The Depressing News from Asia

Marcus Noland (PIIE), Sherman Robinson (International Food Policy Research Institute) and Zhi Wang (US International Trade Commission)

Policy Brief
September 1998

The World Bank began this decade talking of Asian miracles. It may enter the next one speaking of Asian depressions. After World Bank Vice President for Asia Jean-Michel Severino used the “d” word in May to describe the course of events in developing Asia, the organization was forced to come up with a technical definition for what had been a term of art.1

Depressing the news has been. Since July 1997, many of the region's bourses have fallen by half or more. The region's currencies have declined by even greater magnitudes against the US dollar. Growth forecasts have been steadily ratcheted down. Now, there is fear of a ripple of contagion throughout the rest of the world.

In the United States, the effect of the Asian crisis has been to slow growth, generate real exchange rate appreciation, and contribute to a growing trade imbalance. In Global Economic Effects of the Asian Currency Devaluations, we analyzed the impact of events in Asia on both the region and the rest of the world.2 In this piece, we update our analysis to take into account several events—most notably, the dramatic decline in the Japanese yen—that have occurred since the earlier work went to press.

Analytical Background

To analyze the impact of the Asian crisis on production and trade across the world economy, we developed a computable general equilibrium (CGE) model that divides the world economy into 17 regions and 14 sectors. In the model, the regions are linked by trade in differentiated products, and there are detailed data on sectoral output, employment, and trade. The model does not contain an explicit time dimension and, hence, it is not well suited for quarterly forecasting. Instead, it is more useful for analyzing the underlying structural changes that accompany adjustment. The Asian crisis is modeled as a set of real effective exchange rate depreciations across Asia that are paired with supply-side contractions to capture the impact of financial disintermediation on the ability of these economies to respond to the exchange rate change. In Global Economic Effects of the Asian Currency Devaluations, we calibrated these shocks on the basis of developments between July 1997 and February 1998 and presented “low,” “high,” and “medium” shock scenarios.

. . . the case of Japan has been
more surprising and, at least from the standpoint
of the rest of the world, more disturbing.

Our analysis indicated that, broadly speaking, countries fall into four groups: the most affected countries (Thailand, Indonesia, Malaysia, the Philippines, and South Korea); intermediate countries (Japan, Taiwan, and Singapore); the rest of the industrial world (Western Europe, the United States, and Australia and New Zealand); and the unique case of China. We predicted that the most affected countries experience substantial falls in output and significant positive swings in their trade balances, in large part due to import compression. Consequently, domestic absorption (the resources available for domestic use or aggregate final demand) falls dramatically in these countries. Investment, in particular, collapses, and with it demand for local construction services, imported capital goods, and industrial intermediates. We obtained these qualitative results regardless of the scenario-specific pattern of exchange rate and supply-side contractions that we assumed.

The results for the second set of economies were more subtle and sensitive to variations in the underlying assumptions, even in qualitative terms. These economies experience real exchange rate appreciations vis-à-vis the most affected countries, but real depreciations against the rest of the world. The net impact of the crisis on these countries depends on their unique, economy-specific, trade links with other regions. So, for example, in our “medium shock” scenario, the Japanese machinery sector experiences a decline in output as important export markets in developing Asia dry up. In contrast, the motor vehicle industry—less tightly linked to the rest of Asia—expands as Japanese automakers increased their relative competitiveness in markets such as the United States and Western Europe.

The results in the rest of the industrial world were a mirror image of the adjustments in the most affected parts of Asia. In this group of countries, the traded-goods sectors contract due to a decline in exports of capital goods to Asia, where investment had collapsed, and due to increased competition from imports in a broad range of manufactures. For example, our model showed that the US machinery sector contracts by around 3 percent because of the crisis. At the same time, the nontraded goods sector actually expands, because sectors such as construction are stimulated by falling interest rates, by the positive income effect of the appreciating exchange rate, and by the release of resources from the traded-goods sector. In the United States, housing and construction activity increases by around 1 percent directly as a result of the Asian crisis.

In aggregate, the greatest impact is felt in Western Europe, the largest single region in the model. The greatest relative impact is felt in Australia and New Zealand, the two countries most intimately bound to Asia. Indeed, the recent collapse of the Australia dollar is evidence of the Australian economy's inability to adjust to a shock of the magnitude that it has experienced because of the Asian crisis. The update on our model that we provide in the next section of this policy brief takes into account this collapse.

One congressman claimed that
the increase in the trade deficit ‘is a
direct result of poor trade policies'
and ‘is at an all-time high
because of our government's
irresponsible trade policy.'

In our original model, changes in world trade patterns due to the Asian crisis affect the United States, generating significant deterioration in its balance of trade and changes in the structure of exports, imports, and production. Competing domestic producers notice increased imports and often react by seeking assistance from the government. Our analysis of trade policy issues indicated that bilateral trade tensions are largely a function of bilateral trade imbalances—the bigger the US-Japanese trade imbalance, the more that US producers complain about competition from Japanese imports.

While economists argue that the aggregate trade balance is a macroeconomic phenomenon and that bilateral balances are irrelevant, our econometric analysis indicated that any deterioration in the trade balance leads to increased trade tension with the countries most affected. Our results indicated that the Asian crisis generates a significant increase in the US trade deficit, especially with Japan and South Korea, and we concluded that an increase in trade disputes could be expected on the basis of the historical record. The process has already begun. One congressman greeted the May trade figures with a press release claiming that the increase in the trade deficit “is a direct result of poor trade policies” and “is at an all-time high because of our government's irresponsible trade policy.”3 Other congressmen will undoubtedly follow, and the political pressure to close the US market will surely grow.

Finally, there is China—the shoe that has yet to drop. We estimated that the crisis has a relatively modest impact on China's external balances. Ceteris paribus, a real devaluation of under 10 percent is sufficient to restore China's precrisis trade surplus. Yet, given the enormous internal problems that China confronts and the likelihood that a Chinese devaluation would spark another round of financial turbulence in Asia, we counseled (and continue to counsel) that China pursue alternatives to currency devaluation to address the need to maintain aggregate demand while restructuring its state-owned enterprises. At least to this point, Chinese policy has followed that advice.

Updating the Results

For some countries, such as Thailand and South Korea, the “medium shock” assumptions have proved remarkably prescient. For other countries, most notably Indonesia and Japan, performance since February has been worse than even our “high shock” case predicted. (In the case of Indonesia, this was to be expected—it was clear that the political transition in Indonesia was well beyond “normal” economic experience, and the country's adjustment to the crisis could take forms such as hyperinflation or famine that simply could not be captured by the model.)

However, the case of Japan has been more surprising and, at least from the standpoint of the rest of the world, more disturbing. Japan's real exchange rate remained roughly stable from July 1997 through March 1998 but then began a steady slide, falling by more than 10 percent between March and June. At the same time, Japan entered a recession, registering negative growth in the fourth quarter of 1997 and the first quarter of 1998. (Results for the second quarter of 1998 have not yet been released.)

Table 1, based on our updated model, reports the changes in global trade balances that arise from several Asian crisis scenarios, while table 2, also based on our updated model, reports some key bilateral balances. The first line, labeled “scenario 1,” reports the model's results when we program the exchange rate and supply-side shocks that were experienced between July 1997 and June 1998 in developing Asia and Australia and New Zealand but do not program any Japanese adjustment. (We annualize the numbers once adjustment has occurred. One could interpret these as approximating what might be observed in calendar year 1999.) In this mental experiment, the developing Asia countries generate increased trade surpluses of more than $150 billion dollars.

Scenario 2 includes the exchange rate and supply-side shocks that were experienced between July 1997 and June 1998 in developing Asia and Australia and New Zealand as well as a 10 percent depreciation of Japan's real effective exchange rate and a 2 percent supply-side contraction of the Japanese economy, approximating what has actually occurred. (Given the other economies' exchange rate changes, this is equivalent to a yen-dollar rate of around 135.) The impact is to increase Japan's trade balance by nearly $90 billion, reversing Japan's trade balance losses in the first experiment and generating a surplus of $55 billion relative to the baseline solution. Relative to the thought experiment represented by scenario 1, Japan “takes back” more than $20 billion from the trade balances of developing Asia, because the positive swings in their trade balances with Japan in scenario 1 are eroded and Japan gains market share in third country markets such as the United States. Indeed, developing Asia loses more in scenario 2 than the $19 billion Japan pledged in the “second line of defense” commitments as part of the International Monetary Fund's programs with Thailand, Indonesia, and South Korea.

The global positions of the United States and Western Europe decline by roughly $25 billion each in this scenario. For the United States, this implies that the Asian crisis to date adds nearly $60 billion to the trade deficit in nominal terms and even more in real terms (relevant to production and employment) when the fall in import prices associated with the real appreciation are taken into account. As shown in table 2, most of this increase is in US bilateral deficits with Japan and South Korea, which grow by $25 billion and $10 billion, respectively.

In scenario 3, the Japanese real exchange rate depreciates by 20 percent (roughly equivalent to a yen-dollar rate of 155), and the economy experiences a negative 4 percent supply-side shock. This experiment goes beyond what we have observed in 1998 and indicates what could happen if the problems in Japan intensify. The Japanese surplus increases by another $80 billion, or nearly $170 billion relative to the base. The external balances of the United States, Western Europe, and developing Asia each deteriorate by an additional $15-20 billion. As shown in table 2, South Korean gains vis-à-vis Japan are completely wiped out under this scenario.

Some have argued that a significant depreciation of the yen is necessary to restore economic growth in Japan,4 and one commentator has gone so far as to argue that “all Asian economies and the world economy would be far better off with Japan moving to a 2-3 percent sustainable growth rate and the yen somewhere between 150 and 200 per dollar than with continued stagnation and a stable yen.”5 Our model allows us to examine this proposition directly. In scenario 4, we depreciate the yen 30 percent in real terms and augment Japanese total factor productivity by 3 percent.

The results reported in tables 1 and 2 do not support this contention. Although the Japanese output increase partly offsets the impact of the depreciation, the relative price effect through the exchange rate change still predominates, and the Japanese surplus increases nearly $130 billion relative to scenario 2, the adjustment which has occurred thus far, almost $40 billion of which is accounted for by an increase in the Japanese surplus with the United States. In this scenario, Japan reduces the global trade surpluses of Thailand, Indonesia, and South Korea by more than its bilateral aid commitments. While it is certainly true that more robust growth in Japan could have beneficial externalities not captured by this model (e.g., by facilitating resolution of the Japanese banking problems and restoring capital flows to the rest of Asia), the evidence from the real side of the economy does not support the proposition that Asia and the rest of the world would be better off with substantially more yen depreciation even if accompanied by growth in Japan. Rather than relying on export-led recovery to reinvigorate its economy at the expense of the rest of Asia, Japan should undertake policies as outlined in Adam Posen's Restoring Japan's Economic Growth to generate a domestic demand-led expansion and pull itself and the rest of the region out of recession.6


This policy brief summarizes and updates our recent work on the global economic effects of the crisis in Asia, the most important new development being the deterioration of conditions in Japan. While oceans of ink have been spilled on a prospective devaluation in China, the ongoing depreciation of the yen is a much greater threat to the region and the world at large.

While a depreciation of the yen may be a necessary component of Japanese adjustment, it is critical that Japan not rely predominately on external demand to extricate itself from recession. Such a development in Japan would significantly impair the ability of the poorer, more deeply affected, Asian countries to recover from the crisis. In the United States, the projected record trade deficit and the rising deficits with Japan will undoubtedly contribute to protectionist political pressures. For Japan to pursue such a course of action when other alternatives are available is folly. 

Table 1: Change in global trade balances (billions of dollars)

  United States Western Europe Australia and New Zealand Japan China Developing Asia Rest of World

Scenario 1
(Asian adjustment through June 1998, except Japan)
-34.6 -50.4 8.2 -33.1 -11.0 151.8 -30.8
Scenario 2
(10 percent Japanese depreciation plus 2 percent negative supply-side shock)
-58.0 -75.8 6.0 55.2 -15.8 129.5 -41.1
Scenario 3
(20 percent Japanese depreciation plus 4 percent negative supply-side shock)
-79.5 -100.0 4.1 135.2 -19.9 110.1 -50.0
Scenario 4
(30 percent Japanese devaluation plus 3 percent positive supply-side shock)
-96.4 -120.4 3.7 184.9 -21.0 99.7 -50.5

Table 2: Change in key bilateral trade balances (billions of dollars)

  US-Japan US-China US-South Korea Japan-South Korea Japan-China

Scenario 1
(Asian adjustment through June 1998, except Japan)
-2.0 0.2 -10.0 -9.0 0.7
Scenario 2
(10 percent Japanese depreciation plus 2 percent negative supply-side shock)
-25.0 -0.2 -10.3 -4.1 8.7
Scenario 3
(20 percent Japanese depreciation plus 4 percent negative supply-side shock)
-46.3 -0.5 -10.8 0.0 15.9
Scenario 4
(30 percent Japanese depreciation plus 3 percent positive supply-side shock)
-62.7 -0.8 -10.7 -2.0 19.8


1. The World Bank settled on the following technical definition: A fall in income equal to or greater than the 26 percent decline in output experienced by the US between 1929 and 1932. For the United States, the National Bureau of Economic Research (NBER) defines a “recession” as occurring when there are two consecutive quarters of GDP decline.

2. Marcus Noland, LiGang Liu, Sherman Robinson, and Zhi Wang. 1998. The Global Economic Effects of the Asian Currency Devaluations. Washington: Institute for International Economics.

3. Representative Sherrod Brown, “Increased Trade Deficit Points to Failed U.S. Trade Policies,” 17 July 1998.

4. Allan Meltzer, “Time To Print Money,” Financial Times, 17 July 1998.

5. John Makin. 1998 Mr. Greenspan's Dilemma. Washington: American Enterprise Institute for Public Policy Research (August).

6. Adam Posen. 1998. Restoring Japan's Economic Growth. Washington: Institute for International Economics (September).