In addition to the effects of exchange rate variability on the aggregate volume and pattern of international trade, which have been reviewed in the preceding section, fluctuations in exchange rates can have a potential effect on the structure of output and investment in the domestic economy.
Adjustment
Exchange rate variations may lead to shifts in the allocation of resources as producers gain and lose international competitiveness. To the extent that such variations are temporary and reversible, there may be adjustment costs, since resource shifts often involve temporary periods of unemployment, retraining of manpower, adaptation of plant and equipment, etc. (see Kreinin and Heller, 1974). It may be, however, that producers do not react to exchange rate shifts until they are viewed as being “permanent,” so that the resource reallocation that takes place as a result of short-run exchange rate variability is actually quite limited (McKinnon, 1978). The long and distributed lags found in most studies of the effect of relative prices on trade flows suggest that sudden and substantial resource shifts are unlikely in the face of exchange rate movements that are reversed within a short time. The consequence of such short-run movements is more likely to be apparent in the time path of company profits (which may, of course, be regarded as undesirable).
For exchange rate movements of longer duration, it is plausible that trade flows will be affected. The questions we therefore wish to ask are:
(i) Is the volume of resource shifts related to the amplitude of exchange rate movements?
(ii) Has the volume of these shifts been excessive in terms of the operation of the international adjustment process?
The addition of the second question is important, since resource shifts into and out of the foreign sector should be viewed differently when they are responding to an adjustment need (say, a permanent shift in the price of oil) than when they result from temporary and reversible factors that bring about a short-term exchange rate movement.
There appears to be relatively little empirical evidence on this question. Thursby (1981) found no evidence that the adjustment costs faced by Canada were greater under flexible rates than under fixed rates. Adjustment costs in her model were proxied by the volume of resources shifted into and out of the export sector as a result of exchange rate changes. In a later extension of the same model to 19 other industrial countries (Thursby and Thursby, 1981), a generally similar result was found, though in the case of 2 countries in the group, there did appear to be significant costs. However, this test is not quite the same as asking whether exchange rate variability per se (i.e., whatever the exchange rate regime) has had undesirably large adjustment effects.
It does appear that exchange rate variability has been accompanied by greater shifts of resources into and out of the foreign sector. Table 1 presents two alternative measures of the extent of these shifts for the seven major industrial countries during the periods 1960–70 and 1974–82. The first two columns of the table show the variance of the real net foreign balance as a percentage of GDP, while in the last two columns average year-to-year changes in the net foreign balance as a percentage of GDP are set out. Because they comprise variations in both export and import shares, movements in the net foreign balance (in real terms) are a comprehensive indicator of foreign sector resource shifts.
Major Industrial Countries: Fluctuations in the Real Net Foreign Balance as a Proportion of Gross Domestic Product1
(In percent)
Data marked by asterisks are significantly different at the 95 percent confidence level.
Absolute change from preceding year.
Major Industrial Countries: Fluctuations in the Real Net Foreign Balance as a Proportion of Gross Domestic Product1
(In percent)
| Annual Variance |
Average Annual Change2 |
|||
|---|---|---|---|---|
| 1960–70 | 1974–82 | 1961–70 | 1974–82 | |
| Canada | 1.06* | 3.89* | 0.8 | 1.3 |
| United States | 0.43 | 0.47 | 0.3* | 0.8* |
| Japan | 0.39* | 7.05* | 0.8* | 1.5* |
| France | 0.77 | 0.93 | 0.6* | 1.1* |
| Germany, Fed. Rep. of | 0.90 | 1.78 | 0.9 | 1.2 |
| Italy | 1.90 | 2.85 | 0.9 | 1.2 |
| United Kingdom | 0.48* | 1.80* | 0.7* | 1.3* |
Data marked by asterisks are significantly different at the 95 percent confidence level.
Absolute change from preceding year.
Major Industrial Countries: Fluctuations in the Real Net Foreign Balance as a Proportion of Gross Domestic Product1
(In percent)
| Annual Variance |
Average Annual Change2 |
|||
|---|---|---|---|---|
| 1960–70 | 1974–82 | 1961–70 | 1974–82 | |
| Canada | 1.06* | 3.89* | 0.8 | 1.3 |
| United States | 0.43 | 0.47 | 0.3* | 0.8* |
| Japan | 0.39* | 7.05* | 0.8* | 1.5* |
| France | 0.77 | 0.93 | 0.6* | 1.1* |
| Germany, Fed. Rep. of | 0.90 | 1.78 | 0.9 | 1.2 |
| Italy | 1.90 | 2.85 | 0.9 | 1.2 |
| United Kingdom | 0.48* | 1.80* | 0.7* | 1.3* |
Data marked by asterisks are significantly different at the 95 percent confidence level.
Absolute change from preceding year.
For each of the seven countries, and according to both of the measures, variations in the net foreign balance were greater during the floating exchange rate period than during the 1960s. This, of course, is not the same as saying that the greater fluctuations were caused by increased exchange rate variability. Other economic developments of recent years, such as the sharp rise in energy prices and higher inflation, also undoubtedly played a role. While it may be impossible to sort out the differential effects with any degree of certainty, it is probably not unreasonable to suggest that exchange rate shifts were a contributing factor, at least to some extent.
Certainly the increase in the variance of the net foreign balance from the earlier to the later period has been quite pronounced. The average year-to-year change during 1974–82 was over 1 percent of GDP in every country except the United States, and in Japan it was 1.5 percent. In contrast, during the earlier period these changes averaged only 0.7 percent of GDP for all the countries of the group. Taking the square root of the variance to find the standard deviation of the net foreign balance, the increase in the average for the group was from 0.89 percent of GDP during 1960–70 to 1.52 percent during 1974–82.
Company Structure
It is sometimes argued that exchange rate variability bears most harshly on smaller companies, since the kind of uncertainty that it generates imposes proportionately greater costs on small units than on large units (Helleiner, 1981). There are two main reasons for this. First, exchange rate variability imposes certain costs that are essentially fixed in nature–for example, the need to assign management time to the appraisal of foreign exchange trends and the making of foreign exchange decisions. Since these costs do not necessarily rise with the volume of international transactions a company undertakes, they represent a smaller share of costs for companies with a large volume of international transactions. Second, larger companies find it easier and more practicable to hedge against the risks of currency fluctuations. Being large, they are likely to have more diversified production sources and markets in the first place; and if they do not, it is perhaps easier for them to plan the growth of their operations in such a way as to minimize their exposure to currency fluctuations. The existence of production and/or marketing subsidiaries in a number of different currency areas also enables large companies to allocate their financial assets and liabilities more easily toward currencies that offer security against loss or the possibility of profit.
McCulloch (1983) notes that input price uncertainty is a recognized motive for vertical integration. To the extent that exchange rate variability adds to this uncertainty, it would constitute an added motive for foreign direct investment, and a stimulus to the growth of vertically integrated and geographically diversified multinational enterprises. McCulloch considers that the expansion of U.S. direct investment in Canada during the period in which the Canadian dollar was floating, as well as more recent Japanese investments in the United States, fits this explanation. However, it is not difficult to think of alternative plausible explanations for such an investment pattern, such as the growing size, on technological grounds, of optimum production units, the desire to build security against protectionist measures, and simply changes in comparative advantage.
It is also sometimes argued that the sheer size of large companies gives them greater strength to withstand the losses that are brought about by temporary adverse movements in exchange rates. However, since the losses of large firms are themselves likely to be larger, this argument depends on large firms systematically holding (or having access to) financial reserves that are larger, as a proportion of sales, than smaller companies. Even if such a tendency existed, it should be noted that smaller firms typically have more flexibility to cut back costs (by releasing labor or reducing their unit cost) during a downturn in demand than do larger firms.
Evidence on these various subjects is scanty. In two surveys (Duerr, 1977; Blin, Greenbaum, and Jacobs, 1981) respondents were specifically asked how they had reacted to the risks created by greater currency volatility. The great majority of nonfinancial trading companies had undertaken administrative reorganizations that involved either the creation of new departments to monitor foreign exchange exposure or the allocation of greater amounts of senior management time to analysis and decisions involving foreign exchange exposure. They also noted that a variety of hedging mechanisms had been employed to protect their firms against the risk of loss or to enhance the chance of profit. These mechanisms included greater use of forward cover, more systematic distribution of liquid assets among different currencies, diversification of production sources, and others.
Respondents to surveys generally discounted the impact that uncertainty, and the need to hedge against it, had on their external operations. They were, of course, for the most part large corporations, so that it is difficult to tell whether effects that seemed small to them would loom larger for smaller units. At an anecdotal level, it is possible to identify small producers that have traced their difficulties to currency fluctuations. De Lorean Motor Company and Laker Airways, for example, both saw themselves in part as victims of the strong (and subsequently partially reversed) rise of sterling. Their competitors were mostly larger companies with more diversified costs and revenue sources, as well as more substantial financial resources.
Individual examples, while suggestive, do not in themselves make the case for the adverse effects of currency fluctuations, since it is impossible to know the extent to which a particular company’s difficulties stem from other factors.
International Specialization
A mechanism similar to the one that generates a concentration of output in large firms could also work to encourage a greater degree of international specialization. Where producers of a particular commodity in a given country are only marginally competitive (whether because of the factor endowment of the country or because of efficiency considerations in the industry concerned), the additional loss of competitiveness stemming from an adverse shift in the exchange rate could cause them to drop out of international competition. Given the large start-up costs in many modern industries, exchange rate swings may accentuate the tendency toward international concentration in production.
This factor would, of course, tend to increase trade levels over time, as countries specialized more in the production of commodities in which they received increasing economies of concentration and scale. However, the diminution of competition, and its incentive to increasing productivity and technological innovation, could eventually be detrimental to output and welfare. Such considerations can give rise to protectionist pressures, such as those coming from certain sectors of the automobile and electronics industries in Europe and North America, which feel that the market shares recently lost to Japanese producers will be hard to recoup, even if the Japanese exchange rate were to move up again.
Evidence on the effect of exchange rate variability on international specialization is not easy to collect. Table 2 provides information on the concentration of exports within a group of 14 industrial countries. Taking 19 separate commodity categories at the two-digit Standard International Trade Classification (SITC) level, there is little evidence that country concentration (defined as the share of exports accounted for by the three largest exporters) has increased for the majority of commodities over the period of floating exchange rates. In only 4 out of the 19 cases did this measure of concentration rise. During the prefloating period, by contrast (i.e., from 1961–63 to 1968–70), export concentration increased in half the country categories.
Industrial Countries: Changes in Export Concentration, 1961–811
(In percent)
Countries are Austria, Belgium, Canada, Denmark, France, the Federal Republic of Germany, Italy, Japan, the Netherlands, Norway, Sweden, Switzerland, the United Kingdom, and the United States.
Standard International Trade Classification (SITC) codes are 61 leather and leather manufactures; 62 rubber manufactures; 63 wood manufactures (excluding furniture); 64 paper and paper manufactures; 65 textile yarn and fabrics; 66 nonmetallic mineral manufactures; 67 iron and steel; 68 nonferrous metals; 69 metal manufactures; 71 nonelectrical machinery; 72 electrical machinery and appliances; 73 transport equipment; 81 plumbing, heating, and lighting fixtures and fittings; 82 furniture; 83 travel goods and handbags; 84 clothing; 85 footwear; 86 professional and scientific instruments, photographic and optical goods, and watches and clocks; and 89 miscellaneous manufactured articles.
Industrial Countries: Changes in Export Concentration, 1961–811
(In percent)
| Proportion of Total Exports Accounted for by Three Largest Exporters in 1968–70 |
|||||
|---|---|---|---|---|---|
| SITC Code2 | 1961–63 | 1968–70 | Change | 1979–81 | Change |
| 61 | 53.0 | 50.0 | −3.0 | 44.7 | −5.3 |
| 62 | 50.0 | 46.8 | −3.2 | 45.4 | −1.4 |
| 63 | 49.0 | 46.3 | −2.7 | 32.6 | −13.7 |
| 64 | 66.2 | 58.0 | −8.2 | 47.2 | −10.8 |
| 65 | 38.9 | 45.2 | 6.3 | 44.7 | −0.5 |
| 66 | 48.6 | 54.4 | 5.8 | 53.8 | −0.6 |
| 67 | 48.1 | 54.2 | 6.1 | 56.7 | 2.5 |
| 68 | 54.9 | 49.7 | −5.2 | 35.2 | −14.5 |
| 69 | 49.6 | 49.5 | −0.1 | 45.6 | −3.9 |
| 71 | 69.5 | 61.5 | −8.0 | 55.6 | −5.9 |
| 72 | 53.9 | 55.1 | 1.2 | 59.0 | 3.9 |
| 73 | 51.3 | 55.7 | 4.4 | 45.2 | −10.5 |
| 81 | 49.2 | 50.8 | 1.6 | 49.8 | −1.0 |
| 82 | 34.6 | 53.9 | 19.3 | 56.7 | 2.8 |
| 83 | 57.9 | 65.4 | 7.5 | 65.1 | −0.3 |
| 84 | 48.2 | 49.5 | 1.3 | 47.3 | −2.2 |
| 85 | 69.4 | 72.9 | 3.5 | 73.1 | 0.2 |
| 86 | 64.3 | 58.7 | −5.6 | 43.0 | −15.7 |
| 89 | 57.7 | 52.0 | −5.7 | 49.8 | −2.2 |
Countries are Austria, Belgium, Canada, Denmark, France, the Federal Republic of Germany, Italy, Japan, the Netherlands, Norway, Sweden, Switzerland, the United Kingdom, and the United States.
Standard International Trade Classification (SITC) codes are 61 leather and leather manufactures; 62 rubber manufactures; 63 wood manufactures (excluding furniture); 64 paper and paper manufactures; 65 textile yarn and fabrics; 66 nonmetallic mineral manufactures; 67 iron and steel; 68 nonferrous metals; 69 metal manufactures; 71 nonelectrical machinery; 72 electrical machinery and appliances; 73 transport equipment; 81 plumbing, heating, and lighting fixtures and fittings; 82 furniture; 83 travel goods and handbags; 84 clothing; 85 footwear; 86 professional and scientific instruments, photographic and optical goods, and watches and clocks; and 89 miscellaneous manufactured articles.
Industrial Countries: Changes in Export Concentration, 1961–811
(In percent)
| Proportion of Total Exports Accounted for by Three Largest Exporters in 1968–70 |
|||||
|---|---|---|---|---|---|
| SITC Code2 | 1961–63 | 1968–70 | Change | 1979–81 | Change |
| 61 | 53.0 | 50.0 | −3.0 | 44.7 | −5.3 |
| 62 | 50.0 | 46.8 | −3.2 | 45.4 | −1.4 |
| 63 | 49.0 | 46.3 | −2.7 | 32.6 | −13.7 |
| 64 | 66.2 | 58.0 | −8.2 | 47.2 | −10.8 |
| 65 | 38.9 | 45.2 | 6.3 | 44.7 | −0.5 |
| 66 | 48.6 | 54.4 | 5.8 | 53.8 | −0.6 |
| 67 | 48.1 | 54.2 | 6.1 | 56.7 | 2.5 |
| 68 | 54.9 | 49.7 | −5.2 | 35.2 | −14.5 |
| 69 | 49.6 | 49.5 | −0.1 | 45.6 | −3.9 |
| 71 | 69.5 | 61.5 | −8.0 | 55.6 | −5.9 |
| 72 | 53.9 | 55.1 | 1.2 | 59.0 | 3.9 |
| 73 | 51.3 | 55.7 | 4.4 | 45.2 | −10.5 |
| 81 | 49.2 | 50.8 | 1.6 | 49.8 | −1.0 |
| 82 | 34.6 | 53.9 | 19.3 | 56.7 | 2.8 |
| 83 | 57.9 | 65.4 | 7.5 | 65.1 | −0.3 |
| 84 | 48.2 | 49.5 | 1.3 | 47.3 | −2.2 |
| 85 | 69.4 | 72.9 | 3.5 | 73.1 | 0.2 |
| 86 | 64.3 | 58.7 | −5.6 | 43.0 | −15.7 |
| 89 | 57.7 | 52.0 | −5.7 | 49.8 | −2.2 |
Countries are Austria, Belgium, Canada, Denmark, France, the Federal Republic of Germany, Italy, Japan, the Netherlands, Norway, Sweden, Switzerland, the United Kingdom, and the United States.
Standard International Trade Classification (SITC) codes are 61 leather and leather manufactures; 62 rubber manufactures; 63 wood manufactures (excluding furniture); 64 paper and paper manufactures; 65 textile yarn and fabrics; 66 nonmetallic mineral manufactures; 67 iron and steel; 68 nonferrous metals; 69 metal manufactures; 71 nonelectrical machinery; 72 electrical machinery and appliances; 73 transport equipment; 81 plumbing, heating, and lighting fixtures and fittings; 82 furniture; 83 travel goods and handbags; 84 clothing; 85 footwear; 86 professional and scientific instruments, photographic and optical goods, and watches and clocks; and 89 miscellaneous manufactured articles.
Sectoral Allocation of Resources
As well as influencing specialization within the industrial sector, the uncertainty costs associated with exchange rate variability could influence the allocation of resources between traded goods and nontraded goods production within the economy. It is, after all, not simply the production of goods for export that is vulnerable to the uncertainties of exchange rate movements, but of any goods that enter into competition with foreign sources of supply, whether in domestic or export markets. Indeed, in many cases, it is production for the domestic market (e.g., automobiles, electronic equipment) where the problems of loss of international competitiveness have given rise to the most difficult adjustment requirements.
If the price at which traded goods can be marketed becomes more uncertain, this uncertainty will be reflected either in a higher average price or a reduced level of supply. In either case, it could be expected that there would be a transfer of resources toward the less volatile sector of nontraded goods production. Such a tendency would show up, statistically, in a reduction in the size of the traded goods sector, and also in an increase in its average profitability (to compensate for the greater variance in profits). Another consequence could be a reduction in the level of investment. If manufacturing industries have higher capital/labor ratios than the service sectors, then a shift in the pattern of output away from the capital-intensive industries would result in a declining rate of capital formation associated with a given level of output. It might also be expected that, even within industries, there would be a tendency for investment to decline, as businessmen hesitated to put in place long-lived capital equipment that might be rendered uneconomic (temporarily or permanently) by unexpected movements in exchange rates.
Testing these hypotheses presents a number of difficulties. In the first place, there is no clear distinction between industries producing traded goods and those producing nontraded goods. Rather, a country’s output can be thought of as lying along a spectrum of greater or lesser exposure to foreign competition. In Chart 6, two alternative approximations of the traded goods sector of the economy are presented–the manufacturing sector alone and manufacturing plus agriculture and mining–and the output of these sectors as a proportion of GDP in the seven major industrial countries is plotted for the period 1960–81. A second problem in interpreting the data in Chart 6 is that, other things being equal, the shares of these sectors would probably not be expected to remain constant over time. Just as Engel’s Law predicts a declining share of expenditure on food as income rises, the income elasticity of demand for services is believed to exceed that for goods, which would cause a secular fall in the proportion of total output emanating from the goods sector. A third complication is that manufacturing and mining production is more strongly affected by the business cycle than is the output of the economy as a whole. These cyclical effects are particularly apparent in Chart 6 in the periods 1973–75 and 1980–81.
Looking at manufacturing production only, it does appear that its share of aggregate output in the major industrial country group dropped more rapidly during the 1974–81 period of floating exchange rates than during the 1960s. If one looks at the broader measure of the traded goods sector, the situation seems to be the reverse, with the relative decline of traded goods output slowing during the floating exchange rate period. To abstract from cyclical influences by examining only periods of more or less steady growth, 1960–69 (average growth of 5.0 percent) can be compared with 1977–79 (4.1 percent average growth). Manufacturing production declined relative to GDP at about the same pace in both periods, whereas the GDP share of manufacturing, agriculture, and mining stabilized during 1977–79, in contrast with its rather steady fall during the earlier period.
While the evidence in Chart 6 is by no means conclusive, it does not provide any strong support for the proposition that the uncertainty associated with exchange rate variability has accelerated the shift of resources out of the traded goods sector. Such an effect might, however, be revealed if it were possible to construct a more reliable measure of traded goods output.
Investment
Empirical evidence on whether exchange rate variability has had a deleterious effect on investment is difficult to come by. The level of investment in an economy is determined by many factors, both social and economic, and the effect of any individual factor is not easy to isolate. Moreover, investment decisions also seem to be heavily influenced by such intangibles as “business confidence.”
The alleged negative impact of exchange rate fluctuations on investment (Hieronymi, 1983) could occur through several channels. By increasing business uncertainty, exchange rate variability could directly dampen producers’ willingness to undertake longer-term commitments to expand productive capacity. And if, as a consequence of greater risk in traded goods production, businessmen become more inward-looking, the resultant focus of output in the nontraded goods sector could depress profit margins and the incentive to invest there also. Moreover, nontraded goods output is concentrated in services, which generally have lower capital/labor ratios than traded goods production. Thus, investment requirements may be less in the nontraded goods sector, and the scope for productivity-increasing technological innovation may also be less.
Kenen (1979) investigated the effect of exchange rate variability on investment by testing whether the rate of growth of investment in 16 industrial countries between 1973–74 and 1976 was systematically associated with the month-to-month volatility in exchange rates. Each of the four equations he estimated had the right (negative) sign, and one achieved significance at the 95 percent confidence level. (Similar results are obtained when the investment growth considered is that between 1975 and 1976). More puzzling is the fact that the significant association is found with nominal but not with real exchange rate variability.
To further investigate the relationship between exchange rate volatility and investment, Table 3 compares nonresidential fixed capital formation in the seven major industrial countries in the periods before and after the move to more flexible rates. While the data in Table 3 do not shed light on the extent to which the focus of investment might have shifted to the nontraded goods sector during the floating exchange rate period, they do cast doubt on the often-stated contention that exchange rate turbulence in recent years has led to a sharp reduction in the rate of business investment. This table shows gross fixed capital formation (excluding residential construction) as a proportion of GDP for the periods 1960–70 and 1974–82. While investment rates were down significantly in the latter (floating rate) period in the Federal Republic of Germany and Italy, and somewhat in the United States, they were up sharply in Japan and slightly in Canada, France, and the United Kingdom; for the group as a whole the rate of investment was little changed.
Major Industrial Countries: Real Nonresidential Gross Fixed Capital Formation as a Proportion of Gross Domestic Product, 1960–82
(In percent)
Weighted according to the U.S. dollar value of gross domestic product in 1982.
Major Industrial Countries: Real Nonresidential Gross Fixed Capital Formation as a Proportion of Gross Domestic Product, 1960–82
(In percent)
| Average | Average | |||
|---|---|---|---|---|
| 1960–70 | 1974–82 | Change | ||
| Canada | 17.3 | 17.6 | 0.3 | |
| United States | 14.7 | 14.2 | −0.5 | |
| Japan | 22.0 | 25.4 | 3.4 | |
| France | 15.7 | 15.9 | 0.2 | |
| Germany, Fed. Rep. of | 16.2 | 14.9 | −1.3 | |
| Italy | 15.3 | 13.1 | −2.2 | |
| United Kingdom | 14.9 | 15.2 | 0.3 | |
| Weighted average1 | 16.3 | 16.5 | 0.2 | |
Weighted according to the U.S. dollar value of gross domestic product in 1982.
Major Industrial Countries: Real Nonresidential Gross Fixed Capital Formation as a Proportion of Gross Domestic Product, 1960–82
(In percent)
| Average | Average | |||
|---|---|---|---|---|
| 1960–70 | 1974–82 | Change | ||
| Canada | 17.3 | 17.6 | 0.3 | |
| United States | 14.7 | 14.2 | −0.5 | |
| Japan | 22.0 | 25.4 | 3.4 | |
| France | 15.7 | 15.9 | 0.2 | |
| Germany, Fed. Rep. of | 16.2 | 14.9 | −1.3 | |
| Italy | 15.3 | 13.1 | −2.2 | |
| United Kingdom | 14.9 | 15.2 | 0.3 | |
| Weighted average1 | 16.3 | 16.5 | 0.2 | |
Weighted according to the U.S. dollar value of gross domestic product in 1982.
These data are, of course, subject to differing interpretation. The substantial increase in investment in Japan may have reflected the country’s attempt to adjust to higher energy prices, and, if those prices rendered a large portion of the existing capital stock uneconomic, net investment may have risen considerably less than gross investment (or even have declined). Unfortunately, capital consumption allowances do not provide a reliable measure of the actual retirement of plant and equipment, and therefore accurate information on net additions to the capital stock cannot be derived from the national accounts data. On the other hand, very sharp increases in real wages in several European countries during the early 1970s are believed to have seriously undermined investment incentives in those countries, and the weakness of German and Italian investment rates during 1974–82 may have reflected this factor more than the impact of fluctuating exchange rates. Moreover, other things being equal, one might have expected the recessions of 1974–75 and 1980–82 to have led to lower investment ratios during the post-1974 period. In view of these opposing factors, about all that can be said with certainty is that there is no real evidence that the rate of investment in the industrial world as a whole has been weaker during the period of floating exchange rates than it was earlier.
