Abstract

Poor outcomes with regard to both growth and investment have been widespread among highly indebted countries since 1982, aggravating the burden of foreign obligations relative to domestic resources and worsening the debt situation. In part, this disappointing economic performance undoubtedly reflects the policy imbalances that gave rise to the debt problem in the first place.1 There is also a widespread view, however, that the debt burden has itself exacerbated the economic situation in highly indebted countries. This view is based on the observation that the significant reduction in the current account deficit of these countries since 1982 was achieved through a large drop in domestic investment, which presumably had adverse effects on their growth performance. Proponents of this view, which is sometimes labeled the “debt overhang” hypothesis, argue that when foreign debt becomes “excessive,” actual payments to creditors become linked to the economic performance of the debtor country. Therefore, potential increases in debt payments depress the returns to productive investment and discourage capital formation.

Supplementary Notes

Supplementary Note 1: Investment and Growth in Heavily Indebted Countries

Poor outcomes with regard to both growth and investment have been widespread among highly indebted countries since 1982, aggravating the burden of foreign obligations relative to domestic resources and worsening the debt situation. In part, this disappointing economic performance undoubtedly reflects the policy imbalances that gave rise to the debt problem in the first place.1 There is also a widespread view, however, that the debt burden has itself exacerbated the economic situation in highly indebted countries. This view is based on the observation that the significant reduction in the current account deficit of these countries since 1982 was achieved through a large drop in domestic investment, which presumably had adverse effects on their growth performance. Proponents of this view, which is sometimes labeled the “debt overhang” hypothesis, argue that when foreign debt becomes “excessive,” actual payments to creditors become linked to the economic performance of the debtor country. Therefore, potential increases in debt payments depress the returns to productive investment and discourage capital formation.

This note discusses the links between high foreign debt, investment, and growth, by reviewing the main theoretical arguments and by presenting a preliminary empirical evaluation of certain implications of the debt overhang hypothesis. The note is divided into four parts. First, the recent experience of highly indebted developing countries with regard to growth and investment is reviewed. This is followed by an examination of the analytical arguments that link debt problems to reduced growth rates via depressed rates of investment. Certain implications of these arguments are drawn out and evaluated empirically in the third section. Finally, some rough calculations are presented to assess how much of the growth slowdown can be accounted for by the drop in investment, and other possible causes of the slowdown are suggested.

Background

Following the sharp reduction in the flow of external financing beginning in 1982. debtor countries found it necessary to adjust drastically their external current account balances. For example, the aggregate current account deficit for the group of 15 heavily indebted countries identified in the Baker Initiative fell from an annual average rate of almost $36 billion in 1978–82 to just over $8 billion in 1983–87. This large reduction coincided with weak growth and an investment performance that was poor both by historical standards and in relation to other developing countries, leading to the presumption that these developments might be related. In Charts 25 and 26, average rates of growth of GDP and investment ratios for the period 1976–81 are compared with the corresponding values for the period 1982–88 for the 15 heavily indebted countries and for a group of other middle-income countries. The latter group consists of indebted countries that have not confronted serious debt-servicing difficulties in the sense of having undergone frequent rescheduling of debt obligations or having faced limited access to international financial markets, at least since 1982. This group consists of Indonesia, Korea, Malaysia, Thailand, and Turkey.

It is clear from Chart 25 that the growth performance in the period since 1982 was considerably weaker than in the previous six-year period, particularly among the 15 heavily indebted countries. Eleven of the 15 countries in this group experienced average annual GDP growth rates below 2 percent over the period 1982–88 while the 5 debtor countries without servicing difficulties had growth rates averaging at least 4 percent annually over the same period. An even sharper contrast between the two groups of debtors is evident in Chart 26. While the investment to GDP ratios of countries without debt-servicing difficulties are substantially unchanged, all but one of the problem debtors have experienced a reduction in investment in relation to GDP.2

Chart 25.
Chart 25.

GDP Growth Rates for Selected Developing Countries: 1982–88 vs. 1976–811

(In percent)

1 Points above the 45 degree line indicate higher growth in 1982–88. and points below the 45 degree line indicate higher growth in 1976–81.

The evidence presented in Charts 25 and 26 should be interpreted with caution. First, it should be noted that even though the worsening of the debt situation was spread over several years, the bulk of the decline in investment to GDP ratios took place in 1982–83. Also, it is not possible to infer from the charts whether the poor performance of investment and growth caused debt problems or whether the buildup of external debt itself reduced incentives to invest which, in turn, hindered economic growth. Indeed, the observed drop in investment could have resulted from high domestic interest rates and weak capital inflows, with little or no contribution from the debt overhang. A further analysis of the debt situation is needed in order to distinguish among these alternative hypotheses.

Chart 26.
Chart 26.

Investment/GDP Ratios for Selected Developing Countries: 1982–88 vs. 1976–811

(In percent)

1 Points above the 45 degree line indicate higher growth in 198288. and points below the 45 degree line indicate higher growth in 1976–81.

Nature of the Debt Overhang Problem

In situations in which a debtor country is not fully performing on its external obligations, debt payments (or the conditions obtained in a rescheduling agreement) are likely to become linked to the economic performance of the country. For example, if the debtor country suffers unfavorable international price developments or any other adverse shock, its debt-service payments might decline. Conversely if the value of its exports increases, a part of the additional resources may have to be used to service the debt.3 In such circumstances, the debtor country shares only partially in any increase in output and exports because a fraction of that increase will be used to service the foreign debt. This may weaken incentives to invest from the point of view of the debtor country as a whole, because the effective return to investment is reduced. As will be discussed below, this disincentive may also apply to the private sector, even when the debt is held by the government.

This disincentive problem, which has been referred to as the “debt overhang,” arises only if the debtor country is unable to meet its payment obligations in a normal way, and involuntary lending takes place. In contrast to normal debt service, the amounts actually paid to service the foreign debt will depend on the outcome of direct negotiations between the debtor country and its creditors rather than on scheduled interest and amortization. This process may lead to a situation in which increases in production and in exports are at least partly used for payments to foreign creditors, which gives rise to the disincentive effect.4 Therefore, the debt overhang hypothesis does not describe a situation in which foreign debt is merely large, but one in which the existence of foreign debt distorts the relevant margins considered for production and investment decisions.

Even when all foreign debt is held by the government, the debt overhang problem may spill over to private saving and investment. This is because the government would have little incentive to pursue policies that stimulate private savings and investment when debt payments absorb most of the gains to the country. By the same token, the government would experience a weakened resolve to undertake any adjustment policy that implies a reduction in current consumption (public or private) or other politically costly measures in exchange for improvements in future growth and export earnings.

The theoretical case for the investment disincentive hypothesis is not conclusive, however, because offsetting mechanisms may also be at work. Suppose, for example, that an explicit or implicit rule has been reached for a nonperforming debtor under which debt payments are a small fraction of GDP, say 2 percent. It is unlikely that the government of the debtor country would greatly reduce its growth objectives, or that the private sector would face severe disincentives, because of this kind of debt service. Also, a consumption-smoothing mechanism may operate. Existing debt implies future payments which must be met out of lower consumption. In anticipation of those payments, consumers (or the government) may reduce current spending and accumulate productive assets so as to smooth out their effects on consumption. In this case, domestic saving would increase, as would domestic investment if capital mobility were imperfect and if capital flight could be ruled out.

It is important to identify the major determinants of the drop in investment, distinguishing between the debt overhang and other factors (perhaps only indirectly related to the debt level), because they may have entirely different policy implications. For example, a vital question is whether the economy of a debtor country would benefit more from debt reduction or an increase in international lending. If the fall in investment is partly a consequence of the debt overhang, in the sense defined above, a significant reduction in the stock of foreign debt would generate an increase in private investment by removing an important disincentive. Paradoxically, an increase in lending in this case could diminish incentives to invest by the private sector, if it did not generate a commensurate increase in government revenues or in the tax base in the debtor’s economy. This would be the case, for example, of wasteful public investment or even of otherwise sound public investment projects that did not generate revenue for the government, such as roads and bridges, or schools and hospitals. By contrast, if the drop in investment has been the result of weak capital inflows rather than of a debt overhang problem, a reduction in the stock of debt without a large change in current net capital flows would not be expected to increase investment greatly, since a debt reduction would not increase the supply of foreign savings to the debtor country or reduce domestic real interest rates.

Some Evidence on Debt-Related Investment Disincentives

While the “debt overhang” proposition provides a plausible explanation for the decline in the investment-output ratio in developing countries after 1981, it has not yet been subjected to systematic empirical verification. This task remains on the agenda for future research. It is possible, however, to provide some preliminary evidence of the likely empirical importance of the hypothesis. This section examines three types of evidence that shed some light on the possible role of the debt overhang in explaining the recent behavior of investment in developing countries. This evidence is based on comparisons of consumption, investment, and the composition of investment between the public and private sectors across different groups of countries.

The first piece of evidence comes from the behavior of the consumption-output ratio. As a result of the sharp reduction in the availability of external financing to developing countries after 1981. the indebted countries were required to generate substantial trade surpluses in order to service their debt. As the trade balance is the difference between output and absorption, adjustment implies a reduction in the ratio of absorption to GDP. Adjustment in the trade balance can be achieved, however, by reducing the consumption-output ratio or the investment-output ratio. The actual form of the adjustment would likely depend on the nature of the shock. For example, an exogenous increase in debt would imply a reduction in national wealth, and would thus generate a large fall in consumption as agents adjust to the permanently lower level of wealth. Similarly, an autonomous desire on the part of the debtor country to improve its trade balance would lead to a reduction in consumption rather than investment, because investment should be determined by an evaluation of its profitability alone.

Events in the heavily indebted countries have not unfolded in accordance with either of the scenarios described above, suggesting that neither the large stock of debt nor the attempt to improve the trade balance can in and of themselves explain the behavior of investment in these countries. Panel A of Table 15 presents the consumption-output and investment-output ratios for the 15 heavily indebted countries over the past several years. It is evident that the reduction in absorption relative to output in these countries after 1981 has been fully reflected in a lower investment-to-GDP ratio. Contrary to what is suggested by the simple requirements of trade balance improvement or by the wealth effects of external debt, the consumption-output ratio in these countries has not only failed to decline in proportion to the investment-output ratio, but has actually risen on average. This behavior of the consumption-output ratio is consistent with the presence of “debt overhang” disincentives to investment, but it is not a predictable consequence of the need to improve the trade balance, or of the negative effects of debt on wealth. Alternatives to the “debt overhang” hypothesis must be able to provide an explanation for this behavior of consumption, as well as that of investment.

Table 15.

Tests of “Debt Overhang” Implications in Developing Countries, 1975–87

article image

The “debt overhang” hypothesis suggests that disincentives to invest should be particularly strong when a country is not fully servicing its debt, because in this case the amount of debt service is likely to be related to a country’s performance, so that part of the marginal return to investment accrues to external creditors. Disincentive effects are therefore related not only to the size of the debt, but also to the existence of debt-servicing problems. Thus the empirical relevance of the hypothesis can also be evaluated by contrasting the behavior of investment-output ratios between groups of countries with and without recent debt-servicing problems. (See Panel B of Table 15.) For countries with debt-servicing problems, the average investment-output ratio fell from about 25 percent in 1980–81 to about 18½ percent in 1987, a reduction of almost a third. By contrast, the group of countries without debt-servicing problems experienced very little change in its investment-output ratio, from about 28 percent in 1980–81 to 27¼ percent in 1987. Moreover, if the more recent numbers are compared with the situation prior to 1980. the contrast is even more striking, since countries with debt-servicing problems exhibited even higher investment-output ratios in 1975–79. while for those without debt-servicing problems the years 1980—81 were a period of unusually strong investment.

Assuming that the availability of external finance has been curtailed (albeit not equally) for all market borrowers, this test may be much weakened by the inclusion in the group of countries without debt-servicing problems of a large number of countries which are not market borrowers. Such countries have not been affected in the same way as market borrowers by the post-1981 reduction in private capital flows. Panel B therefore restricts the comparison to market borrowers, thus comparing two groups of countries which have been affected by the curtailment of private lending, but which differ with regard to their debt-service experience. Again, those market borrowers with recent debt-servicing problems have experienced sharp declines in their investment-output ratios, consistent with the presence of “debt overhang” disincentives. In this case there is evidence that factors other than investment disincentives may be at work, since market borrowers without debt-servicing problems have also experienced a drop in investment-output ratios, from 30¾ percent in 1980–81 to 26 percent in 1986. The reduction, however, appears to be considerably less pronounced for this group if the more recent period is compared with 1975–79, when the investment-output ratio of these countries averaged 28½ percent. Thus, the cause of the drop in investment appears to be largely specific to countries with debt-servicing problems, rather than to heavily indebted market borrowers per se.

A third indication of the potential empirical importance of debt-related investment disincentives is provided by changes in the composition of investment in the indebted countries. If debt service depends on overall macroeconomic performance, disincentives should apply to both the private and public sectors. In contrast, a “fiscal crunch” related to debt-servicing difficulties in the public sector combined with reluctance to cut government consumption for political reasons would lead to a squeeze on public sector investment but not necessarily on private investment. This scenario may even be consistent with an increase in private investment if a cut in government outlays would normally “crowd in” private investment.

Panel C of Table 15 presents information for a group of indebted developing countries for which data on the public-private composition of investment are readily available. The disaggregated behavior of investment is broadly consistent with the existence of “debt overhang” disincentives, in the sense that both public and private investment ratios drop from 1981 to 1984. There is some support, however, for the “fiscal crunch” view as well, because the reduction in the investment ratio is more severe for public than for private investment. If public spending typically crowds out private investment,5 however, the “fiscal crunch” view cannot account for all the facts, since it would have predicted that the reduction in public investment ratios would have been associated with a rise in private investment ratios. Thus, the data leave room for a debt-disincentive mechanism in this case as well.

Altogether, the evidence reviewed above suggests that the poor performance of investment in countries with debt-servicing problems is generally consistent with the presence of debt overhang disincentives. The behavior of the consumption-output ratio in such countries, the contrast between the investment-output ratio in these countries and in countries without debt-servicing problems, and the sectoral evolution of investment between the public and private sectors are all broadly consistent with the presence of such disincentives. They are not fully consistent with explanations that emphasize the effects of the debt on national wealth in debtor countries, the need for trade balance adjustment, the reduced availability of external financing to market borrowers, and the “fiscal crunch” induced by public external debt. However, there is room for alternative explanations. In particular, the evidence examined above does not rule out the possibility that macroeconomic mismanagement and lack of confidence in government policies may have led to poor investment, low growth, and debt-servicing problems in certain highly indebted countries. Nevertheless, the crude tests presented above suggest that the “debt overhang” hypothesis cannot easily be dismissed on empirical grounds.

Accounting for the Growth Shortfall

Do debt disincentive effects provide an explanation for the growth slowdown that debt-problems countries have experienced since 1982? The experience of a number of countries suggests that high investment rates are not necessarily associated with high growth rates.6 Thus, even if the debt overhang contributed to the investment shortfall, other factors may have had a more important role in explaining the growth slowdown. In fact, some illustrative calculations suggest that the decline in investment accounts for less than half of the growth shortfall. Other possible explanations—in most cases complementary to debt-disincentive effects on investment—are described below.

The annual growth rate of real GDP for the 15 heavily indebted countries averaged 4.6 percent per year over the period 1975–81. before declining to 1.5 percent during 1982–87. The extent to which the drop of investment caused this decline in the growth of output can be gauged on the basis of the following procedure: Using gross investment figures for the highly indebted countries over the years 1965–81 and assuming a rate of depreciation of 5 percent, four alternative capital stock series were constructed for this group of countries corresponding to initial 1965 capital-output ratios of 2, 3, 4, and 5. A growth equation of the form:

y=α0eα+1K(1)α2

where y denotes real output, K the capital stock, and the αi’s are positive parameters, was then estimated for the group under alternative assumptions about the share of capital (α2 - 0.2, 0.3, and 0.4) for the period 1965–81, yielding three sets of estimates of the parameters α0 and α1, for each of the four capital stock series. The contribution of the investment shortfall to the growth slowdown was then calculated under each of these twelve pairs of estimates. The capital stock series for the years 1982–87 were projected by assuming (i) that the investment-output ratio was constant at 26 percent over those years, and (ii) that the ratio assumed its historical values. The levels of output for the years 1982–87 were calculated corresponding to the two alternative capital stock series using the estimates of α0 and α1 and the assumed values of α2. The difference between the average annual growth rates over the period 1982–87 of these two sets of output projections represents the contribution of the investment shortfall to the reduction of output growth.

The results of this exercise, reported in Table 16, imply that only a fraction of the growth shortfall can be attributed directly to the lower investment rates. Each cell in Table 16 contains the estimated contribution of the investment shortfall to the reduction in growth (measured in percentage points) under alternative assumptions about the initial capital-output ratio (identified along the top row) and the share of capital (identified in the first column). As indicated above, the estimated contribution of the investment shortfall depends on assumptions made about the share of capital and the initial capital-output ratio. The larger the share of capital, the larger is the effect of a given change in the capital stock on the growth of output; the smaller the capital-output ratio, the larger is the percentage change in the capital stock that results from a given change in the investment-output ratio. However, even in the most favorable case (i.e., that with the smallest initial capital-output ratio and largest share of capital), the investment shortfall can account for only about a third (1.1 percentage points) of the slowdown in growth.

Table 16.

Reductions in Growth Rates Accounted for by Investment Shortfalls in Heavily Indebted Countries, 1982–87

(In percent)

article image

If the behavior of investment accounts for at most one third of the growth shortfall, it is clear that other factors must have played a significant role:

a. Restrictive aggregate-demand policies in heavily indebted countries seeking to improve their trade balances may have resulted in prolonged underutilization of productive capacity in these countries. Thus, while the investment slowdown may have resulted in a reduction in the rate of growth of capacity, actual output may have grown even more slowly owing to deficient demand.

b. In the same vein, the drop in investment demand may have had a multiplier effect on aggregate demand and output, to the extent that capital goods are produced domestically. In this case, the drop in investment would affect output through the demand side of the economy, as opposed to the supply side effect which was discussed above.

c. Seeking to improve the trade balance, many countries have resorted to quantitative restrictions on imports and foreign exchange rationing. To the extent that such measures have restricted the availability of imported inputs, and distorted resource allocation, this would lead to an adverse supply shock, reducing output growth beyond what could be explained by slower capital accumulation.

d. To the extent that incentive-distorting policies have been pursued in the quest for trade-balance adjustment, overall economic efficiency will have suffered in indebted developing countries. As indicated above, poor policies may have contributed to excessive debt accumulation in some cases, as well as to the post-1981 investment slowdown. In addition to their effects on growth through slower capital accumulation, however, such policies would have adversely affected the growth of total factor productivity, resulting once again in a growth slowdown greater than can be explained by a reduced rate of capital accumulation.

e. Deterioration in the terms of trade suffered by most debtor countries may have contributed to the fall in households’ real income and in firms’ profitability. Both would tend to lead to declines in aggregate demand, in account of reductions in investment and consumption spending.

The role of these various factors in explaining the slowdown of growth in the heavily indebted countries remains a subject for future research. Most of the effects listed above are related at least in part to the need for external adjustment stemming from debt-servicing difficulties. In this sense, an improvement in the debt situation, coupled with appropriate macroeconomic policies and efficiency-enhancing microeconomic policies, may contribute much more to reversing the slowdown in developing-country output growth than can be accounted for by supply-side effects of investment alone. Obviously, the first requirement for achieving sustained growth is for each country to pursue prudent domestic policies that promote the full and efficient use of resources. The evidence discussed above suggests, however, that debt overhang problem may have made this task more difficult.

Supplementary Note 2 Strength of Fixed Investment in Major Industrial Countries

The surge in business fixed investment has been a key factor behind the unexpected strength of economic activity in the industrial countries in 1988. This note examines the reasons for the strength of investment, evaluates its characteristics, and discusses its sustainability.

Characteristics of the Strong Performance of Fixed Investment

Real business fixed investment in 1988 is estimated to have grown by over 5 percent a year in every Group of Seven country, and over 10 percent a year in Canada, the United States, Japan, and the United Kingdom. In no other year in the 1980s have the industrial countries experienced such a broadly based expansion in fixed investment. The size and pervasiveness of the boom make an investigation of its causes significant.

Table 17 shows ratios of real investment to real GNP for total fixed investment, public investment, residential construction, non-residential construction, and machinery and equipment. It indicates that the recent rise in investment has been concentrated in machinery and equipment: as a ratio to GNP, investment in this category is at high levels, both in comparison to the recent past and by historical standards. Comparing the first half of 1988 with the average for 1986, real investment in machinery and equipment has risen by at least half of a percent of output in every Group of Seven economy, accounting for the whole of the rise in the ratio of total fixed investment to output over this period in the United States and Italy, and for over half of the increase in the other Group of Seven economies. The machinery and equipment ratio is also at its highest level since 1970 in five of the Group of Seven countries. No other component of fixed investment shows similar strength. In particular, nonresidential construction has not been particularly strong, suggesting that the present investment surge may have more to do with replacement investment and capital deepening than with capital widening.

Table 18 indicates that nominal ratios of total fixed investment and of machinery and equipment investment to output are substantially lower than the corresponding constant price1 ratios in Table 17, and are not particularly high by historical standards, except in Japan. The higher real ratios are associated with declines in the relative price of certain high technology investment goods over the last few years, principally computing equipment (the ratio of real investment to real GNP equals the nominal ratio divided by the ratio of the investment deflator to the GNP deflator). However, Table 18 also shows that, even in nominal terms, investment/output ratios have increased recently, particularly for machinery and equipment.

The different behavior of investment/output ratios in current and constant prices complicates interpretation of the data. In the face of a large reduction in the relative price of certain capital goods, the use of base year prices that do not reflect the subsequent fall in goods prices tends to exaggerate the weight of these goods in a constant price index. Thus, such indices probably overstate the relative importance of investment in new technology, and hence the actual expansion in total investment. A more recent base year would give a lower weight to these goods, implying a less dramatic increase in real fixed investment, and a somewhat lower ratio to GNP. A second statistical problem relates to the difficulty of calculating an accurate price deflator in the face of rapid changes in the quality of a product. Such changes make it increasingly difficult to define the price of a homogeneous product over time. In addition, different national accounting systems may use different assumptions in calculating the volume series complicating comparisons across countries.

The recent surge in investment, while anticipated, was underestimated in magnitude, as indicated by data on 1988 investment intentions for the Group of Seven countries, which have generally risen over time (Table 19). The data fall into two types; for Canada, the United States, and Japan they are estimates of the planned amount of investment in 1988, while for the four European countries the numbers refer to the percentage balance of respondents who expect capacity to be more than sufficient in relation to production expectations. Given the widespread upward revision of investment forecasts and the poor forecasting performance of estimated investment equations, the following discussion of the factors behind the investment surge will be necessarily tentative.

Table 17.

Ratio of Various Categories of Real Fixed Investment to Real GNP/GDP, 1976-881

(In percent)

article image
Note: The Japanese data for machinery, equipment, and residential construction exclude public investment, whereas it is included for the other countries.Source: OECD Quarterly National Accounts, except the 1988 data I’m the Federal Republic of Germany, which are staff calculations.

The data are in constant prices: the base year is 1980 for Japan, France, the Federal Republic of Germany, and Italy, 1981 for Canada, 1982 for the United Slates, and 1985 for the United Kingdom.

Private sector.

Includes non-residential construction.

Table 18.

Ratio of Various Categories of Nominal Fixed Investment to Nominal GNP/GDP, 1976-88

(In percent)

article image
Source: OECD Quarterly National Accounts.

Private sector.

Includes non-residential construction.

1988 data seasonally adjusted by staff.

Table 19.

Investment Survey Results, 1987-88

article image
Sources: U.S. Department of Commerce Survey of Current Business, Bank of Japan Economic Statistics Monthly. Department of Regional Expansion Canada, and European Economy, Supplement B.

Plant and equipment.

Projections for fiscal year, which starts 1st April.

Data collected from selected large companies.

Explanations for the Strength of Fixed Investment in 1988

The current strength of fixed investment in the industrial countries appears to reflect a number of related factors: advances in technology, capacity constraints, high profitability, and the decline in oil prices.

Important technological advances have lowered the prices and improved the quality of certain capital goods in the last several years, particularly in the area of computing equipment.2 Reductions in the price of these goods relative to output have lowered the overall cost of capital to firms. which could be expected to stimulate investment. Moreover, the fall in the price of computers and other high technology equipment relative to other factors of production has probably induced some substitution toward these capital goods, and thereby increased investment in machinery and equipment. Other elements of the cost of capital—which depends upon the level of nominal interest rates, inflation expectations, and the effective rate of corporate taxation as well as the price of investment goods relative to output prices—do not appear to explain the strength of investment. Table A15 in the Statistical Appendix shows nominal short- and long-term interest rates for the Group of Seven countries: short rates rose in 1988, while long rates stayed broadly level. Long-term inflation expectations appear to have changed little in 1988; for example, survey data published by Drexel Burnham Lambert on the expected rate of inflation in the United States over the next decade actually fell slightly over the course of the year. Nor do changes in corporate taxation appear to have been important in reducing the cost of capital; indeed, the most major recent change, the U.S. Tax Reduction Act of 1986, is estimated to have raised the average cost of capital.3

For the United States, unpublished Commerce Department data indicate that real investment in high technology equipment4 contributed almost three quarters of the total rise in machinery and equipment investment between 1986 and the first half of 1988. Even taking into account problems related to the use of 1982 prices, it is clear that high technology equipment has played an important part in the investment boom in the United States.5 For the other Group of Seven countries, similar detail on investment patterns is not readily available. However, if the rise in investment includes a large amount of high technology items whose relative price has fallen, there should be a divergence in the behavior of the real and nominal investment/output ratios. A comparison of the last columns of Tables 17 and 18, which show changes in investment/output ratios over the recent past, indicates that such a divergence exists for the United States, Canada, and Japan: for the European countries, however, the effect is less pronounced.6 Investment in new technology appears to have had a large impact in North America and Japan, but may be a less important factor in Europe.

Table 20 shows that rates of capacity utilization in the manufacturing sector have been rising in all seven major industrial countries, approaching—and, in the case of the United Kingdom, exceeding-levels reached in the cyclical peaks of 1973 and 1979. Part of the recent surge in investment can therefore be seen as a cyclical response to the strength of demand. The early 1980s saw depressed levels of investment, as the capital stock adjusted to lower growth prospects.7 More recently, however, robust growth in demand has resulted in capacity bottlenecks, shifting the composition of demand toward investment, as in earlier cyclical peaks.8

Table 20.

Capacity Utilization Indices: Manufacturing, 1972-88

(In percent)

article image
Sources: OECD Main Economic Indicators. For Japan, Economic Statistics Monthly. Bank of Japan.

1985 = 100.

Seasonally unadjusted.

Percentage of firms operating at full capacity.

One important question is the degree to which the strength of fixed investment over the past year will ease the capacity constraints faced by industry. A recent Bank of Japan paper on the strength of corporate investment9 notes that in Japan investment usually affects the capacity utilization index with lag of about three quarters, indicating that the burst of investment in 1988 will start to affect the capacity figures soon. On the other hand, the data in Table 19 for the European countries indicate that the proportion of respondents who believed that new capacity was sufficient to accommodate their production plan was generally on the decline in 1988, suggesting that capacity constraints may not be reduced in the near future in those countries.

Another explanation for the current strength of investment is the current high level of corporate profits. The data on profits in manufacturing industry as a percentage of output in Table 21 show some shift in income toward capital over the past few years, particularly in Europe. While data for 1988 are not available, there are indications that this trend has continued. These high profits represent an increase in the return on productive assets.10 Since investment depends upon a comparison of the expected rate of return on assets with the cost of capital, what matters is whether high levels of profitability are expected to continue. Here, the evidence is sketchy. Share prices, which can be interpreted as the stock markets’ assessment of the present value of the future stream of dividends, are an indicator of the expected future level of company profits. Since the October 1987 stock market crash, the performance of share prices has been robust in Japan, France, and Italy, but sluggish elsewhere.

Table 21.

Profit Shares in Manufacturing Output, 1972-871

(In percent)

article image
Sources: U.S. Bureau of Labor Statistics.

Non-labor income of manufacturing companies as a percentage of gross nominal output.

The fall in oil prices in 1986 could have affected investment in two ways. First, it may have boosted domestic demand through improvements in the terms of trade, thereby stimulating investment through the high rates of capacity utilization discussed above. The second effect comes through the supply side of the economy. The long-run effect of this supply response on the capital stock is uncertain, depending upon the type of technology, and in particular on whether capital, labor, and intermediate inputs are substitutes or complements in production. However, the short-run effects are less ambiguous. If the fall in oil prices does not induce a rise in real wages, its primary effect is to lower industrial costs and hence increase profitability, data for which are discussed above. In the short run, with the capital stock fixed, a fall in costs will induce producers to raise output. The positive effects on demands for labor and capital from this rise in output will probably dominate any substitution effects caused by the change in relative prices, hence the actual (and expected) rate of return on capital will rise, stimulating investment.

Though it has been difficult to establish an empirical link between changes in oil prices and large and sustained trends in productivity, such as the slowdown in the 1970s, these price changes could help to explain the smaller changes in activity over the last few years. However, the major drop in oil prices occurred in early 1986. While some lag is to be expected, particularly in view of the stock market crash of 1987, it is not clear why the beneficial effects on investment from the fall in oil prices should have become such a factor in 1988.

Another factor that is impossible to quantify, but that may be important in explaining the strength of investment is business confidence, which may have been boosted by a number of favorable factors over the past year. These include the robust performance of the world economy in the wake of the stock market crash, the relative stability of exchange rates over 1988, and structural policies taken by industrial countries.11

Conclusions

Fixed investment has been unexpectedly strong in the past year throughout the industrial countries, primarily owing to substantial increases in investment in machinery and equipment. The strength of this sector, compared to nonresidential construction, points to a significant role for replacement investment and capital deepening, rather than capital widening, in the present boom. Apart from confidence effects, which may well have had an impact but are difficult to measure, major factors appear to have been investment in new technology, particularly in North America and Japan, and capacity constraints stemming from the robust performance of demand. Supply effects from the fall in oil prices and increased profitability of industry, particularly in Europe, may have also played a role.

An important issue is the effect of the recent rise in investment on industrial capacity; consideration of this question is complicated by a number of measurement problems. Traditional straight-line depreciation calculations may not capture some features of the present situation adequately, the use of constant price data may overstate the strength of the real investment, and rapid technological advances in computing equipment may have speeded up technological obsolescence. As a result, measured depreciation rates may be too low. Despite these caveats, the size of the investment surge points to a significant increase in productive capacity.

Looking to the future, an assessment of the sustainability of the investment boom is difficult to make. Nominal investment/output ratios are not particularly high, and survey data for Europe indicate that businesses foresee the need for continued expansion in capacity. However, household saving rates are low in many countries, oil prices have firmed since their low point in 1986, short-term interest rates are rising, and prospects for technological advances in computing equipment are uncertain. Overall, it seems unlikely that investment will contribute as much to the growth in real demand in the near future as it did in 1988.

Supplementary Note 3 Public Debt and Fiscal Policy in Major Industrial Countries, 1977-87

Public debt expanded rapidly during the last decade in most industrial countries. Unlike the increases experienced during the great depression and the two world wars, the recent accumulation of public debt relative to output took place in a peacetime environment—albeit characterized by supply shocks and a recession. This note examines the factors that contributed to the aggravation of the debt burden in the Group of Seven countries. It concludes that despite comparable trends in the debt-to-GNP ratio, the relative contributions of various components and the policy responses differed widely among industrial economies.

Major Trends

In the 1980s both gross and net debt ratios increased for practically all major industrial countries and followed broadly similar patterns (Table 22).1 The net debt ratio increased most rapidly in Italy and Canada. In Italy, spurred on by large budget deficits, the ratio moved ahead by more than 35 percentage points during 1977-87 (to 81 percent of GDP) for the general government and almost tripled (to 77 percent of GDP) for the central government. In Canada the ratio grew rapidly in the 1981-82 recession and its aftermath before moderating somewhat in the last few years. This growth reflected a deterioration in the financial position of the federal government, which only recently has improved; the growth in net debt amounted to about 30 percentage points during 1977-87, to a ratio of 39 percent of GNP at the end of the period.

In France, the Federal Republic of Germany, Japan, and the United States the net debt ratio rose by an average of 12-15 percentage points at the general government level. In Japan and the Federal Republic of Germany the increase occurred despite a steady correction of fiscal imbalances during most of the period under consideration. In Japan, the general government budget balance as a ratio to GNP improved from a deficit of 5½ percent to a surplus of ½ percent, thus contributing to a stabilization of and a subsequent reduction in the ratio of the net debt to GNP. In both countries, and in the United Kingdom, which also achieved a steady reduction in its fiscal deficit, net debt ratios were in fact lower in 1987 than in 1984. By contrast, the accumulation of debt in France and the United States resulted from a significant deterioration in fiscal position from virtual balance in the late 1970s to a deficit of around 3 percent of GNP at the general government level in the first half of the 1980s. Even wider deficits were incurred at the central government level in the United States during the latter period.

Determinants of Growth in Debt Burden

The growth of the public debt ratio can be decomposed within an accounting framework based on the budget-financing identity. Specifically, the budget deficit in period t can be expressed as

bt=gt+itdt1τt=Δdt+(nt+πt)dt1,(1)

where b is the budget deficit, g and τ are noninterest government expenditure and revenue, respectively, and d is the stock of net government debt outstanding (including claims by the monetary authorities), all in terms of ratios to GNP; in addition, n denotes the rate of growth of real GNP, π is the rate of inflation, and i is the nominal rate of interest. Rearranging terms and solving for Δd.

Table 22.

Major Industrial Countries: Debt Outstanding at Central and Government Levels, 1977–871

(In percent of GNP/GDP)

article image
Source: Fund staff estimates based on the following national publications: United States: Board of Governors of the Federal Reserve System. Flow of Fund Accounts Financial Assets and Liabilities, Year-End, 1963–86, and Federal Reserve Bulletin (various issues); Japan: Economic Planning Agency, Government of Japan, Annual Report on National Accounts (various issues); Federal Republic of Germany: Statistical Supplements to the Monthly Reports of the Deutsche Bundesbank (May issue); France: Institut de Prévisions Economiques et Financières pour le Developpement de Entreprises, Revue de l’Ipecode, No. 11 (March 1986); Italy: Assemblea Generale Ordinaria Dei Participant, Banca d’Italia; Canada: The National Balance Sheet Accounts 1961-84, Statistics Canada (June 1985), and The Fiscal Plan, Department of Finance (February 1988).

Book value of debt outstanding at the end of the year.

Preliminary staff estimates, subject to future revision.

For the Federal Republic of Germany, central government net debt series are not available; in the decomposition in Table 24, the gross debt series has been used as a proxy for the net debt series, since its relations with the annual deficit series suggest it is a good proxy for the net debt series.

Estimated by adding the fiscal deficit to the corresponding stock of net debt in the previous year.

Δdt=gtτt+(itntπt)dt1,(2)

which shows the change in the debt ratio as determined by the primary (noninterest) budget deficit, interest payments, economic growth, and inflation. In turn, the primary deficit can be broken down into cyclical and noncyclical components2

gtτt=(τ0ytpτ0)(τ0g0)ytp+ft,(3)

where g0 and τ0 are ratios of noninterest expenditure and revenue to GNP in the base period, yp is the ratio of potential to actual GNP and/is a measure of the fiscal stance. The first term in parentheses represents the cyclical component, the second is the structural component (measured in relation to the base year), and the third component represents the fiscal stance.

The cyclical component of the budget deficit reflects variations in revenue resulting from deviations of actual output from its potential level. The structural component is defined in terms of the cyclically adjusted primary deficit in the base year; the fiscal stance reflects the effect of both automatic stabilizers and discretionary policy changes since that year.3 The fiscal stance as defined here is deemed expansionary (contractionary) relative to the base year if the actual primary deficit exceeds (falls short of) the cyclically adjusted deficit, where the base year is a period in which actual output and potential output are assumed to be equivalent, and the fiscal stance is neutral. The fiscal stance, in this sense, is the magnitude of the stimulus injected (withdrawn) by the budget into (from) aggregate demand above and beyond the primary deficit implied by the structure of budgetary operations in the base period.4 After substituting equation (3) into (2). changes in the public debt ratio may be expressed as

Δdt=(τ0ytpτ0)+(g0τ0)ytp+ft+itdt1(nt+πt)dt1,(4)

The change in the debt-to-GNP ratio can thus be seen as the sum of four components: a cyclical effect, a noninterest structural effect, which reflects the extent to which the initial fiscal imbalance is magnified by the growth of potential GNP, the change in the fiscal stance measured during the period, and the extent to which the difference between the interest rate and the growth of nominal GNP tends to raise (or diminish) the initial debt ratio. In essence, except for the cyclical effect, all the other components reflect, directly or indirectly, the adequacy of the fiscal stance.

As can be seen in the first term of equation (4), the cyclical component, as defined in this note, is the cumulative contributions of government revenue shortfalls resulting from the level of aggregate output falling below that of potential output. One could argue that the contribution of the cyclical factor to the growth of debt should be positive during recession years, but negative during boom years, and thus it should be negligible over a long period. In this respect, the methodology used in this note, which is critically dependent on the measure of potential output rather than concepts such as a moving average of output, will tend to overestimate the contribution of the cyclical factors to the growth of debt. In particular, the concept of potential output used here is not equal to trend output, but is rather a measure of full capacity output; as a result, output will usually be below potential.5

Table 23 and Chart 27 show the decomposition of the public debt burden accumulated over the period 1977–87, on the basis of equation (4).6 The results indicate that the contribution of various factors to the growth of the net debt ratio differed significantly among countries. Except for the United States and Canada at the general government level, the noninterest structural component and, to a lesser extent the cyclical component, contributed to the growth in the debt ratio in all countries. This growth was only partially offset through a corrective fiscal stance: since 1982, only in the United Kingdom and Japan did the debt ratio actually fall for the general government. The budget deficits in prior years led in turn to a roughly proportional additional growth in the debt ratio through a sharp rise in interest payments. The latter more than offset the influence of output growth and inflation, except in Italy and the United States for the general government, where, on average. GNP growth exceeded the effective interest rate on net debt (Table 24).

Table 23.

Major Industrial Countries: Decomposition of Changes in the Net Public Debt Ratio, 1977–87

(In percent of GNP/GDP)

article image
Source: Fund staff estimates.

Attributable to measurement differences owing to factors such as valuation changes and timing.

Budget deficit at the central government level includes disbursements of the Fiscal Investment and Loan program, except those to financial institutions. Thus, the noninterest structural component has been derived as a residual to make the components of deficit consistent with the increase in debt reported on a national accounts basis.

Debt data for the United Kingdom at both levels of government, and for the United States at the general government level, are available on a cash basis. Because of the potential noncomparability with the deficits reported on a national accounts basis the noninterest structural components have been derived residually.

Apparently in the United States the cyclical component—driven mainly by the 1981-82 recession and not sufficiently offset by the strong subsequent expansion—accounted for a large increase in the debt burden for the general government and some one third of the debt of the central government. The noninterest structural component was relatively small or even negative, given a small primary central government deficit and a primary general government surplus in the base year. Despite a built-in tendency for the methodology used in this note to yield an underestimation of the fiscal policy stance component, the results indicate that fiscal policy contributed substantially to the growth of the general government debt. As indicated earlier, since the decomposition between the cyclical and fiscal stance components is highly sensitive to the potential GNP path, these results should be interpreted with caution. Indeed, a small adjustment in the assumed level or path of potential GNP would affect substantially the cyclical and fiscal stance components. In particular, a lower level of potential output relative to actual GNP would raise the contribution of an expansionary fiscal stance to the rising U.S. debt ratio.

In Japan the main source of growth in the debt ratio was a comparatively large noninterest structural component, although this was partially offset by an uninterrupted contractionary fiscal stance (Chart 27). Because the Japanese economy was growing at a rate close to its potential, the cyclical component was marginal. By contrast, in the Federal Republic of Germany the cyclical component increased during the recession and was not fully reversed in the following years as economic growth remained below potential. High interest rates relative to nominal GNP growth contributed significantly to the buildup in the debt burden. Unlike in Japan and the United States, however, the cumulative contribution of the noninterest structural component was broadly cancelled by the fiscal stance.

Compared with most other countries, France’s debt ratio increased on the basis of primary deficits in the early 1980s and because interest payments exceeded output growth thereafter. However, the cyclical component was important at the general government level where the tax base is twice as large as at the central government level. Meanwhile, the noninterest structural component for the general government was more than offset by a fiscal stance that was on balance contractionary. In the United Kingdom the debt ratio declined at the general government level and rose slightly at the central government level, with corrective action more than offsetting a sharp increase in the cyclical and net debt-service components.

Chart 27.
Chart 27.

Major Industrial Countries: Cumulative Changes in Public Debt Disaggregated into Major Components, 1978–87

(In percent of GNP/GDP)

1 The sum of the cyclical and noninterest structural components; the unexplained residual is also included in this category.
Table 24.

Major Industrial Countries: Average Nominal GNP Growth Rates and Interest Rates, 1977–87

(In percent a year)

article image
Source: Fund staff estimates based on national publications.

Calculated as the interest rate on gross debt adjusted for the differential interest rate paid on assets held at the general government level. The interest differential is weighted by the ratio of assets to liabilities.

In Italy the marked growth in the debt burden was attributable to the noninterest structural component—the largest among Group of Seven countries. This was offset in part by the debt-service component, net of output growth effects, as the effective interest rate on government debt was lower than the growth of nominal GNP (Table 24), reflecting the extensive use of captive financing at a negative average real interest rate during much of the period under consideration.

In Canada the debt-service component exceeded the effect of nominal GNP growth by a large margin and was the dominant factor behind the rapid increase in debt burden. This was partly because government assets earned substantially below-market financial returns, as against relatively high interest rates paid on liabilities, resulting in an exceptionally high average interest rate on net debt. The noninterest structural component contributed to the accumulation of central government debt, while exercising a dampening influence on the general government debt.

Fiscal Policy Response

The above calculations suggest that the corrective stance of fiscal policy for the most part was insufficient to counteract contributions of cyclical, structural, and net interest components to the growth of public debt in major industrial countries. However, the measure of fiscal stance used in this note cannot be expected to capture a systematic policy response to the debt buildup. To determine whether fiscal policy has indeed attempted to counter the rise in the debt burden, a simple hypothesis was tested in regard to the government’s reaction to the growth of the debt burden. In the test, budget deficits (adjusted for cyclical variations) were regressed on debt ratios. The results suggest that major industrial countries responded differently to the growing burden of public debt during the period 1977-87. Moreover, estimates for subperiods characterized by an identifiable and significant change in government or announced reorientation in fiscal strategy indicate that most countries have adopted increasingly active fiscal rules in recent years to resist further buildup of debt.

Supplementary Note 4 Monetary Policy in Major Industrial Countries

In recent years, monetary policy in the major industrial countries has faced the difficult task of balancing two separate sets of objectives. First, the monetary authorities have remained responsible for maintaining conditions conducive to a noninflationary expansion of domestic economic activity. Second, they have played an active role in promoting the exchange rate objectives established under the international cooperative process.

The task of balancing internal and external objectives is by no means a new one for central banks. Because the instruments of monetary policy are more flexible and can be adjusted more promptly than those of fiscal policy, monetary authorities have been responsible for a major share of the short-term stabilization measures adopted over the years, in response to both internal and external pressures. Nevertheless, the renewed emphasis on exchange rate objectives, in the face of large current account imbalances and uncertainties about fiscal prospects in some countries, has placed considerable strain on monetary policy.

Judged on the basis of recent macroeconomic performance, monetary authorities have met the challenge well. The expansion of employment and economic activity has been fairly strong over the past year, and where strains on labor markets and capacity have started to emerge, financial market conditions have been tightened to slow the expansion. Even so, the outlook for inflation has become a greater concern in a number of the major countries, especially Canada, the United Kingdom, and the United States.

In the past, the rate of monetary growth has often been relied upon as a principal guide for the conduct of policy, and most of the major industrial countries established target growth rates for their monetary aggregates. In recent years, the increased emphasis on exchange rate objectives, together with the increased difficulty of interpreting the monetary aggregates in an environment of financial innovation and deregulation, has led some countries to de-emphasize their monetary targets and other countries to tolerate increased variability of monetary growth.

For the largest countries, the deviations of monetary growth from the paces sought by the authorities were somewhat smaller in 1988 than in 1987 (Table 25). In 1987, the U.S. Federal Open Market Committee had allowed a sharp slowdown in the growth of its broad money aggregates to rates below the M2 target range and at the lower bound of the M3 target range. By contrast, the German Bundesbank had allowed its money target to be exceeded by 2 percentage points, and the Bank of Japan, which has never adopted a formal monetary target, had tolerated double-digit monetary growth. In 1988, both of the targeted U.S. aggregates were held within their designated ranges, while the Bundesbank’s M3 target was exceeded by less than I percentage point. In Japan, the money stock continued to expand rapidly in 1988, although at a somewhat slower pace than in 1987. Both the United States and Germany have announced monetary targets for 1989. The Bundesbank has continued to express its monetary aggregates objectives in terms of the broad money stock M3 (after changing over from a central bank money target at the beginning of 1988) and has established a growth target of about 5 percent for 1989. The U.S. authorities have specified objectives of 3 percent to 7 percent for M2 growth during 1989 and 3½ percent to 7½ percent for M3.

Among the other major industrial countries. France and Italy held the growth rates of their targeted M2 aggregates within the pre-specified policy ranges for 1988 (although Italy’s target growth rate for total credit to the nonstate sector was substantially overshot); both countries have announced unchanged monetary targets for 1989. By contrast, in the United Kingdom the 12-month growth rate of MO was allowed to run several percentage points above its target during most of fiscal year 1988-89: the target for 1989-90 has been left unchanged. In Canada, the practice of targeting Ml was abandoned in 1982 after an increasing instability was experienced in the relationship of that aggregate to nominal income and interest rates. More recently. M2 has played a role as an indicative policy guide. The growth of M2 quickened from about 7½ percent during 1987 to 10 percent during 1988. despite steps taken by the Bank of Canada toward restraint, although at least I percentage point of this increase reflected the pattern of sales of Canada Savings Bonds.

Table 25.

Major Industrial Countries: Targets and Outcomes for Monetary Growth, 1987–1989

(In percentage change from fourth quarter to fourth quarter)

article image

Percentage change from October–November to October–November.

Percentage change from December to December.

In response to developments in the economy and to signals from both the monetary aggregates and exchange rates, monetary authorities in a number of the major industrial countries took actions to raise short-term nominal interest rates substantially during 1988 and early 1989. The increases were most pronounced in the United Kingdom (approximately 4¼ percentage points, end-December 1987 to end-February 1989) and Canada (more than 3 percentage points). In addition, the authorities in Germany and the United States acted on a number of occasions to tighten financial market conditions.

While short-term interest rates rose substantially in a number of the major industrial countries, long-term rates remained remarkably stable. At the end of February 1989, long-term interest rates were significantly lower than short-term rates in Canada and the United Kingdom, while in France, Germany, and the United States the differentials between long and short rates were considerably less than they had been at the end of 1987. One possible interpretation of these developments is that the authorities in those countries where monetary conditions were tightened substantially acted at an early stage to maintain confidence and prevent a worsening of inflationary expectations, hence establishing the basis for monetary conditions to be held relatively steady in the future (or eased in countries where yield curves became inverted). This interpretation would seem consistent with the degree of business confidence reflected in the strong performance of investment last year.

Of course, the expectations about future levels of interest rates that appear to be reflected in market yield curves can only be interpreted, broadly speaking, as central points of probability distributions. Accordingly, if market participants in fact feel quite uncertain about the interest rate outlook and attach a significant probability to the prospect that monetary policy will not succeed in preventing an upsurge in inflation, they must also attach a significant probability to the opposite outcome—namely, that fiscal adjustment, or perhaps a relatively sharp economic slowdown or contraction, will lead to lower-than-expected levels of interest rates. Thus, a second possible interpretation of the term structure of yields is that market participants are not confident that monetary policy will succeed in prolonging the noninflationary expansion, but that they consider the possibility of downward pressure on interest rates at least as significant as the risks of upward pressure. Uncertainty about how forcefully the new Administration in the United States will act on the budget deficit, together with memories of both past inflationary episodes on the one hand and the October 1987 stock market crash on the other hand, suggest that this interpretation is plausible.

The difficulties of prolonging the noninflationary expansion should not be underestimated. Success hinges critically on whether the authorities can continue to correctly interpret and anticipate the strength of economic activity and inflationary pressures. Thus, from one perspective, the noninflationary expansion is threatened by the difficulties of assessing and forecasting macroeconomic developments in the context of large internal and external imbalances.

From a different perspective, macroeconomic prospects depend importantly on whether the private sector behaves in a relatively stable and predictable pattern. This consideration suggests the desirability of maintaining consistent and clearly articulated objectives for monetary policy, and perhaps of avoiding certain types of fiscal policy adjustments. Surprises and uncertainty about the interest rate levels that monetary authorities seek to establish can significantly affect private sector spending patterns in ways that the authorities themselves may initially find difficult to interpret—for example, via changes in savings rates or through an acceleration or postponement of spending. Similarly, some types of major adjustments in tax rates and investment incentives—particularly when they become effective with a lag—can also have significant effects on the intertemporal pattern of private spending. In turn, changes in private sector spending patterns, when difficult to interpret, can introduce significant errors into the authorities’ assessments and forecasts of the strength of the economy. Such errors carry the risk that inflationary or recessionary forces may take hold before the errors are clearly perceived, and that abrupt “stop” or “go” changes in policy may subsequently be required.

It can be conjectured, in this context, that the relatively credible and coherent behavior of monetary policy in recent years has contributed importantly to the prolonged noninflationary expansion in the major industrial countries. While the authorities have given less emphasis to “rules” for monetary aggregates, they have also devoted considerable effort to articulating their near-term objectives and priorities, to explaining how they would react to various changes in macroeconomic conditions, and to adjusting monetary conditions promptly and incrementally rather than with excessive caution at first and then abruptly.

In addition to depending on accurate assessments of the underlying strength of economic activity and inflationary pressures, success in continuing to keep monetary policy on a relatively predictable and steady course in the period ahead will hinge critically on avoiding major conflicts between policy objectives. These include conflicts between the policy objectives of different countries, between the objectives of the monetary and fiscal authorities within individual countries, and between the different objectives that individual monetary authorities may attempt to pursue. Policy conflicts of any kind are potentially a major source of market uncertainty and abrupt shifts in market expectations. Private sector responses to perceptions of conflicting policy objectives can generate instability either directly through changes in private sector positions in financial markets or indirectly by adding “noise” to the patterns of aggregate demand and supply that the authorities must interpret and forecast with reasonable accuracy if stability is to be maintained.

These considerations suggest that the use of monetary policy to promote exchange rate objectives may be constructive at times and counterproductive at other times. Exchange rate objectives, like targets for the monetary aggregates, may be useful guides to the extent that they give rise to macroeconomic outcomes that are desirable from both a national and an international perspective. But unless fiscal policy can be adapted along with monetary policy, exchange rate objectives that become inconsistent with other national policy objectives can be counterproductive. It can be argued, for example, that the failure or inability of the major industrial countries to rely more heavily on fiscal policy for resisting mounting inflationary pressures in the summer of 1987, in the context of the common exchange rate objectives established under the international cooperative process, resulted in a ratcheting upward of interest rates that may have contributed to the October 1987 stock market crash. More generally, when monetary policy must aim at achieving both internal and external balance without much help from fiscal policy, conflicts between the different objectives seem bound to arise from time to time under any monetary strategy.

To some extent, the monetary authorities have been able to cope with the various difficulties that have developed, and with changes in the relative strength of different policy concerns, by modifying their strategies over time. For example, after money demand instability led the Canadian authorities to abandon the practice of monetary targeting in the early 1980s, the Bank of Canada followed a strategy that placed considerable weight on avoiding sharp depreciations of the exchange rate as a possible source of inflation during a period of substantial slack in the economy. Subsequently, after the monetary aggregates had become more predictable, and after economic expansion had raised concern over the possibility of inflation arising from excessive demand pressures, the Canadian authorities again began to rely more heavily on monetary aggregates as indicative policy guides.

Looking ahead, it can be argued that if fiscal policy remains relatively nonadaptable. the prospects for prolonging the noninflationary expansion will be weakened should monetary policy become locked into a strategy that can be changed only after a crisis develops. Greater adaptability of fiscal policy is clearly desirable. But in the absence of substantial progress in this direction, macroeconomic stability may continue to depend crucially on the adaptability of monetary policy strategies.

Supplementary Note 5 Capital Account Developments in Japan and the Federal Republic of Germany: Institutional Influences and Structural Changes

This note provides a brief description of the capital accounts of the two major surplus countries—Japan and the Federal Republic of Germany. Particular attention is given to the private capital flows that in recent years have been the main counterparts to the current account surpluses of these countries and to whether these flows can be associated with capital inflows into the United States. In understanding the capital account patterns of Japan and Germany, emphasis is given to the institutional structures of their financial systems and to certain fundamental forces affecting them.

Capital Flow Structure in Japan and Germany

Both Japan and the Federal Republic of Germany have seen large increases in net private capital outflows over the last several years (Table 26). However, a breakdown of these flows, by maturity,1 type of instrument, and sector shows a number of significant differences. In Japan, increases in long-term portfolio investment outflows are predominant, although other forms of long-term capital outflows—e.g., direct investment—also have increased steadily.2 Moreover, large inflows of short-term capital have occurred, especially through banks. In Germany, overall inflows on long-term portfolio investment were registered from 1983 to 1987 owing to large purchases of German securities by foreigners. As a result, total long-term outflows were small in most years.3 Short-term capital flows have been negative (outflows), in most years. These differences between Japan and Germany reflect differences in structure of their financial sectors and in behavior of investors, as discussed below.

The increases in net private outflows from Japan and Germany have been associated with the large current account deficits of the United States. It is difficult to determine from available data to what extent U.S. deficits are ultimately financed by capital flows from Japan and Germany. It would appear, however, that the channels from Japan to the United States are more direct than those from Germany to the United States, at least with respect to long-term flows (Table 27).4

Institutional Arrangements in Japan and Germany

The institutional arrangements governing financial markets are important to an understanding of capital outflows from Japan and Germany. Characteristics of these arrangements include the relative sizes of financial intermediaries, prudential regulations on foreign currency positions, the depth of short-term money markets, the degree of integration with offshore centers, and the degree of segmentation of financial activities.

One important similarity between Japan and Germany (and indeed among all major industrial countries) is the existence of prudential guidelines on the foreign exchange positions of banks aimed at preventing currency fluctuations from compromising the stability of the banking system. In addition, both countries have long-term bond markets that are deeper and more flexible than short-term money markets. The similarities appear to have led both countries to eliminate most foreign exchange risk from the banking system and to channel such risk predominantly through markets for long-term instruments.

Table 26.

Japan and the Federal Republic of Germany: Private Capital Flows. 1980–88

(In billions of U.S. dollars)

article image
Sources: Bank of Japan, Balance of Payments Monthly. and Bundesbank, Monthly Report. Supplement 3.

Including errors and omissions.

Figures convened to dollars at average exchange rates.

Table 27.

Selected Regional Payment Flows, 1985–87

(In billions of U.S. dollars)

article image
Sources: Bank of Japan. Balance of Payments Monthly, and Bundesbank. Monthly Bulletin. Supplement 3.

Converted to U.S. dollars at average exchange rates.

Japan and Germany have differed in terms of the degrees of integration between domestic and offshore financial centers, with Japan having been less integrated and having a shorter tradition of use of offshore centers. Although technology has reduced the importance of this difference, it has not disappeared altogether. In addition, the two countries differ in terms of the legal segmentation of financial activities: Japan largely separates banking, securities, and trust operations among independent institutions while Germany allows most operations within a single institution. As a result, banks, which are regulated not to assume foreign exchange risk, play a larger role as financial intermediaries in Germany than in Japan. These two differences appear to have led to Japan’s having a relatively larger proportion of exchange risk borne by residents. A more detailed examination of the financial system in the two countries may help to explain how these institutional differences have led to different capital account structures.

Japan

Most intermediaries in Japan have increased the share of foreign securities in their portfolios in recent years (Table 28).5 Yet there are important differences among groups of investors, related chiefly to whether institutions are involved in maturity transformation of external funds or currency transformation of domestic funds.

Commercial banks in Japan have been involved chiefly in maturity transformation of external funds. This is reflected in the form of large short-term external borrowings and purchases of foreign securities by banks. That is, Japanese banks have borrowed short-term funds overseas in foreign currencies and have invested funds in long-term instruments overseas in foreign currencies. This behavior reflects a combination of two factors: prudential regulations that limit banks” net foreign exchange exposure6 and the relatively limited availability of attractive yen-denominated foreign investments.7

In contrast to banks, a number of Japanese institutions are highly active in currency transformation, that is, holding foreign, dollar-denominated assets to back domestic, yen-denominated liabilities. The most important of these institutions are the trust banks, which act as fiduciaries for pensions and other types of trust accounts, and the life insurance companies, neither of which faces regulations on matching foreign currency assets and liabilities. By their nature trust banks and insurance companies have traditionally matched a long-term liability structure with a long-term asset structure, with relatively less attention given to short-term exchange rate prospects.8 Moreover, accounting rules that favor high-coupon investment have made foreign securities investment more attractive, given the higher coupons available on foreign securities in recent years. Insurance companies have been affected by liability risks that are in some ways peculiar to Japan, especially the risk of earthquakes in major metropolitan centers. Thus companies have an incentive to diversify into foreign currency assets that would allow them to sustain operations if such domestic risks were to materialize. In the public sector, an active investor in foreign assets in recent years has been the Postal Life Insurance System, which also has a long maturity structure for liabilities. Other institutional investors with increasing foreign investments include the securities investment trusts, which are mutual funds usually administered by the subsidiaries of securities companies, and the thrift institutions, acting either directly or through central organizations.

The Federal Republic of Germany

In Germany, in contrast to Japan, banks appear to have been the major vehicle for recycling German current account surpluses, but they have acted indirectly through offshore banks (Table 29). Lending to other banks as a proportion of the assets of German banks has grown steadily through the 1980s; much of this growth has been accounted for by lending to foreign banks, which has risen from 14 percent of all lending to banks in 1981 to 22 percent in 1988. Investments in nonbank foreign securities, however, have declined as a portion of total investment in securities. On the liability side, no major increase of foreign exposure has occurred; indeed, deposits from foreign banks have declined slightly as a proportion of interbank deposit liabilities. Behind these developments lie currency exposure regulations that are stricter than those in Japan. In Germany, the regulations prohibit net foreign currency positions from exceeding certain limits not only across maturities but also in instruments with maturities of one month or less and in instruments with maturities of six months or less. Thus, recycling of current surpluses in German banks occurs in a way that expatriates not only much of the currency-transformation activity but also much of foreign-currency maturity-transformation activity.

Table 28.

Japan: Investment in Foreign Securities by Banks and Institutional Investors. 1981–88

(At yearend, in percent)

article image
Sources: Bank of Japan, Economic Statistics Monthly, and Economic Statistics Annual, various issues.

Including banking accounts of trust banks.

End of fiscal year.

Excluding banking accounts of trust banks and securities investment trusts.

Table 29.

Federal Republic of Germany: Selected Ratios from Balance Sheets of All Banks. 1981–88

(Average of month-end data, in percent)

article image
Source: Bundesbank, Monthly Bulletin. Statistical Tables III.2 and III.3.