Abstract

In setting out the elements of a common strategy for the industrial countries to strengthen growth and reduce unemployment, the Madrid Declaration stressed the importance of maintaining price stability, strengthening fiscal consolidation efforts, and advancing structural reforms, including in labor markets and trade. A review of the progress toward these goals indicates some successes and some areas that need more attention. Success has been clearest in the areas of inflation and trade, with inflation remaining subdued in most countries and the Uruguay Round of the General Agreement on Tariffs and Trade (GATT) now being implemented. But industrial countries have had much more limited success in managing their budgets and reforming their labor markets, as shown by budget deficits and structural unemployment rates that remain too high in many countries.

In setting out the elements of a common strategy for the industrial countries to strengthen growth and reduce unemployment, the Madrid Declaration stressed the importance of maintaining price stability, strengthening fiscal consolidation efforts, and advancing structural reforms, including in labor markets and trade. A review of the progress toward these goals indicates some successes and some areas that need more attention. Success has been clearest in the areas of inflation and trade, with inflation remaining subdued in most countries and the Uruguay Round of the General Agreement on Tariffs and Trade (GATT) now being implemented. But industrial countries have had much more limited success in managing their budgets and reforming their labor markets, as shown by budget deficits and structural unemployment rates that remain too high in many countries.

The fiscal problems of the industrial countries, including the problems associated with population aging that lie ahead, were discussed at length in the May 1996 World Economic Outlook. For most countries, current and prospective budgetary imbalances are such that fiscal policy, for all practical purposes, must be geared toward fiscal consolidation. Discretionary measures to stimulate demand are not a feasible policy option in most countries. At best, the role that fiscal policy can play in directly supporting demand is limited to allowing automatic stabilizers to work, and even this may be feasible only in those countries where underlying deficits have been reduced sufficiently to avoid adverse market reactions to the cyclical widening of budget deficits. Progress in fiscal consolidation should, over time, expand the room for operation of the automatic stabilizers and permit, in some instances, discretionary fiscal actions to stimulate activity in times of weakness. But, for some time to come, fiscal policy in the industrial countries will need to remain firmly focused on the general need to reduce structural budget deficits.

Maintaining reasonable price stability is the primary goal of monetary policy, because it is the best contribution that monetary policy can make to supporting the highest sustainable growth paths for output and employment and to minimizing cyclical fluctuations around these growth paths. When monetary policy fails to contain inflationary pressure, thereby necessitating stronger tightening at a later date, the eventual effects on both the trend growth rate and the stability of output and employment are likely to be negative. These consequences were demonstrated once again in the experience of the industrial countries in the late 1980s and early 1990s.

The proper conduct of monetary policy, however, does not consist only of raising short-term interest rates to forestall a potential rise of inflationary pressures. Rather, monetary conditions need to follow a broadly cyclical pattern—tightening when inflationary risks rise as levels of activity approach potential, and easing when inflationary risks abate and slack emerges in the economy. Indeed, as economic information becomes available, central banks, to achieve their policy objectives for the medium and longer term, must frequently assess the need for adjustments in monetary conditions. This is a demanding task, particularly in view of the uncertainties about present and future economic conditions and the lags and ambiguities in the effects of changes in monetary conditions. Because of these difficulties, monetary policy cannot hope to smooth out all of the bumps and wiggles in the paths of the price level and output. Overambition in smoothing output, without adequate regard for inflationary risks, is counterproductive for the stability of both price levels and output. But overambition in reducing inflation too rapidly also has unwarranted adverse consequences for output and employment.

The experience of the past few years indicates that these challenges for monetary policy can be successfully met and that success is enhanced by the credible commitment of monetary policy to the primary goal of reasonable price stability. With its anti-inflation credibility reestablished by the tough actions of the early 1980s and reinforced by action to contain resurgent inflationary pressures in the late 1980s, the Federal Reserve had the flexibility to ease the federal funds rate down to 3 percent during the summer of 1992. This further easing of monetary conditions helped to revive the recovery and move the U.S. economy back to full employment. Tightening of monetary conditions, beginning in early 1994, assured continued low inflation. In Japan, short-term interest rates were lowered substantially during 1994 and 1995 as the economy faltered under the impact of a sharply appreciating yen. This easing of monetary conditions (including its impact on the yen) has clearly aided in the recovery now apparent in Japan. In Europe, too, central banks eased monetary conditions as inflationary pressures abated during the recessions of 1992–93. As it became clear that the European recovery of 1994 was stalling out during 1995 and early 1996, monetary conditions were eased significantly further, in the expectation that this would aid the renewal of recovery without generating inflationary risks.

Because of the importance of monetary policy in helping not only to establish reasonable price stability but also to stabilize the economy in a low-inflation environment, this chapter examines recent evidence on the monetary transmission mechanism. The main focus is on the question of whether the transmission mechanism may have changed in the 1990s compared with the 1980s. Deregulation and globalization of financial markets and rapidly rising stocks of financial wealth and debt have undoubtedly altered the way monetary policy affects an economy, and the extent of these changes is a matter of ongoing research and debate. Although some have argued that these changes may have served to enfeeble monetary policy, this does not appear to be the case. In fact, for a number of countries, monetary policy may be even more powerful in the 1990s than it was in the 1980s.

This chapter also discusses the relationship between banking sector health and macroeconomic performance. Banking sector risks are inevitably an important consideration for monetary authorities and are carefully monitored by them. What are the policy implications of recent banking sector problems?

Finally, the chapter argues that labor market rigidities have impeded the appropriate responses of the economy to policy and other changes in many countries. It discusses the problems caused by labor market rigidities, describes how they have impeded necessary reallocations of resources, and how they have served to frustrate policy efforts to reduce unemployment and encourage growth, especially in Europe. It also argues that good macroeconomic management is important for the reduction of structural unemployment.

Changes in Transmission of Monetary Policy

Traditionally monetary policy has been seen as influencing an economy in the short run, mainly through cost of capital effects—essentially the effects of changes in short-term interest rates on domestic saving and investment decisions—and through wealth effects and exchange rate effects. But other channels of influence, including through central bank credibility (or inflation expectations), and creditworthiness also may be important. Central banks have a direct influence on short-term interest rates, but can only indirectly influence long-term rates. Depending on the credibility of the authorities, a cut in short-term rates could lead to either an increase or a decrease in long-term rates. Credibility is therefore an important determinant of the effectiveness of policy actions.

Many changes have been taking place that affect the channels of monetary policy transmission. Financial structures have been changing owing to competitive pressures and financial deregulation, for example, and the increasing globalization of the world’s financial markets means that the effects of monetary policy may increasingly play out through capital flows and exchange rate movements. Meanwhile, the increased scale of financial market transactions has heightened the importance of the reactions of markets and expectations to changes in policy. For such reasons, it is worth looking at whether the changes in the channels of transmission are significant, and what they might mean for monetary policy. After a brief review of the theoretical basis for some of these mechanisms, several key changes that might affect the responses to monetary policy actions are examined. The overall impact of these kinds of changes are then assessed.

Channels of Influence of Monetary Policy

The channels through which monetary policy actions are transmitted into changes in real GDP and inflation have continued to be the subject of widespread theoretical and empirical analysis.7 The cost of capital channel works by changing the opportunity cost of borrowing funds to finance expenditure. A higher interest rate prompts a reduction in current expenditure on investment goods and probably on consumption (including housing and durables) and therefore lowers domestic demand. The associated reduction in activity increases excess capacity and unemployment, which in turn put downward pressure on inflation. Changes in interest rates also affect asset prices, including real estate prices and the value of corporate equities, partly by changing the rate at which expected future earnings from assets are discounted; this gives rise to wealth effects. A rise in interest rates lowers the value of an existing portfolio of assets, thus decreasing the expected lifetime resources of consumers and dampening consumption. In addition, a lowered value of equities, by reducing the market value of firms, makes it relatively cheaper to acquire capital through takeovers rather than through new investment. This is another channel implying that investment spending will be lowered.

An important transmission mechanism of monetary policy operates through the exchange rate. In an open economy with free capital flows and a flexible exchange rate, an increase in domestic interest rates that is not offset by a change in exchange rate expectations adverse to the domestic currency will amount to a relative increase in yields on domestic currency assets, and this will tend to cause the home currency to appreciate. This appreciation will reduce the international cost competitiveness of domestic producers, which will cause the trade balance to deteriorate, magnifying the depressive effect of higher interest rates on the pace of activity in the home economy.

The central bank credibility effect influences the real economy by affecting agents’ inflation expectations. If the persistent anti-inflation actions of a central bank cause it to be viewed as having high anti-inflation credibility, then a tightening policy move intended to reduce inflation could contribute directly to the lowering of inflation expectations and wage demands. Lower labor costs, in turn, could reduce inflation more and at lower output cost, than might have been the case if the central bank had less anti-inflation credibility.

The credit channel is often split into two paths: the bank-lending channel and the balance sheet channel. The bank-lending channel focuses on the fact that bank loans and funds raised in capital markets are not perfect substitutes. Certain types of borrowers, particularly small firms, lack access to capital markets and rely on banks, which have the capability to monitor and screen their activities. But this capability is imperfect—borrowers have more information on their ability to repay a loan than do banks—so that banks also use nonprice rationing devices, such as security checks and evaluation of creditworthiness and deposit performance, as part of the loan approval process.8 When monetary policy is tightened and bank reserves are compressed, the supply of bank lending will be reduced partly through these devices, because banks know that raising lending rates alone would have adverse selection effects. Borrowers who are dependent on banks will then be particularly affected. The balance sheet effect can reinforce this result. When interest rates rise and equity prices decline for the reasons described above, the net worth of firms is reduced, and this tends to reduce the value of their collateral for loans. Increases in interest rates also reduce the cash flow of firms, which increases their risk of default. Prudent lenders therefore lend less and investment spending declines.

Have These Channels of Influence Changed in the 1990s?

While not taking into account all of the factors that could explain changes in the effectiveness of monetary policy in the 1990s, there are several analytical constructs that might be taken as evidence about how these transmission channels are changing.

Structure of Financial Systems

A good beginning point for examining whether the cost of capital channel of monetary policy transmission has changed over time is to look at whether the banking intermediation process has changed. A change in this channel could be reflected in a change in the spread between commercial banks’ lending rates and the interest rates determined directly by monetary policy. A statistical examination of the relationship between these two interest rates for the major industrial countries over two periods, 1975–89 and 1990–96, indicates that there may have been some loosening in this relationship in the 1990s (Table 14). A number of factors may be at work, but changes in the structure and competitiveness of the banking industry in each country are probably important. The linkage between policy rates and lending rates appears to have weakened most in Germany, Italy, and the United States, while there has been relatively little change in the relationship in Japan and Canada.

Table 14.

Selected Major Industrial Countries: Responsiveness of Bank-Lending Rate to a Change in the Policy Interest Rate

article image
Source: These bank-lending rates are from International Financial Statistics (IFS) and are defined as the rates that usually meet the short- and medium-term financing needs of the private sector. For the United States, it is the bank prime rate. The policy interest rates are the federal funds rate for the United States and the money market rate (as defined in IFS) for the other countries. Results are based upon regression analysis. Coefficients on policy rates were significant at the 95 percent confidence level in all cases.

For Japan, data are only available from the fourth quarter of 1980 onward.

For Germany, data are only available from the third quarter of 1977 onward.

For Italy, data are only available from the third quarter of 1982 onward.

The growth of nonbank financial intermediation and capital markets may have caused a change in the transmission of monetary policy. If the banking sector is losing market share in an economy’s overall credit market to bond and equity financing, the potency of changes in the short-term interest rates determined by monetary policy may decline or increase, depending on the responses of financial markets. What is the evidence of such changes? Recent studies show that over the past decade the banking sector has been losing market share to other financial institutions in the United States, the United Kingdom, and Sweden, but not in continental Europe.9 In Japan, by contrast, the banking sector has actually been gaining market share. Ratios of the broad money stock (banks’ monetary liabilities) and equity market capitalization to GDP also suggest that the United States, France, and Sweden have seen increases in the relative importance of their capital markets over the period from 1986 to 1995, with the United States and Sweden having declining ratios of broad money to GDP (Table 15). Meanwhile for Japan, Germany, and Italy, there were smaller changes in the ratios of broad money or market capitalization to GDP over these years. Overall, for countries such as the United States and Sweden, there may have been some reduction in the influence of the bank lending channel in the 1990s, while the power of the cost of capital channel may have become more dependent on the way in which changes in short-term interest rates are transmitted to financial markets, and thus on market expectations.

Table 15.

Changes in the Structure of the Financial System

(In percent of GDP: end of period)

article image
Sources: IMF, International Finance Statistics: and International Finance Corporation. Emerging Stock Markets Factbook 1996.

National definitions. Data for 1995 refer to 1994.

Changes in the term structure of private sector borrowing may also carry implications for monetary policy transmission. If borrowers choose to rely more on long-term bonds and other forms of market borrowing than on shorter-term bank credit, then central bank interest rate movements might have a somewhat more dulled impact on investment spending and real activity. Again, there appears to have been a general movement toward more use of long-term credit in the industrial countries since the early 1980s. Among all of the major industrial countries, for example, the share of private sector borrowing from all sources with a maturity exceeding one year increased between 1983 and 1993, with the increases being particularly noticeable for countries such as the United Kingdom, Sweden, and Japan.10 Because long-term interest rates apply to a larger portion of total liabilities, the power of the changes in short-term rates engineered by monetary policy may be somewhat reduced in most major industrial countries.

Changes in the Central Bank Credibility Effect

On the other hand, the effectiveness of monetary policy may well have been enhanced since the 1980s by improvements in the anti-inflation credibility of monetary authorities. It is, of course, difficult to estimate such credibility and changes in it. But it seems clear that the surest foundation for anti-inflation credibility is establishing a record of maintaining low inflation over time. That foundation has been strengthened significantly in the industrial countries in recent decades, as discussed in detail in Chapter VI. So far in the 1990s, inflation in the industrial countries has averaged just 3½ percent a year, compared with 6 percent on average in the 1980s. And the decline in long-term interest rates that has occurred—from over 10 percent on average in the 1980s to 7.2 percent so far in the 1990s for the industrial countries as a group—may be assumed in large part to be a reflection of declines in expected future inflation (related to the fall in inflation that has actually occurred) and in inflation risk. These and other indicators of expected inflation suggest that the anti-inflation credibility of monetary authorities in the industrial countries has increased in the 1990s.

Changes in the Role of the Exchange Rate

Finally, the increased international integration of financial markets may mean that movements in exchange rates could be more important in influencing how monetary policy affects the real economy and could reinforce the cost of capital effect. A higher policy interest rate would normally slow domestic demand, but it would also tend to cause the domestic currency to appreciate, other things being equal. This would tend to boost imports and slow exports, thereby reducing the net trade balance and providing another contractionary impulse. In practice, it is often difficult to identify the positive relationship between real short-term interest rates and real effective exchange rates that theory suggests because of the effects of changes in exchange rate expectations, monetary policy reactions to exchange rate pressures, changes in foreign interest rates, and so forth. Nonetheless, changes in monetary policy are frequently reflected rather quickly and substantially in the exchange rate. Such a relationship seems visible in quarterly data, for example, in the United States for most of the 1970s and 1980s (a well-known exception being 1984–85), in Canada in much of the period since the mid-1980s, and in Italy and the United Kingdom in the period since 1992, when they dropped out of the ERM of the European Monetary System (EMS) (Chart 18). Higher real short-term interest rates in these cases did tend to lead to real exchange rate appreciation, so that both the cost of capital effect and the exchange rate effect did move in the same direction. But for France, Germany, and Italy and the United Kingdom (for most of the 1980s) such an exchange rate effect is less apparent, partly because of the use of monetary policy to defend pegged exchange rates.

Chart 18.
Chart 18.

Major Industrial Countries: Real Short-Term Interest Rate Differential Versus Real Exchange Rates

1 Differential with Germany.

Has the Overall Impact of Monetary Policy Changed?

Given these developments in the various channels of the monetary policy transmission mechanism, how has the overall impact of monetary policy changed? Are monetary policy movements having slower or faster effects on the real economy? As evidence about changes in the overall transmission of monetary policy in the industrial countries, consider the results of a simple statistical modeling exercise that examines the dynamic effects of a monetary policy shock and the way it influences the level of real GDP over two periods: 1970–89 and 1990–96.11 As shown in Chart 19, simulated responses of real GDP to changes in the policy interest rate were either unchanged or, if anything, somewhat larger in the period 1990–96 than in the period 1970—89 for most of these countries. In terms of timing, these results also suggest that monetary policy affected real GDP somewhat faster in the 1990s than it did in the 1970s and 1980s (see also Box 2).

Chart 19.
Chart 19.

Major Industrial Countries: Simulated Response of Real Output to a Monetary Policy Tightening

Source: IMF staff estimates, based upon simple econometric models.1 The initial (but statistically insignificant) response is a reflection that monetary policy actions often do not have an immediate impact on aggregate activity (see Box 3).

While it is of course impossible to provide a definitive answer to the question about how the transmission of monetary policy has changed in the 1990s, this simulation exercise suggests that monetary policy has not become less potent in most industrial countries.12 The right context for evaluating this issue, however, may be a broader one. Arguably the biggest change in the making of monetary policy in the industrial countries in the 1990s has been the widespread success of central bankers in taming inflation. In this context, it would seem illogical to think of monetary policy as having become enfeebled over the past half decade. Even the rapid growth of the nonbank capital markets in countries such as the United Kingdom, the United States, and Sweden would not seem to undermine the effectiveness of monetary policy. The reason is that capital markets can quickly and powerfully reinforce a change in central bank action, perhaps responding even quicker than can the bank lending channel. The rapid increases in interest rates, both in the United States and worldwide, in response to the Federal Reserve’s tightening action in 1994 seemed to illustrate this dramatically.

The success of monetary policy in maintaining reasonable price stability and in helping to smooth cyclical fluctuations is likely to be greater if a central bank has strong credibility and has delivered a low inflation environment in the past. On balance, monetary policy in the 1990s has demonstrated considerable power to achieve its key objectives.

Banking Sector Risks and Economic Performance

Banking sector difficulties also may influence the transmission of monetary policy and therefore need to be taken into account. Several factors seem to underlie banking problems in the industrial countries, including an oversupply of banking services brought on by financial liberalization and innovation, and the general decline in global real estate prices in the early 1990s. What are the key banking sector risks today? If there are banking system problems in some industrial countries, what might be the economic effects, and what should policymakers be doing to limit risks?

Using the Slope of the Yield Curve to Estimate Lags in the Monetary Transmission Mechanism

It is well understood that the macroeconomic effects of monetary policy occur with significant lags. Consequently, having an early predictor of economic developments is useful in helping monetary authorities to determine the appropriate policy stance. It has been recognized for some time that the yield curve, which shows the term structure of interest rates prevailing in an economy at any point in time, contains information that can be used as an indicator of economic prospects.1 This is because the term structure reflects both the settings of the instruments of monetary policy, as shown in the level of short-term interest rates, and the market’s expectations of future short-term rates, and hence of future growth and inflation. An easing of monetary conditions, by lowering short-term rates, will both tend to steepen a normally upward-sloping yield curve and raise prospective growth and inflation; the opposite would occur for a tightening of monetary conditions. Similarly, a shift in market expectations toward higher inflation or growth, for given settings of short-term interest rates, will tend to raise long-term rates and steepen the yield curve. This indicates why the slope of the yield curve—the difference between long-term and short-term rates—may be a useful leading indicator of macroeconomic developments.

Studies show, for example, that the slope of the yield curve in the United States has significant information content for predicting the future quarterly growth path of real GDP, real consumption, real consumer durables spending, and real investment.2 Indeed, statistical analysis of the relationship between the slope of the yield curve and these measures of economic activity support the view that significant lags exist between a tightening in monetary conditions and its effects on real GDP.

In recent research by IMF staff, this analysis has been extended to changes in unemployment and inflation, as well as real GDP, and to a broader range of industrial countries.3 For each country, cumulative changes in these variables were regressed on lagged values of the difference between the long-term (usually ten-year) government bond yield and a three-month interest rate, with lags of between 1 and 20 quarters. For each country, there was found to be a strong, positive relationship between the yield gap (long minus short rate) and both GDP growth and inflation, and a negative relationship between the yield gap and changes in unemployment. The table presents the lag, in quarters, that was found to maximize the predictive power of the yield gap for each series in each country.4 The estimated lags may be seen to vary considerably across countries. This is to be expected, especially since there are many country-specific factors that affect the extent to which the yield curve reflects market expectations of future economic variables. For example, administrative controls on interest rates, credit controls, and other institutional factors that influence the way in which monetary policy affects the economy will also all tend to affect the relationship. However, on average, the results suggest that in the industrial countries, the maximum “effect” on real GDP growth of changes in the yield gap occurs after about 5 quarters, on unemployment after 8 quarters, and on inflation after 14 quarters.

Number of Quarters Between Changes in the Yield Gap and the Maximum Impact on the Growth of Output, Unemployment, and Inflation

article image

Such a simple econometric exercise cannot be expected to provide a reliable gauge of the timing of the effects of monetary policy. It is interesting, however, to compare the results from this model with the view of the U.S. Federal Reserve that it takes about 18 months in the United States for changes in monetary policy to have their full influence on inflation, and with Bank of England estimates that the corresponding lag in the United Kingdom is about two years.5 The Bank of Canada’s Quarterly Projection Model (QPM) incorporates lags of roughly the durations shown in the table.6 The results for Italy are more questionable, and it seems implausible that the lag to inflation is similar to that for output; these results also contrast with the Bank of Italy’s quarterly model, in which the lag in the effect of monetary policy on real activity is very long (several years).

Notwithstanding the caveats, evidence on the predictive power of the yield curve suggests that data on the term structure may help policymakers avoid inflationary risks and destabilizing policy corrections. The evidence presented also supports a forward-looking approach to policy formulation and points to the appropriateness of a medium-term framework for policies in the presence of “long and variable lags.”

1

See “Information Content of the Yield Curve,” May 1994 World Economie Outlook, Annex II, pp. 89–92.

2

For instance, Arturo Estrella and Gikas A. Hardouvelis, “The Term Structure as a Predictor of Real Economic Activity.” Journal of Finance, Vol. 46 (1991), pp. 555–76; and Zuliu Hu, “The Yield Curve and Real Activity,” Staff Papers, IMF, Vol. 40 (December 1993), pp. 781–806.

3

Irene Chan. “The Term Structure as a Predictor of Economic Activity” (unpublished, IMF, 1996).

4

It should be emphasized that these estimates represent average responses and that in practice the response to any particular change in the yield gap, including the lag, will depend on the origin of the change, and in particular on the timing and magnitude of any monetary policy action that may be associated with it.

5

See “In Search of a Holy Grail.” Economist, September 17, 1994, p. 80.

6

Pierre Duguay and Stephen Poloz, “The Role of Economic Projections in Canadian Monetary Policy Formulation.” Canadian Public Policy, Vol. 20 (June 1994). pp. 189–99.

Banking sector problems that developed in Scandinavia and the United States in the early 1990s provide a valuable frame of reference for looking at current potential vulnerabilities and suggest the magnitude of bad loan burdens that imply trouble. In response to the bursting of the world real estate market bubble of the late 1980s and the subsequent global slowdown, waves of bad loans started hitting Scandinavian banks in the early 1990s. By 1991–92, the annual loan-loss provision burden amounted to 6 percent of total loans outstanding in Norway, 5 percent in Finland, and 7 percent in Sweden. These burdens were simply not sustainable. Numerous banks faced failure, there were large-scale restructurings with forced mergers and consolidations, and governments were saddled with heavy bailout payments. The U.S. banking problems over the same period were precipitated by many of the same factors and the loan-loss burden for commercial banks peaked in 1991 at 2¾ percent of loans outstanding. While it is impossible to establish any hard and fast rule about when banking sector problems become worrisome, these experiences suggest that when an industrial country’s banking system reaches an annual loan-loss burden of 3–4 percent of loans outstanding, action may be needed to prevent a crisis.

Three major industrial countries today have banking systems that are experiencing significant bad loan difficulties: Japan, Italy, and France. In Japan, bad loans arose in large part from the bursting of the asset bubble in the early 1990s, including the collapse of land prices, and the prolonged economic slowdown. Bad loans are officially reported to equal 5 or 6 percent of total loans outstanding, and are estimated by private sources to be as high as 9 or 10 percent.13 In the most recently completed fiscal year, the major Japanese banks made total loan-loss provisions of about percent of loans outstanding. In Italy, where bad loans are also estimated to be on the order of 10 percent of loans outstanding, the bad loan problem is heavily concentrated in the southern part of the country. Poor bank performance there reflects longstanding structural problems and has been aggravated by the regionally unbalanced nature of the recovery. While overall national loan-loss provisions and write offs were only about 1½ percent of loans outstanding in 1995, they may increase in coming years. Authorities have provided about $2 billion of financial support for bank restructuring, and many small banks have been merged with stronger banks. In France, stresses in the banking sector have been due in large part to the decline of real estate values in recent years. In 1995, new loan-loss provisions by the commercial banks amounted to 1 percent of the stock of loans outstanding. The process of provisioning is now considered to be well advanced; the bad loan ratio declined over 1995 from 8.9 percent to 8.1 percent of loans outstanding, with provisions on average in excess of 50 percent against these loans. Underlying banking profitability in France remains rather weak at present, however, for a combination of cyclical and structural reasons.

Two main questions arise. First, if a banking crisis were to occur, what might it mean for a banking system’s supply of credit and a country’s macroeconomic performance? And second, what can policymakers do to reduce the risks of a banking crisis, or to deal with one if it does occur? Japan, a country that has displayed some evidence of a credit squeeze over the past couple of years, provides an interesting case study.

How Bad Loan Problems Affect Credit Supply and the Economy: Japan

In practical terms, bad loan problems and credit supply are linked through the need for banks to maintain an adequate capital reserve. Consider the following evidence from Japan. Japanese banks have been making large loan-loss provisions over the past couple of years, and although profits have supplied the bulk of the funds for these provisions, large amounts of bank capital have also been used. In FY 1995, for example, the 21 largest banks used nearly ¥2 trillion of core capital (as defined by the Basle Committee on Banking Supervision)—about 8 percent of their total core capital—for write-off purposes. This reduced the aggregate core capital ratio for these banks from about percent of total loans to roughly 4½ percent. To continue satisfying core capital requirements, banks have had an incentive to rebalance their assets away from loans, which may have a full capital charge, and toward government securities, for which no capital charge is imposed. It was precisely this asset substitution to repair capital ratios that has been offered as an explanation for the credit squeeze in the United States during 1989–93.14 In addition, internationally active Japanese banks must allocate additional capital to cover market risk in their trading books by the end of 1997, as required by the Amendment to the Basle Capital Accord to Cover Market Risks, although the Japanese authorities hold the view that the additional required capital will not impose a heavy burden on banks. These two factors suggest that lending by Japanese banks will remain under pressure over the next couple of years.

There is already evidence of some credit tightness in Japan, especially among small and medium-sized enterprises.15 Survey data, such as the Bank of Japan Tankan report on banks’ willingness to lend, indicate that although larger firms are reporting a noticeable easing in credit conditions over the past couple of years, small and medium-sized firms have seen little improvement. Previously, credit conditions for these two categories of firms have moved more closely together. Another indicator of credit tightness is the interest rate spread between bank lending rates and a proxy for banks’ cost of funds.16 These spreads widened noticeably over the period from 1991 to 1994; also there has been a noticeable increase in the interest rates charged on loans to small and medium-sized enterprises (SMEs) relative to rates on loans to large firms.17 Finally, a number of empirical studies have found evidence that poor bank health, as captured by credit ratings or the ratio of nonperforming loans to total loans, does have a negative impact on bank lending and investment activity.18 The range of estimates is fairly wide, with poor bank health associated with anywhere from a negligible to a 50 percent cutback in lending or customer investment spending, or both. Again, the impact on small and medium-sized firms is consistently found to be greater than the impact on larger firms.

Investment spending by small and medium-sized enterprises usually provides the leading impetus to growth in normal Japanese economic recoveries—increasing by over 40 percent, on average, in the first two years after a business cycle trough (Chart 20). Given the evidence that there has already been some form of credit squeeze in Japan in recent years, and that it seems to have been concentrated in the small and medium-sized business sector, it is therefore perhaps not surprising that investment spending by these firms has continued to decline and that the pace of the recovery to date has been disappointing. The fact that even medium-sized firms have been relatively unsuccessful in raising capital in the corporate bond market in Japan has been another factor. The risk is that the credit squeeze could worsen over the next couple of years if banks attempt to cut back on their lending in an attempt to strengthen their balance sheets. If investment spending remains bogged down because of a lack of credit, this could significantly retard growth.

Chart 20.
Chart 20.

Japan: Investment Spending Over Recovery Cycles, Small and Medium-Sized Enterprises

(Four-quarter moving average, indexed to recession trough quarter indicated)

Source: Japan Development Bank; small and medium-sized enterprises are firms with share capital of ¥1 billion or less.

What Policy Should Do When Banking Sector Problems Occur

Banking system problems can be costly for an economy and can limit the effectiveness of monetary policy. They can be costly because the efficiency of financial intermediation is reduced when banks are being closed on a large scale or are focusing on strengthening their balance sheets, and when they constrict the flow of credit to worthy borrowers and force the cancellation of otherwise profitable investment projects. Banking sector problems can reduce the effectiveness of monetary policy because banks that are in trouble may not react normally to interest rate changes, and because central banks may become too cautious about tightening policy in situations that require it, for fear of damaging fragile banks. Policymakers need to monitor banking sector risks carefully to try to minimize these losses. Prevention is always the best first step. This means that monetary policy should try to prevent the buildup of asset bubbles in the first place, whether in real estate or equity markets. Admittedly this can be a tall order because of the difficulty of assessing inflationary pressures and the significance of asset market pressures. It may also be desirable to have some diversification of a country’s financial system beyond just one main class of intermediary and its regulators. Corporate bond and equity markets, which do not need to be regulated in the same way as banks, offer a country useful financial diversification.

It is particularly critical to ensure that the banking sector is regulated and supervised prudently and critically. There will always be a natural tension between bank regulators and the banking industry, because the industry comprises firms whose objectives and priorities differ from those of the authorities. Regulators must allow banks to pursue profitable business and to develop new instruments and services, yet at the same time they must insist that banks carefully manage their assets and make sure that the riskiness of banking operations is acceptable. They must engage in a constant process of probing, analyzing, and questioning banks’ operations, and assessing overall bank risk. When a banking crisis does occur, the banking authority, as the lender of last resort, has an obligation to deal with the consequences at minimum current and future cost, whether through closure, merger, or some other form of rehabilitation. Giving the proper signals is key to containing such problems. Owners, managers, and depositors of failed institutions need to face the prospect of appropriate penalties to ensure that moral hazard problems are minimized. The use of fiscal resources in the bank-restructuring process may be justified in some cases, in part because it may prevent a banking problem from overburdening other policy tools. Thus, for example, a banking problem that authorities attempted to deal with by easing monetary policy might conflict with the objective of price stability.

On the other hand, achieving reasonable price stability would probably make it easier for monetary policy to help offset the potentially contractionary effects of a banking crisis on credit extension, without risking a loss of anti-inflation credibility. To gauge the appropriate stance of monetary policy, policymakers of course pay attention to early warning signals of credit tightness, such as surveys of lending practices by banks, bank lending spreads overall and lending rates changed to different classes of borrowers, and divergent investment patterns by large and small firms. If the bank lending channel of credit transmission were to be constricted and a credit squeeze were to develop, investment spending would be reduced, GDP growth would tend to slow, the output gap would tend to grow larger, and inflation would normally decline. An easier monetary policy made feasible by low inflation would mean a lower cost of funds for banks, and that would tend to boost bank profitability, which in turn would help to repair bank balance sheets and stimulate the bank lending channel. There are cases where this seems to have helped. In the United States, real short-term interest rates were reduced to roughly zero over the year and a half from the summer of 1992 to the end of 1993, and this development, at a time of banking sector difficulty, clearly assisted in the eventual recovery of the U.S. banking sector. In Japan also, exceptionally low short-term interest rates are helping the banking sector to recover.

Easier monetary policy made feasible by low inflation would also help to improve corporate balance sheets, and therefore have a positive balance sheet effect. This, too, would help to restore economic growth. There are also good arguments for financial authorities to clarify the containment of responsibilities for bad debts so that healthier banks and institutions can get on with their lending operations, and not be saddled with handicaps unrelated to their own operations. The temporary emergence of the “Japan premium” in 1995, which added to the costs not only of relatively troubled, but also of relatively strong and healthy banks, illustrates the problems that delays in resolving systemic banking sector imbalances can cause.

Strengthening the Functioning of Labor Markets

Another key policy challenge for industrial countries is the promotion of high employment. The maintenance of high levels of employment is a primary objective of economic policy in all countries and is essential to avoid social exclusion and the waste of economic resources, to combat poverty, and to promote economic welfare. Labor market flexibility helps to maintain high employment in the face of economic disturbances and the changing fortunes of economic sectors and regions. In a rapidly changing global economy in which market conditions are constantly evolving, responsive labor and product markets ensure that resources are quickly matched to their highest value uses, thus promoting efficiency. There are also macroeconomic policy benefits. More flexible labor and product markets allow an economy to adjust to adverse shocks and shifts in policy more rapidly, and hence allow policymakers to achieve economic goals more fully at lesser costs in terms of unemployed resources and forgone output and income.

Differences Between U.S. and European Labor Market Experiences

From a long-term perspective, the remarkable divergence in labor market conditions between the United States and Europe helps to identify the reasons for labor market shortcomings. U.S. job creation has been quite impressive over the past twenty-five years, with employment having expanded by roughly 60 percent, in step with the rapid growth in the labor force, in part reflecting a sharp increase in the participation rate of women. The U.S. unemployment rate has thus remained relatively stable, although oscillating somewhat with the economic cycle (Chart 21). Taking the same 25-year view, the Japanese economy also has tended to operate at a high level of employment and a relatively low level of unemployment. In Europe, however, employment has increased by far less than the labor force over the period since 1970: total employment is up a bare 11 percent, with private sector employment up by just 5 percent, and the unemployment rate has climbed steadily, from a little above 2 percent in the early 1970s to 11–12 percent today. Estimates of cyclically adjusted structural unemployment rates in Europe have climbed in correspondence with overall unemployment rates. Rigid wage structures and inflexible labor markets are widely understood to be the primary causes of these increases in structural unemployment. Persistently weak cyclical performance in Europe has also probably contributed to the structural unemployment problem.

Chart 21.
Chart 21.

European Union and the United States: Output, Capital, Employment, and Compensation

(1970 = 100: unless noted otherwise)

Sources: OECD, Analytical Data Bank; and IMF staff estimates.1 Excluding Greece, Luxembourg, and Portugal. Real GDP. capital stocks, and real compensation are calculated using PPP-based weights.2 The increase in employment in 1991 is largely due to the unification of Germany.

Real compensation per worker in Europe has increased noticeably relative to the United States—by almost 2 percent a year over the past twenty-five years compared with roughly ½ of 1 percent annual growth in U.S. real earnings. Some of this relative gain in European earnings probably reflects a continuing catch-up to U.S. productivity levels in the 1970s and 1980s. Higher rates of investment and saving have also contributed to European earnings gains by allowing more rapid increases in capital-labor ratios than has been possible in the United States, where saving has remained relatively low. To some extent, however, the relative rise in European capital-labor ratios has also reflected the desire of European businesses to substitute capital for high cost labor. This may have contributed to a vicious circle, whereby this rising capital-labor ratio has increased the productivity of the workers who remain employed and given them a strong position in wage negotiations. This may help to explain the upward pressures on wages that seem to characterize many European economies at times of high unemployment.

While the U.S. labor market has been flexible and has delivered steady job growth, many observers have criticized the quality of the jobs, the real wage stagnation, and the income inequality that the U.S. “model” is widely perceived to have delivered. Some recent statistical analysis, however, has found that the jobs being created in the United States—while often in the service sector—are not typically low-paying, low-quality jobs, but rather jobs that pay above-median wages.19 Many jobs in the U.S. service sector, for example, pay relatively high wages—including such categories as accountants, managers, technicians, financial services workers, and computer programmers. Recent data also suggest that the vast majority of new U.S. jobs are full-time rather than part-time, that the share of workers holding multiple jobs has remained roughly constant since the late 1980s, and that the displacement rates for older, white-collar, and better educated workers, though somewhat higher than in the 1980s, are still low relative to the workforce in general. Cross-sectional income distribution does tend to be more unequal in the United States than in Europe, and those with the lowest skill and education levels tend to fare much worse. There is evidence, however, that income mobility may be higher than generally appreciated—both in the United States and in Europe—and that many individuals can expect to rise to substantially higher relative income levels over the course of their working careers.20 This could help to foster greater social cohesion than cross-sectional income distribution data alone might imply.

A particular concern with the U.S. labor market is the perception that a sizable fraction of the workforce at the bottom of the income distribution, including residents of inner cities and those with poor educations and limited work skills, do not have good prospects for moving up the income distribution. The recently enacted reforms to the welfare system in the United States focus on improving work incentives in the system by imposing work requirements on recipients and by establishing a time limit on eligibility for benefits. Thus, an important policy challenge for the United States in this regard is to improve the delivery of education and training for these groups. In the United States, as in all industrial countries, less-skilled workers are at a competitive disadvantage—often facing twice the unemployment rate of workers with higher education qualifications. As an approach to improving the economic well-being of low-income workers, the Earned Income Tax Credit has been effective in providing incentives to encourage work among the less well off, and from an employment policy standpoint, it may be preferable to an increase in the minimum wage, such as has recently been adopted as a means of improving the economic well-being of low-income workers.

Pressures of Globalization and Opportunities

There is growing concern that the increased globalization of the world trading system, especially through the integration of the emerging market countries, is a key factor contributing to higher levels of unemployment in the industrial countries, and in Europe in particular. Is this a valid concern? While the issue is complicated, the evidence to date does not indicate that globalization makes unemployment increase. Globalization does put pressure for change on labor markets and may expose countries with substantial labor market rigidities to risks of higher unemployment. Indeed, the natural working of the forces of competition and comparative advantage means that some groups of workers will always be underpriced by foreign competitors, that some of their countrymen will be international winners outproducing their foreign rivals, and that by adapting to global price signals and shifting the composition of output, overall welfare can be maximized. That is, countries with flexible labor markets can better exploit the gains from trade, and enjoy greater benefits and suffer smaller costs from globalization, than countries with rigid labor markets. If European countries feel that they are losing jobs because of the forces of globalization, the correct policy is to liberalize labor markets and thereby encourage the redeployment of workers into higher value-added sectors. In fact, such redeployment should help to raise national income and enhance living standards.

New Efforts at Labor Market Reform

Among countries that have suffered from rigid labor markets, there are some indications that the costs of protecting workers from the vagaries of the market have risen to levels deemed unsustainable, and that acceptance of the need to increase labor market flexibility may be increasing. New Zealand, for example, has implemented a wide range of labor market reforms, and among European countries, the United Kingdom has implemented far-reaching reforms. Collective bargaining and trade union structures have been reformed in these countries to increase competition, and in the United Kingdom, minimum wages have been abolished for most of the groups of workers to which they applied. A number of reforms have also been introduced in other countries. Recent legislation to make labor markets more flexible includes the allowance of more fixed-term contract workers in Germany, France, and Spain; less-stringent rules for shift work in France and Finland; and easing the conditions for layoffs in France. Various attempts have also been made to reduce the generosity of unemployment insurance (UI) schemes in Canada and most European Union members. In France, for example, reforms in 1993 reduced the overall duration of benefits and caused benefits to decline with the duration of unemployment, while in Canada there are already signs that UI reform efforts have been having salutory effects.

Have these labor reforms made a difference? One way to quantify progress is to look at movements in estimates of the natural rate of unemployment, which can be defined as the lowest rate of unemployment that is consistent with nonaccelerating wage settlements. Such estimates are based upon a range of techniques, including econometric estimation and benchmarking at assumed periods of full employment. These estimates should be treated with caution and should be considered to be subject to wide confidence bands.

Bearing these qualifications in mind, IMF staff estimates suggest that there has been progress in reducing the natural rate of unemployment in countries where substantial measures have been taken to reform their labor markets. In the United Kingdom, for example, the unemployment rate is currently 7½ percent—the lowest in five years—and there are only limited indications to date of building wage pressures, suggesting that the natural rate has declined from earlier levels. In New Zealand, labor market reforms have been widely credited with contributing significantly to the country’s favorable performance in recent years. The unemployment rate has declined from 11 percent in 1992 to about 6 percent currently. And Irish unemployment has fallen sharply from 17 percent in 1993 to 13 percent currently. Despite the continued surge in output and employment in Ireland, inflation remains subdued, suggesting that the structural unemployment rate may still be falling. Even the United States seems to have experienced a modest decline in its natural rate of unemployment over recent years. Whereas most estimates of this rate ranged between 6 percent and 6½ percent as recently as a couple of years ago, the consensus now, given the apparently subdued rate of inflation, is that the range may now be from 5½ to 6 percent and some analysts suggest that it may be even lower. Partial deregulation of the Australian labor market has left that country with a hybrid system of collective and enterprise bargaining that probably represents only a mixed success. While there has been a significant decline in Australian unemployment during the current economic expansion, the natural rate of unemployment is estimated to have remained steady at about 7½ percent. Countries such as New Zealand, Ireland, and the United States have experienced fairly strong economic growth in recent years, which means that their labor markets have been “tested” by actual conditions of tightening demand, and generally they seem to be posting encouraging results. That is, lower unemployment rates do not seem to have created undue wage pressures.

There is little evidence to date to suggest that natural rates of unemployment have fallen much in continental Europe. IMF estimates of natural rates of unemployment are currently on the order of 8 to 10 percent for Germany, France, and Italy, and projections for the year 2000 foresee only modest declines in most countries. Still, there can be little doubt that margins of slack in labor markets in continental Europe are substantial, and there seems to be little risk of an early reemergence of wage pressures even under conditions of relatively robust growth. There has also been underutilization of these labor forces through part-time work, increasing resort to early retirement, and other discouraged-worker effects.

Measures That Might Help

There are a number of measures that might foster reductions in unemployment rates.21 Among the steps that might be considered would be a substantial scaling back of passive support measures, including lower benefit levels for the unemployed; shorter durations and tighter eligibility criteria and job-search requirements for unemployment benefits; and tighter eligibility criteria for disability benefits. Other steps might include replacing passive income support measures with a negative income tax conditional on employment or job search, reductions in payroll taxes (particularly for low-wage workers), reductions in minimum wages, a substantial scaling back of job security legislation, and an end to government extension of collective bargaining agreements to cover workers and enterprises not party to the original agreement. Governments could also implement measures to lower job-search costs by increasing worker mobility and allow wage differentials to better reflect productivity differences—and hence improve employment prospects for new entrants to the labor force and low-skilled workers. Many of these steps would reduce the bargaining power of insiders and increase the influence of outsiders on wages.

Why have European countries seen only modest progress to date in reducing structural unemployment? One reason is that there has been excessive timidity in labor market policies. Another reason may be that a wide range of labor market institutions—including unemployment benefits, job security legislation, and payroll taxes—have strongly complementary effects on unemployment, suggesting that a broad range of reforms designed to reduce rigidities, like those described above, may be needed simultaneously to reduce structural unemployment.22 According to this view, it should not be surprising if piecemeal reforms enjoy only limited success, especially changes that fail to look at overall incentive effects. Unfortunately, fundamental reform is not easy. Any package of reforms will have to address the distributional issues in a straightforward manner and will be politically challenging.

The real risk in Europe is that in the absence of comprehensive labor market reforms, unemployment rates could continue their steady upward march. This underscores the urgency of tackling the problem more forcefully. The challenge policymakers need to confront is to mobilize public support for reform while overcoming fears that a deregulated labor market would seriously jeopardize Europe’s social fabric and mean the end to the European welfare state. Of course, the market cannot solve all problems and governments clearly have an important role in addressing distributional concerns. But such interventions need to minimize labor market distortions by working primarily through the tax system, by ensuring that there are strong incentives to create and seek employment, through reforms that promote mobility—for example, by allowing the portability of pensions—and through enhanced education and training. Meeting equity objectives through such channels should help to reduce resistance to allowing market forces a much greater role in clearing the labor market. The outcome of such reforms would be higher levels of employment and clearly higher levels of welfare for the presently unemployed. From an economic perspective, such reforms are among the most important conditions for the success of the planned monetary union.

* * *

Just as there may be complementarity among various labor market reforms, so is there also a broader complementarity between labor market reforms and macroeconomic policies and performance.23 A flexible and dynamic economy, which responds quickly to changing supply and demand, provides the most favorable structural environment for monetary policy to maintain reasonable price stability while also contributing to the stability of aggregate output and employment. Conversely, a monetary policy that assures reasonable price stability and helps to avoid large deviations of output and employment from their sustainable growth paths provides the best macroeconomic environment for households and businesses to recognize and benefit from the opportunities available in a flexible market-oriented economy.

7

See, for example, the recent set of articles in the Journal of Economic Perspectives, Vol. 9 (Fall 1995), including those by Frederic S. Mishkin, John B. Taylor, Ben S. Bernanke and Mark Gertler, Allan H. Meltzer, and Maurice Obstfeld and Kenneth Rogoff, pp. 3–96.

8

For a more complete explanation of this channel, see William E. Alexander and Francesco Caramazza. “Money Versus Credit: The Role of Banks in the Monetary Transmission Process.” in Frameworks for Monetary Stability Policy Issues and Country Experiences, ed. by Tomás J.T. Baliño and Carlo Cottarelli (IMF, 1994), pp. 397–422.

9

See Claudio Borio, “The Structure of Credit to the Non-Government Sector and the Transmission Mechanism of Monetary Policy: A Cross-Country Comparison,” BIS Working Paper No. 24 (Basle: Bank for International Settlements, April 1995), pp. 1–49.

10

See Borio, “The Structure of Credit.”

11

Simple vector autoregression models of the monetary transmission mechanism were estimated for the seven major industrial countries. Real GDP was specified as a function of the short-term policy interest rate, the broad money supply, the inflation rate, and the unemployment Rate. The interest rate was then increased by one standard deviation and the model was used to trace out a path for real GDP.

12

John Taylor, in his article, “The Monetary Transmission Mechanism: An Empirical Framework,” Journal of Economic Perspectives, Vol. 9 (Fall 1995), pp. 11–26, argues that monetary policy in the United States in the late 1980s and early 1990s may have been somewhat less potent than in earlier years. This is based upon his view that the key channel of monetary policy is through the money supply rather than through the policy interest rate, which the simulation exercise above assumes. Taylor finds little evidence of a change in the overall effectiveness of monetary policy in Japan and Germany.

13

Higher range is estimate of the International Banking Credit Agency, May 1996.

14

This argument has been put forth recently by Anjan Thakor in “Capital Requirements, Monetary Policy, and Aggregate Bank Lending: Theory and Empirical Evidence,” in Journal of Finance, Vol. 51 (March 1996), pp. 279–324.

15

Admittedly it is difficult to distinguish between tight credit conditions caused by a bank capital shortage, on the one hand, and by a decline in the creditworthiness of potential borrowers, on the other—that is, the difference between the bank lending channel and the balance sheet channel described earlier.

16

This indicator is suggested by Ben S. Bernanke and Mark Gertler, “Inside the Black Box: The Credit Channel of Monetary Policy Transmission,” in Journal of Economic Perspectives, Vol. 9 (Fall 1995), pp. 27–48.

17

See “An Analysis of Banks Lending Interest Rate Spreads” (in Japanese), 1995 Annual White Paper (Tokyo: Economic Planning Agency, December 1995), pp. 142–44. This study found roughly a 50 basis point premium on loans to SMEs over loans to large firms, whereas in a similar economic period in the mid-1980s there was almost no SME premium.

18

See Michael S. Gibson, “Can Bank Health Affect Investment? Evidence from Japan,” Journal of Business, Vol. 68 (July 1995), pp. 281—308; Tsutomu Maeda “Slow Growth in Bank Lending in Japan: An Empirical Analysis of Banks‘ Unwillingness to Lend” (in Japanese), Financial Review, Japan Ministry of Finance, Institute of Fiscal and Monetary Policy (March 1996), pp. 131–53; Hiroshi Yoshikawa, Masaru Eto, and Toshihiru Ike, “Unwillingness of Banks to Lend to Small- and Medium-Sized Enterprises” (in Japanese), Japan Economic Planning Agency, Economic Research Institute (March 1994), pp. 1–43. Another Gibson paper, “More Evidence on the Link Between Bank Health and Investment in Japan,” International Finance Discussion Paper No. 549 (Washington: Board of Governors of the Federal Reserve System, May 1996), using more recent data than the first, finds much less significant effects, especially for large firms.

19

See, for example, the U.S. Council of Economic Advisers Report. “Job Creation and Employment Opportunities: The United States Labor Market, 1993–96, April 23, 1996.” available on the Internet at http://www.whitehouse.gov/WH/EOP/CEA/html/labor.html.

20

For example, see Employment Outlook (Paris: OECD, July 1996), pp. 59-l08, which reports that only a minority of low-paid workers continued in low-paying jobs over the five-year period 1986–91. Stephen Rose, in “Declining family Incomes in the 1980s: New Evidence from Longitudinal Data,” in Challenge. Vol. 36 (November-December 1993), pp. 29–36, argues that although upward mobility tor prime-age U.S. adults was lower in the I980s than it was in the 1970s, average real family income increased by 33 percent over the 1980s, with all income quintiles except the bottom quintile posting gains. According to his data, even individuals with fewer than 12 years of education saw their real family income INCREASE BY 14 percent over the 1980s and had one chance in five of seeing their real income increase by at least 70 percent. Young workers, those aged 22–33 in 1989, had about one chance in four of seeing their real income increase by at least 70 percent over the 1980s.

21

One of the most extensive discussions of the problems of structural unemployment to date is in the OECD Jobs Study: Evidence and Explanations (Paris: OECD, 1994). The proposals detailed in this section are consistent with the recommendations in that study.

22

See, for example. David Coe and Dennis Snower. “Policy Complementarities: The Case for Fundamental Labor Market Reform.” IMF Working Paper 96/93 (August 1996).

23

This view is put forth by Assar Lindbeck in “The West European Employment Problem.” in Weltwirtschafliches Archiv (forthcoming).

  • View in gallery

    Major Industrial Countries: Real Short-Term Interest Rate Differential Versus Real Exchange Rates

  • View in gallery

    Major Industrial Countries: Simulated Response of Real Output to a Monetary Policy Tightening

  • View in gallery

    Japan: Investment Spending Over Recovery Cycles, Small and Medium-Sized Enterprises

    (Four-quarter moving average, indexed to recession trough quarter indicated)

  • View in gallery

    European Union and the United States: Output, Capital, Employment, and Compensation

    (1970 = 100: unless noted otherwise)