The appreciation of the yen after 1985 dramatically shaped the structure of Japanese trade flows, with a spectacular rise in manufactured imports
In recent years, Japan’s current account surplus has drawn enormous interest, becoming the rationale, at times, for calls for greater protection against Japanese exports. The assumption has been that Japanese trade flows are not responsive to market signals. Otherwise, the argument goes, Japan would not have been able to amass persistent annual current account surpluses of around $80 billion in the second half of the 1980s.
But, in actuality, the growth of Japanese exports in volume terms during this period was very slow by historical standards, while import volume growth surged—a significant departure from past trends. Indeed, the large surpluses entirely reflected a shift in the terms of trade, owing to lower imported oil prices and a sharp appreciation of the yen. This article examines how Japanese trade flows have changed since 1985 and why, highlighting the role of the yen and the unprecedented burst of foreign direct investment by Japan. It also tries to answer the question: Has import growth been low in light of the rapid appreciation of the yen?
Developments in exports
Japan is the world’s third largest exporter (after the United States and Germany), supplying nearly 10 percent of total world exports in 1989. Over the last three decades, this share has increased dramatically, reflecting mainly the rapid growth of Japan’s economy as a whole (see Chart 1). But in terms of real market penetration—that is, the ratio of Japanese exports to partner countries’ total manufacturing imports adjusted for price changes—exports remained fairly constant, on average, during most of the 1970s and 1980s, with increases largely confined to the 1960s, when Japanese economic development was taking place at a rapid pace.
Even so, real market penetration has fluctuated quite widely since the early 1970s, closely tracking changes in Japanese competitiveness. In general, when competitiveness—measured either by relative unit labor costs or export prices—improved, market penetration rose, only to decline when the competitiveness gain was reversed. The most pronounced and long-lived decline occurred when the yen appreciated sharply after the Plaza Agreement of September 1985, which set the stage for coordinated foreign exchange intervention aimed at curbing the rise of the dollar. During 1985-89, Japanese real export growth was the slowest for any five-year period of the past three decades.
Traditionally, Japanese exports have been composed almost entirely of manufactured goods, concentrated in relatively few commodity categories. But changes in the shares of exports by end-use—industrial supplies, consumer goods, and capital goods—have been underway for some time, reflecting a growing specialization of Japanese industries in more sophisticated, high quality products, and changing comparative advantage (see Table 1). In 1965, industrial supplies accounted for the largest share of exports, followed by capital goods and consumer goods. But by 1985, the emphasis had switched, with the latter two taking up over 75 percent, and the share of industrial supplies slipping to 20 percent.
|Iron and steel||15.3||14.7||11.9||7.7||5.4|
Data processing machines, integrated circuits, and telecommunications equipment.… Data not available.
Data processing machines, integrated circuits, and telecommunications equipment.… Data not available.
Behind these movements were the successes and failures of a few key products. The declining share of industrial supplies in the 1960s reflected, in large part, the decreasing importance of textiles exports as Japanese exporters lost market share to producers in lower-cost countries. For a while, this category benefited from the rise of the iron and steel industry in the second half of the 1960s, but in the 1970s, this industry also fell victim to increased foreign competition. Gradually, the export of trucks, machine tools, and industrial and commercial equipment, as well as automobiles and other consumer durables, gained prominence. The export of motor vehicles, for example, reached its peak in 1985—with over seven million units, of which five million were passenger cars—accounting for 20 percent of all exports.
The decline in the share of industrial supplies continued in the second half of the 1980s, as the rising yen contributed to a further weakening of metals exports and squeezed the already small textiles sector. At the same time, exports of capital equipment continued to grow in importance, led by new technology-intensive products, such as data processing machines, integrated circuits, and telecommunications equipment.
But unlike in the past, the share of consumer goods declined, primarily reflecting a weakness in non-automobile exports (although passenger car exports dropped in numbers, the total share was maintained, in part because voluntary export restraints to major foreign markets helped tilt the composition toward higher quality, higher priced units). The main cause was a loss in cost competitiveness, particularly to producers in the newly industrializing economies (NIEs) of Asia—Hong Kong, the Republic of Korea, Singapore, and Taiwan Province of China. While the emergence of new Japanese exports, such as videocassette recorders, compact disc players, and video cameras, initially helped, the appreciation of the yen boosted the competitiveness of similar products assembled outside Japan.
Indeed, the high value of the yen, along with a desire to overcome foreign protectionism, precipitated a surge in Japanese foreign direct investment (FDI), as Japanese producers sought cheaper overseas production centers (see Chart 2 ). Until the second half of the 1980s, Japanese FDI had lagged behind that of other industrial countries, amounting to the equivalent of only 1-2 percent of total private domestic fixed capital formation; this compares with a US average of nearly 5 percent in the 1970s. But in 1986, Japanese FDI began to soar, accumulating to $112 billion in 1986-89, or roughly 4½ percent of private domestic investment—close to the US total.
Chart 1.Trends in Japanese exports and imports, 1965-89 Chart 2Surge in foreign direct investment, 1970-89
Source: Japanese balance of payments data.
While much of the overseas investment has been in the financial and real estate sectors, a growing proportion has been directed toward the manufacturing sector. One of the most prominent examples has been in the US automobile industry—between 1985 and 1989, production of Japanese cars in the United States rose from under 200,000 units to over 750,000 units, or nearly 30 percent of all Japanese cars sold in that country.
Not surprisingly, this new phenomenon of a high level of FDI became a channel for a number of changes in the nature of Japanese trade flows. Additional demands surfaced for Japanese physical capital to set up overseas operating plants; at the same time, output from these plants, when it came on stream, began to displace exports of consumer goods made in Japan. On the regional level, there was a sharp rise in exports to the NIEs, which were large recipients of Japanese FDI in this period, along with Europe and the United States.
Developments in imports
As with exports, Japanese imports now take up a substantial share—7 percent in 1989—of the world total, a sharp increase from the only 1.5 percent level in 1950. Again, the growth of the Japanese economy accounts for a large part of the underlying increase, although the value of imports has fluctuated quite widely from year to year due to price and currency changes. For example, as oil prices increased sharply in 1973-74 and again in 1979-80, the share of imports in GNP jumped to 13.25 percent in 1980 from around 10 percent in the 1960s and early 1970s. Then in the 1980s, imports declined sharply as a proportion of GNP, owing to weaker oil prices and appreciation of the yen (see Chart 1).
But if the effects of oil and yen movements are removed, real import penetration—measured as the share of imports in total domestic demand adjusted for price changes—is seen to have changed little, on average, from 1975 to 1985, whereas in the second half of the 1980s, such penetration increased markedly. The largely unchanged picture during 1975-85 reflected offsetting trends in manufactured and nonmanufactured imports—real manufactured imports grew faster than overall domestic demand, while real nonmanufactured import penetration declined. Furthermore, although nonmanufactured import growth picked up only modestly after the Plaza Agreement, manufactured import growth surged, substantially redressing the traditional compositional bias against manufactures (see Table 2).
|Non-oil Industrial supplies||60.9||62.7||61.6||54.8||47.3|
|Food and drink||25.8||19.0||20.4||19.8||17.7|
Until recently, imports had been largely composed of nonmanufactured goods—mostly commodities—reflecting Japan’s scarcity of natural resources, with manufactured goods amounting to a comparatively small proportion of GNP by industrial country standards, despite low tariffs and few formal trade barriers. This has provoked much (as yet unresolved) discussion as to whether Japan imports an “abnormally” low amount of manufactured goods. In particular, the concern centers on the low degree of intra-industry trade—other countries produce, export, and import the same commodities to a much greater extent than Japan does. But because there is little theoretical or intuitive basis for this type of trade, such a case is difficult to prove.
One important reason for the relatively strong growth in manufactured imports since 1975 has been the changing sophistication of the Japanese economy and, notably, a shift away from materials-processing activities. Taking the earlier examples of the textiles and iron and steel industries, the demand for imported textile raw materials and iron ore steadily declined and was replaced, first, by imports of textile materials and iron and steel, then subsequently by finished garments and steel products. Thus, the more rapid growth of manufactured imports reflected relatively strong demand for manufactured industrial supplies, as well as capital and consumer goods. In addition, the relative importance of nonmanufactured imports has been eroded since 1975 by energy conservation in the wake of higher real oil prices.
The surge in manufactured import growth in the second half of the 1980s represented an accentuation of these past trends. The growth in consumer goods imports was, in particular, remarkable, quadrupling in volume terms between 1985 and 1989. Significant foreign penetration was achieved in several markets—notably, automobiles and electronic consumer durables—that had previously been supplied almost entirely by domestic producers. Imports of passenger cars, for example, rose to nearly 200,000 units in 1989, compared with 50,000 in 1985. The rise in consumer goods imports partly reflected the effects of FDI, as Japanese producers shifted some of their production facilities offshore; imports from the NIEs increased particularly rapidly in this period.
Clearly, the second half of the 1980s both accentuated past trends and brought forth some new developments. But were these changes in keeping with historical patterns?
Or were there some surprises? To answer these questions, econometric tests were run to determine how imports and exports normally behave following exchange rate changes and alterations in domestic and foreign income. Then these results were used to estimate how the historical behavior could explain developments after 1985.
On the export side, the study shows that for total export volume, the stagnation in 1986-88 was a normal response to the large loss in competitiveness, although export growth should have rebounded somewhat more strongly than it did in 1989 (see Chart 3). For capital goods exports, actual volume growth was a little stronger than predicted, particularly in 1987-88, although growth was a little slower than expected in 1989. By contrast, consumer goods export volume was weaker than anticipated for most of the 1986-89 period. While the yen had the expected effect of depressing the level of exports, it also had a somewhat greater-than-usual impact on export composition. The direction of these compositional changes was consistent with the surge in Japanese FDI.
As for imports, the results suggest that a large part of the surge in manufactured goods in the second half of the 1980s cannot be explained by developments in the domestic and world economic environment. While the yen’s appreciation and strengthening of domestic demand would have stimulated manufactured import growth, the response was much greater than historical experience would have suggested: actual manufactured imports greatly exceeded predicted levels. For non-manufactured imports, demand was a little weaker than expected, although this result may reflect problems in properly accounting for nonprice factors, such as energy conservation and quotas on certain agricultural products.
Chart 3Actual versus predicted exports and imports, 1984-89
Source: IMF staff estimates.
1Excluding nonmonetary gold.
Note: Predicted values in the shaded area refer to regression-based extrapolations of historical trends that take into account actual developments in income and relative prices in the period 1986-89.
An analysis of imports by end-use shows that the underprediction of the strength of manufactured import demand stemmed from the surge in consumer goods imports, which was well in excess of what might have been expected on the basis of developments in overall consumer demand and relative import prices. However, the rapid growth in capital goods imports was largely in line with historical trends and reflected the strong growth in Japanese investment in this period.
During the second half of the 1980s, the high value of the yen exerted a strong influence on the level and structure of Japanese trade flows. The ensuing loss of competitiveness acted as a forceful brake on exports and encouraged a surge in manufactured import demand. The loss of export competitiveness was felt most keenly in the consumer goods sector, whose share of exports declined in contrast to past trends, while capital goods exports remained comparatively buoyant. Import expansion was led by spectacular growth in consumer goods categories, well in excess of their usual response to lower relative import prices and the overall strength of domestic demand. Unusually large outflows of Japanese foreign direct investment, as well as “threshold” effects—that is, the appreciation was viewed as sufficiently large and permanent to make it profitable for importers to undertake the costly exercise of attempting to penetrate a new market—appear to have provided two of the vehicles through which the exchange rate elicited these particular changes.
Thus, far from being unresponsive to market signals, Japanese trade flows adapted in a remarkably flexible way to the economic environment of the late 1980s. Whereas this evidence of flexibility will not satisfy the hardest of critics of Japan’s trade practices, it certainly flies in the face of popular views of an economy that is closed to foreign goods.