speculative pressures that developed against the French franc, in particular, is difficult to explain in view of the high degree of convergence of the French economy since the mid-1980s and its relatively strong external competitive position in recent years. There may have been a perception that the franc was vulnerable because the high unemployment rate in France would limit the scope for significant interest rate increases, as had been the case in the United Kingdom. Perhaps more important, after the exit of the lira and the pound from the ERM, markets appear to have felt that the EMS might break down completely, in which case speculation could have forced the deutsche mark to appreciate against most other European currencies, including the French franc.

In sum, the September crisis appears to have reflected an unusual combination of divergent economic forces and important political events that raised doubts about countries’ commitment to control inflation and to reduce excessive budget deficits and about the prospects for monetary cooperation in Europe. It is unclear, even in retrospect, if the political events would have triggered a crisis had the underlying economic forces been less divergent. However, an exchange rate realignment at an appropriate moment to adjust for divergent fundamentals—with adequate support from strengthened convergence policies together with an intensification of efforts to contain the budget deficit in Germany—could undoubtedly have reduced the strains on the system.

The smaller crisis in November, which led to another realignment of the peseta and to a realignment of the escudo, appears to have reflected market perceptions that these two currencies were vulnerable because of the large inflation differentials in the past and the relatively small realignment of the peseta in September. The decision to float the Swedish krona, which preceded the latest ERM realignment, was related to a failure to reach a new agreement to reduce the excessive fiscal deficit. In addition, notwithstanding the successful defense of the krona in September, the country’s deepening economic and financial crisis left the currency vulnerable to renewed pressures. The decision to float the Norwegian krone in early December reflected mounting pressures in the wake of the depreciation of the Swedish krona.

Several important lessons can be drawn from these episodes of exchange market turmoil. The suspension from the ERM of the lira and sterling has left the reputation of the EMS as a zone of monetary stability badly shaken and the system itself potentially vulnerable in the future—not least in view of the fixed timetable for the EMU process and the continued uncertainty about the ratification of the Maastricht treaty. To reduce the risk of a repetition of the crisis, progress toward convergence will need to be reinforced considerably. In this context, it is important to stress that the Maastricht convergence criteria are minimum requirements and that more ambitious targets are appropriate in many cases. Moreover, although participation in the ERM can be an important element in countries’ convergence strategies, recent experience underscores that a fixed exchange rate is likely to be challenged sooner or later unless it is backed by forceful and successful adjustment efforts. To the extent that convergence is achieved only gradually, it will be necessary to undertake small realignments from time to time to avoid the buildup of pressures for larger, more disruptive adjustments.

The crises have confirmed that fully sterilized intervention has only limited effectiveness in influencing exchange rates. Even where parities were successfully defended, as in the case of France and Sweden (in September), substantial adjustments of interest rates were required to support intervention. The episodes also suggest that defense of fixed exchange rates is most successful when it is closely coordinated by the central banks on opposite sides of exchange market pressures. In addition, the pressures against some currencies such as the French franc and the Danish krone, for which the fundamentals did not justify expectations of a realignment, have underlined both the desirability of ensuring rapid progress toward EMU and the crucial importance of the credibility of countries’ commitment to converge. The events since September have also underscored the need to strengthen the functioning of the EMS during the transition period.

Finally, the recent episodes have illustrated the extent to which the private sector is willing to take positions when a currency is felt to be overvalued. Indeed, both the size of the market and the number of private operators able to inject very large sums into currency markets are clearly much greater now than in the past, and both have probably been underestimated.

Medium-Term Consequences of the Budget Deficit in the United States

In the United States, the federal government deficit on a unified budget basis averaged 3½ percent of GDP during the 1980s and the early 1990s. As a result, the outstanding federal debt rose from 23 percent of GDP in 1980 to 49 percent of GDP in 1992—a development that has placed an increasing burden on future generations, who will have to service the debt either through higher taxes or reduced public services. These sustained deficits, by reducing the amount of domestic saving available to the private sector, have also contributed to high real long-term interest rates and to large current account deficits and growing foreign indebtedness. Consequently, during the 1990–91 recession and the unusually weak recovery that followed, not only has monetary policy been the only available instrument to support economic activity, but the effectiveness of lower short-term interest rates has also been brought into question because of the high level of long-term interest rates and the high degree of uncertainty about future economic policies in the United States. Thus, although the recent enlargement of the federal deficit to almost 5 percent of GDP reflects cyclical and other temporary factors, a credible program to reduce the large structural component of this deficit remains—as emphasized in past issues of the World Economic Outlook and in the IMF’s Article IV consultations with the U.S. authorities—the most important macroeconomic policy problem confronting the United States.

Both theory and experience demonstrate that the best time to approach the difficult task of fiscal consolidation is when economic activity is otherwise expected to be rising at least at a moderate pace. To illustrate this important principle for the United States, it is useful to consider—through an alternative scenario—what would have happened if a gradual program of fiscal consolidation had been rigorously pursued during the long economic expansion of the 1980s, and if, as a consequence, the budget-tightening measures adopted in late 1990 had not been necessary. Since such a program of fiscal consolidation was widely recommended at the time, and indeed was the explicit objective of the Gramm-Rudman-Hollings Act, the simulation exercise should be regarded as something more than second guessing after the fact.

Under the simulated program of gradual fiscal consolidation, the federal budget would have shown a surplus (on a unified basis) in the late 1980s, before falling with the onset of recession in 1990 (Chart 14 and Table 8). The stock of publicly held federal debt would have peaked at less than 35 percent of GDP in 1986 and then headed downward to below 25 percent of GDP by the early 1990s. According to the simulation, growth would have been slowed modestly during 1983–86, when the economy was otherwise expanding fairly vigorously (Chart 15). Note that the downturn of economic activity during the 1990–91 recession is moderated under the simulation. This favorable result does not reflect discretionary fiscal action to combat the recession, but rather the absence of discretionary actions that were actually taken in the Budget Enforcement Act of 1990 because of the pressing need to reduce a deficit that had previously not been adequately controlled. If the simulated consolidation program had been followed, however, the favorable fiscal situation achieved by 1990 would clearly have allowed some room for temporary discretionary fiscal actions to combat the recession. In addition to the direct effect on activity, business and consumer confidence would probably have benefited from the knowledge that such temporary actions could be taken to support activity without impairing longer-term fiscal probity.

Chart 14.United States: Federal Government Budget Balance, Primary Balance, and Debt

(In percent of GDP)

1 Unified budget basis.

2 Held by the private sector.

Table 8.United States: Counterfactual Simulation of Fiscal Consolidation in the 1980s(Percent deviation from baseline unless otherwise noted)
19831985198719891992
United States
Real GDP–0.1–0.20.20.6
Real GNP–0.1–0.20.10.41.0
Real investment0.72.63.93.5
Absorption deflator0.60.2–0.9–2.0–2.3
Long-term interest rate1–0.1–0.8–1.6–1.7–0.9
Real long-term interest rate10.4–0.2–1.2–1.6–1.1
Real effective exchange rate–4.4–5.4–6.6–6.9–5.1
Current account balance20.713.225.249.158.0
Net foreign assets30.61.42.85.3
Federal government balance30.51.63.14.84.8
Federal government debt3–0.6–2.8–7.4–14.4–27.1
Other industrial countries
Real GDP0.10.10.20.40.4
Real investment1.12.22.32.31.6
Absorption deflator–0.3–0.5–0.9–1.1–1.0
Real long-term interest rate1–0.2–0.4–0.5–0.5–0.2
Real effective exchange rate1.21.62.02.11.6
Current account balance23.7–12.1–25.2–40.4–46.7
Note: Compared with the historical baseline, and allowing for effects induced by changes in activity and inflation, the primary deficit of the federal government was reduced by ½ of 1 percent of GDP in 1983 and by a further ½ of 1 percent of GDP compared with the historical baseline in each of the following six years, giving a cumulative reduction equivalent to 3½ percent of GDP by 1989. The reduction in the primary deficit was held constant at 3½ percent of GDP in 1990 and was then reduced to 3 percent of GDP in 1991–92, reflecting the estimated approximate impact of the 1990 Budget Enforcement Act. Half of the reduction in the fiscal deficit is assumed to have come from lower expenditures and half from increased tax revenues. The IMF’s multicountry macroeconomic model, MULTIMOD, does not distinguish between the federal and total government; the changes in the government balance and in government debt are attributable to the federal government on the basis of the assumption that the state and local governments, most of which have constitutional limits on debt accumulation, would have taken actions to maintain deficits unchanged from the historical baseline. Monetary policies in the United States and other countries are assumed to have been neutral in the sense of allowing interest rates to adjust to simulated changes in actual and expected economic developments. Fiscal policies in other countries are assumed to have been what they actually were.

In percentage points.

In billions of U.S. dollars.

In percent of GDP.

Note: Compared with the historical baseline, and allowing for effects induced by changes in activity and inflation, the primary deficit of the federal government was reduced by ½ of 1 percent of GDP in 1983 and by a further ½ of 1 percent of GDP compared with the historical baseline in each of the following six years, giving a cumulative reduction equivalent to 3½ percent of GDP by 1989. The reduction in the primary deficit was held constant at 3½ percent of GDP in 1990 and was then reduced to 3 percent of GDP in 1991–92, reflecting the estimated approximate impact of the 1990 Budget Enforcement Act. Half of the reduction in the fiscal deficit is assumed to have come from lower expenditures and half from increased tax revenues. The IMF’s multicountry macroeconomic model, MULTIMOD, does not distinguish between the federal and total government; the changes in the government balance and in government debt are attributable to the federal government on the basis of the assumption that the state and local governments, most of which have constitutional limits on debt accumulation, would have taken actions to maintain deficits unchanged from the historical baseline. Monetary policies in the United States and other countries are assumed to have been neutral in the sense of allowing interest rates to adjust to simulated changes in actual and expected economic developments. Fiscal policies in other countries are assumed to have been what they actually were.

In percentage points.

In billions of U.S. dollars.

In percent of GDP.

Chart 15.United States: GDP

(In billions of 1987 U.S. dollars)

The implications of this simulation exercise for current U.S. policy are clear. Unquestionably, fiscal consolidation is required over the medium term to contain the growing stock of government debt and to enhance prospects for growth by reducing longer-term interest rates. During 1993 and the next two or three years, it is reasonable to anticipate that the United States will enjoy at least a moderately paced recovery as balance sheet problems are corrected and as lower short-term interest rates take full effect. Some time will be required to enact many of the measures necessary to raise revenue and, especially, to control spending in several critical areas, including health care and other “entitlements.” Thus, it is important to begin now to consider these measures, with a view to their enactment during 1993 and their gradual implementation during 1993 and succeeding years. With this timetable, the main effort of fiscal consolidation—which is absolutely required for the medium and longer term—should take place when the economy is best positioned to absorb it, while still sustaining an acceptable pace of output growth. Otherwise, as in late 1990, fiscal consolidation is likely to be necessary when growth is slowing or a new recession is already in progress.

The above timetable for fiscal adjustment has potentially important and favorable implications for the likely course of U.S. monetary policy. As economic recovery gathers momentum and margins of excess capacity are reduced, inflationary pressures naturally tend to rise. In the late 1980s (beginning in 1987 and extending through the summer of 1990), this tendency was counteracted by a considerable tightening of U.S. monetary policy, as reflected both in the rise of short-term interest rates and in the slowing of growth of monetary aggregates. It would have been much better if fiscal policy had borne at least part of this burden through the pursuit of necessary fiscal consolidation while the economy was expanding and ultimately overheating. This is again a relevant concern as the U.S. economy appears poised to achieve somewhat more rapid recovery, with levels of excess capacity far lower than they were at the end of the 1981–82 recession. Even with an appropriate program of fiscal consolidation, it is likely that U.S. short-term interest rates will rise during the course of the current recovery. With a monetary policy properly attuned to an appropriate fiscal policy, however, it is possible that the inevitable flattening of the currently steep yield curve may occur in part through a reduction in longer-term interest rates. This would be a particularly favorable development from the perspective of longer-term growth.

The need for medium-term fiscal consolidation leaves open the issue of immediate measures to provide short-term fiscal stimulus to the U.S. economy during 1993, as well as the broader questions of structural changes in fiscal policy to enhance longer-term growth. As regards short-term fiscal stimulus, two conditions are essential for this to make any sense for 1993. First, to avoid adverse financial market reaction, which could easily counteract the stimulus measures, these measures would have to be linked to a credible program of medium-term fiscal consolidation. In view of the failure of past efforts to ensure fiscal consolidation, establishing such a credible linkage would clearly not be easy—and would be made even more difficult because the stimulus package would be put in place very early, before much of the medium-term consolidation program could be enacted. Second, because any short-term fiscal stimulus must later be reversed in order to achieve the desired medium-term path of fiscal consolidation, a stimulus package for 1993 would make sense only if there is good reason to believe that economic growth in 1993 is likely to be significantly slower than in subsequent years. Otherwise, the perception is likely to be that the government will always talk about fiscal consolidation for the medium term but will never find the conditions appropriate to convert such intentions into action.

As regards the broad subject of the structure of U.S. fiscal policy, much can be done to shift the orientation of both taxation and spending toward the support of stronger long-term growth, while simultaneously pursuing overall fiscal consolidation. Remembering that tax revenue will have to rise, on balance, if the deficit is to be reduced, it makes sense to reduce excessively heavy rates of taxation on productive investment while raising tax rates on current consumption. Special tax breaks for particular forms of investment, however, can generate significant revenue losses without necessarily raising the level or improving the distribution of investment. On the spending side, greater emphasis on public investment—including support for private investment in human and physical capital—could also aid longer-term growth. Merely classifying an expenditure as an “investment,” however, does not necessarily imply that the return justifies the cost. Also, it is important to avoid the fallacy that large-scale public investments will generate such economic benefits that the additional tax revenues will more than pay the cost of the investment. In the end, public investments must be paid for either with tax revenue or with diminished spending on current government services; and this must be done in the context of a general effort to reduce the federal deficit substantially.

Ongoing Process of Asset Price and Balance Sheet Adjustment

Although asset price declines and balance sheet adjustments are not unusual during economic downturns, they have been particularly important during the current business cycle. In several countries, sharp increases in real estate values or stock prices, fueled in part by a considerable expansion of private sector indebtedness during the 1980s, have been followed by unusually sharp downward adjustments. The consequent increase in private sector indebtedness relative to asset values and income levels during the past two years has been a restraining influence on private spending and has thereby contributed to the sluggishness of the current recovery. The balance sheet positions of many financial institutions have also weakened considerably, leading to more cautious lending practices. These real and financial adjustments are likely to continue to restrain the pace of the economic recovery in the industrial countries.

Asset Price Inflation and Its Reversal

Several structural factors, together with other economic and policy developments, help to explain the excessive expansion of private sector indebtedness and asset price inflation in the 1980s.17 Speculative forces in financial and real estate markets that drove asset prices to levels that bore little relationship to underlying fundamentals were an important factor. The process of financial deregulation and liberalization initiated by many industrial countries in the late 1970s was beneficial because it improved access to credit markets, led to more market-determined interest rates, and fostered greater allocative efficiency. But in the absence of adequate prudential safeguards, it also encouraged greater risk taking and speculation and ultimately created an institutional environment in which there was potential for dramatic, and in some cases unwarranted, increases in both the demand for and supply of credit. Two other structural factors also sharply increased the demand for assets financed by credit. First, demographic changes—most notably the entry into the housing market of the baby-boom generation—increased the demand for housing, automobiles, and other durable goods. Second, new or long-standing tax incentives for debt financing encouraged consumers and investors to finance expenditures with new varieties of debt instruments, many of them with adjustable interest rates.

Monetary policy played a key role in the asset price inflation of the 1980s by inadvertently accommodating the strong demand for credit in many industrial countries. The reshaping of the economic and financial environment described above led to dramatic changes in the scale and composition of financial activity. Asset prices and interest rates became more volatile, and other important economic indicators used in the conduct of monetary policy became less reliable.18 Moreover, since asset prices and their movements were not captured by consumer price and wholesale price indices, monetary authorities may have ignored important information about the short-term and medium-term effects of their policies. In the event, the stock market crash in 1987 was an early warning sign of growing financial imbalances; but the liquidity provided during the crash to support financial markets was not fully removed and contributed to subsequent asset price inflation. Once monetary policies were tightened to eliminate the overheating in 1989–90, there was a sharp correction of both real estate and stock market prices.

The expansion of private sector indebtedness during the 1980s exceeded the expansion in economic activity by a wide margin (Chart 16). Many countries experienced a boom in asset markets that was associated with an unprecedented buildup of debt, sharp increases in asset prices, and related increases in household wealth. The appreciation of real estate values provided homeowners with unrealized capital gains, and in some countries the increases in stock prices after 1987 temporarily improved the capital value of many businesses. These temporary unrealized increases in wealth led to a reduction in both household and business saving rates and probably encouraged speculation on the basis of expectations of further asset appreciation.

Chart 16.Selected Countries: Total Private Nonfinancial Sector Debt

(In percent of GDP, end of period)

Sources: For the United States, Data Resources, Inc. data base; for the United Kingdom, Central Statistical Office, Financial Statistics; and for Japan, Economic Planning Agency, National Income Accounts.

1Total financial liabilities of the private nonfinancial sector less trade credit.

2Total credit market debt outstanding of the private nonfinancial sector. Data for 1992 are through the third quarter.

3Total financial liabilities of the personal and of the industrial and commercial sectors, less outstanding domestic trade credits and ordinary and preference shares.

The asset markets most affected were the residential and commercial real estate markets—most notably in Japan, the United Kingdom, Australia, and the Nordic countries, but also in the United States. Inflation-adjusted residential property prices increased at an annual average rate of 14 percent (19½ percent nominal) in the United Kingdom during 1986–89 and 6 percent (10 percent nominal) in the United States (Chart 17). In Japan, real land prices increased sharply, recording an annual average increase of 20 percent (22 percent nominal) during 1986–90. In some countries there were also sharp increases in stock prices—most notably in Japan but also in others, including the Nordic countries (Chart 17).

Chart 17.Selected Countries: Property Prices and Stock Market Indices

Sources: Property prices for the United States, Data Resources, Inc. data base; for the United Kingdom, Central Statistical Office, Financial Statistics; and for Japan, Real Estate Institute, Bulletin of Japan Land Prices. Stock prices for the United States, Data Resources, Inc. data base; for Japan, Nikkei Services; and for Finland, Norway, and Sweden, IMF, International Financial Statistics.

1Property prices deflated by consumer prices.

2Urban residential land prices in six largest cities.

3Index of prices on dwellings.

4Average price of a new house.

The expansion of indebtedness during the 1980s left the private sector in many countries unusually vulnerable to cyclical developments, high interest rates, and asset price declines. The general rise in interest rates during 1989–90, and the corresponding reductions in income growth and asset values, sharply diminished the private sector’s ability to service its debt. Debt service is most easily managed in an environment of rapidly expanding incomes, relatively low real interest rates, and improving consumer and business confidence. By contrast, during the current slow recovery, income gains have been weak, real interest rates in many countries in recession are still relatively high, and confidence levels are still depressed. As a result, households and businesses have reduced borrowing and increased saving to strengthen their financial positions.

Asset price deflation, in combination with the ongoing structural adjustments in financial sectors related to financial liberalization, has reduced the capital value of many financial institutions and has increased the fragility of the financial system in many countries.19 The most notable examples are the savings and loan crisis in the United States; bank failures in the Nordic countries, where several large banks have been taken over by the government; financial fragility in Japan, where asset price adjustments have adversely affected the financial position of banks and other financial institutions; and the poor earnings and weak profit outlook of financial institutions in the United Kingdom. As a result, financial institutions in all of these countries have adopted cautious lending practices, thereby contributing to a sharp slowing in the growth of credit.

Progress to Date in Selected Countries

Policymakers in some countries have attempted to ease the debt burdens of the private sector, either by lowering interest rates to reduce debt-service payments or by adopting policy measures to support the financial system. It remains difficult, however, to measure the equilibrium levels of asset prices and indebtedness—which may well have increased as a result of greater market access of households and enterprises following financial deregulation. It is therefore difficult to gauge the likely depth and duration of the financial adjustments, and most analysts, including those at the IMF, have underestimated the extent to which growth would be impeded.

The adjustment process in the United States is still under way, although it seems the nearest to completion among the major countries. Compared with Japan and the United Kingdom, the private sector adjustments required in the United States, in retrospect, appear to have been less severe because the runup in real estate prices was less dramatic (Chart 17). While U.S. consumer installment debt as a share of disposable income has been reduced to historical norms, mortgage debt as a share of income has not yet declined after surging to a historical high in 1991 (Chart 18). Although most of this rise seems to reflect the permanent effects of financial deregulation and tax incentives, some of it is likely to be reversed. A large part of the remaining adjustment of debt-income ratios in the United States may occur through income growth once the recovery gains momentum, as occurred during the late 1960s and early 1970s. Moreover, debt-service costs have declined significantly as lower short-term interest rates have gradually lowered adjustable rates on mortgages.

Chart 18.United States: Household Liabilities

(Percent of disposable income)

Source: Data Resources, Inc. data base.

The runup in asset prices in Japan was sharper than in many other industrial countries. Although recent asset price deflation has been dramatic, real estate prices could decline further in the absence of an early recovery of activity. Although the level of consumer indebtedness is low compared with most countries affected by asset price declines, the financial adjustment process has contributed to weaker-than-expected growth in consumption expenditure, and to the recent decline in business investment. Even though interest rates have been reduced significantly in Japan, the need for businesses to refinance equity-linked debt at higher costs has raised the cost of capital significantly.

Against this background, the Japanese government announced in mid-August 1992 a series of policy measures intended to strengthen confidence. The policy package consisted of a clarification of accounting procedures and the announcement of an intention to introduce measures to help financial institutions liquidate nonperforming real estate loans. The authorities subsequently announced on August 28 a stimulus package that included fiscal measures amounting to ¥10.7 trillion (2¼ percent of GDP, or $85 billion), consisting mainly of an increase in public works expenditures and an augmentation of loans from public financial institutions to be implemented during the remainder of the fiscal year. More recently, the Japan Bankers Association has proposed the establishment of the Real Estate Buy-up Corporation to encourage financial institutions to write off existing nonperforming assets and to establish a market for property formerly used as collateral.

Mortgage debt as a share of disposable income rose particularly sharply in the United Kingdom during the 1980s and has remained high in the early 1990s, in part reflecting low income growth during the recession (Chart 19). Meanwhile, real estate prices have declined significantly since 1989. In contrast to developments in the United States, however, house prices have continued to drop in recent months, and the decline to date has reduced the value of many homes below the value of existing mortgages. These developments suggest a risk of further balance sheet adjustment in the period ahead.

Chart 19.United Kingdom: Mortgage Liabilities

(Percent of disposable income)

Source: Central Statistical Office data base.

Although interest rates have been reduced in the United Kingdom, until recently the authorities have had limited flexibility to lessen debt-service burdens because of the exchange rate constraint. Household and business spending has remained depressed; where earnings and profits have improved, consumers and businesses have chosen to strengthen their financial positions by reducing existing debt burdens. Business conditions have shown little improvement, and consumer and business confidence remains weak. Since September, however, the further reduction of short-term interest rates allowed by sterling’s withdrawal from the ERM has helped to improve the outlook and should speed up the necessary adjustment, since most houses in the United Kingdom are financed with floating-rate mortgages.20

In view of the uncertainty surrounding the financial adjustments still under way, it is difficult to avoid the conclusion that the risks to the short-term projections for countries that have experienced asset price deflation are still predominantly on the downside. The adjustment process appears to have progressed further in the United States than elsewhere. In contrast, the possibility of further decline in asset prices cannot be excluded in Japan, where the private sector financial adjustment process is less advanced. Notwithstanding the recent decline in interest rates, there are also downside risks in the United Kingdom, especially in view of the high level of mortgage debt relative to household income. There is considerable uncertainty about the adjustment process in several smaller industrial countries, where policy flexibility is limited.

Before the September 1992 crisis, the most recent previous realignment took place in January 1987, which was the eleventh realignment since the inception of the EMS in 1979.

The peseta joined the ERM in June 1989, the pound sterling in October 1990, and the Portuguese escudo in April 1992; all three were subject to “wide” (6 percent) fluctuation bands. The other EC currencies (except the drachma) were part of the ERM from its inception in 1979, and all were subject to the “narrow” (2½ percent) band, with the exception of the Italian lira, which was subject to a wide band up to January 1990.

The relevant criteria are (1) consumer price inflation must not exceed, by more than 1½ percentage points, the average for those three member states with the lowest inflation rates; (2) interest rates on long-term government securities must not be more than 2 percentage points higher than those in the same three member states; (3) the currency must have been held within the narrow band of the ERM for two years without a realignment at the initiative of the member state in question; and (4) the financial position must be sustainable, which is defined as a general government deficit no greater than 3 percent of GDP and a public debt GDP ratio of no more than 60 percent. The treaty requires a substantial and continuous decline of fiscal deficits toward the reference value, and the debt-GDP ratio must be approaching the benchmark at a “satisfactory pace.” See “The Maastricht Agreement on Economic and Monetary Union,” Annex II in the May 1992 World Economic Outlook, pp. 52–55.

The real bilateral exchange rate calculations referred to here are based on relative unit labor costs in manufacturing for all countries mentioned except Greece and Portugal, for which relative consumer prices have been used.

A small adjustment of the lira’s central rate occurred in January 1990, when the lira entered into the narrow band of the ERM.

To the extent that the U.S. dollar can be expected to appreciate over the medium term (as implied by the large interest differential in favor of European currencies), the real effective exchange rate of the pound would tend to depreciate more than other European currencies because of the dollar’s relatively large weight in the United Kingdom’s basket of competing currencies.

The recent increase in the U.K. budget deficit may also have influenced market perceptions of the sustainability of sterling’s position in the ERM. The U.K. general government deficit is expected to exceed 6 percent of GDP in 1992, but since the deficit has widened, mainly because of the deep recession, it can be expected to fall as the economy recovers. Belgium is also expected to have a deficit of about 6 percent of GDP in 1992, but Belgium’s inflation rate has been quite low in recent years.

The relatively strong positions within the ERM of currencies of high-interest-rate countries demonstrated the phenomenon of “excess credibility” arising from foreign capital inflows motivated by large interest differentials and the absence of pressures for an early realignment. As sentiment suddenly changed, the reversal of these capital inflows contributed to the subsequent downward pressure on some currencies.

In this context the earlier IMF analysis of the macroeconomic implications of German unification has been broadly confirmed by subsequent events. See the October 1990 issue of the World Economic Outlook, pp. 44–50; and Paul R. Masson and Guy Meredith, “Domestic and International Macroeconomic Con-sequences of German Unification,” in German Unification: Economic Issues, Occasional Paper 75 (IMF, December 1990), pp. 93–114.

That the pressures against the escudo were quite limited may be attributed to the relatively small market for assets denominated in that currency. The introduction of capital control measures may also have played a role.

The process of asset price and balance sheet adjustments has been covered extensively in previous issues of the World Economic Outlook. See “Balance Sheet Constraints and the Sluggishness of the Current Recovery,” Annex I in the May 1992 issue, pp. 47–51, and “Asset Price Deflation, Balance Sheet Adjustment, and Financial Fragility,” Annex I in the October 1992 issue, pp. 57–68.

For a comprehensive survey of this issue, see Bank for International Settlements, Recent Innovations in International Banking (Basle, 1986).

A brief discussion of the policy implications of these developments can be found in the October 1992 World Economic Outlook, pp. 24–25. See also Annex I of that volume, pp. 57–68, for a detailed discussion of adjustments in the financial sectors of several industrial countries.

For floating-rate mortgages in the United Kingdom, changes in short-term interest rates are passed through to mortgage payments within one or two months in many cases. This suggests a relatively large and immediate impact of lower interest rates on consumer confidence and spending in the United Kingdom compared with most other industrial countries.

    Other Resources Citing This Publication