The world economy has been in serious recession since the financial crisis began. According to IMF, world GDP grew 3.2% in 2008, is expected to shrink 1.3% in 2009, and will remain low in 2010. Among the many nations harmed by the crisis, Japan has been one of the most seriously affected. Japan’s GDP fell by 0.6% in 2008, and is predicted to fall by 6.2% in 2009. US GDP, by comparison, grew 1.1% in 2008 and is predicted to fall by only 2.8% in 2009.
In the early stages of the crisis, it seemed that Japan might escape a serious recession, since its financial sector was relatively small. However, Japan’s economic recession eventually became the most serious among OECD countries in terms of GDP growth rates.
The seriousness of the Japanese recession stemmed mainly from the sharp decline in Japanese exports as well as the reduction of private capital investment. The decline in US import demand sent shockwaves through export-dependent countries around the world. But the impact of the reduction of US demand on trade varied across countries because of differences in comparative advantage.
This chapter presents new evidence on why Japanese exports declined so massively due to the change in US demand.
Variation of the US imports by partner and product
After recovering from the recession caused by the crash of the IT bubble, the US economy recorded a significant increase in imports – almost 10% annually since 2002. This boom reversed suddenly when the financial crisis hit US demand. The shock affected many nations, but in an uneven manner (Table 1). For example:
- US imports from Japan dropped by 40% in the first quarter of 2009, while total US imports were off by 30% (quarter on quarter).
- Imports from China decreased by only 10%.
- Imports from Canada decreased almost as sharply as those from Japan.
These differences reflect bilateral exchange rate movements but also partner-specific effects that stem from a combination of an uneven drop in imports by product-group, and an uneven distribution of these product-groups that the various partners’ export to the US.
As shown in Table 1a, some of the biggest drops in US imports occurred in the products that account for a very large share of Japanese exports to the US (especially automobiles, electronics and electric goods, and machinery industries). US import demand plummeted by over 45% for automotive vehicles, parts, and engines, and industrial supplies and materials between the 1st quarter of 2009 and 1st quarter of 2008. Oil imports dropped even more – by more than 50% -- but the reduction in imports of food, consumer durables and nondurables was comparatively small.
In 2008, the share of automobiles and parts in Japanese exports to the US was one-third, electronics and electric machinery 16%, and machinery one-fifth of total exports to the US. In total, exports in these three product categories amounted to over 70% of total bilateral exports.
Table 1a Change in US imports, by country and product (%).
Source: US data from Statistics of International Transaction, US Department of Commerce.
As Table 1a shows, Canada – whose exports to the US depend heavily on automotive vehicles, parts, and engines – experienced a drop of sales to the US almost as large as Japan’s. By contrast, the relatively small response of consumer goods to the US drop in income helps explain why Chinese exports to the US fell by much less.
Table 1b shows directly that Japanese exports to the US and Europe declined significantly more than average. This US-specific impact was especially important for Japan; the US is Japan’s largest export market, absorbing one-fifth of her exports in 2008. The figures also confirm that the US’s big drop in auto trade is also found in Japanese export statistics. While the average drop was 47% in 2009 Q1, it was almost 60% in autos and parts. Capital goods more generally fell in line with the average, but the fall-off was less severe in consumer goods and industrial supplies.
Table 1b Change in Japanese exports by country and product (%).
Source: Japanese data are from Trade Statistics, Ministry of Finance.
Reflecting the internationalisation of Japan’s industrial supply chains, Table 1c shows that the decline of US demand indirectly caused a sharp decline of trade between Japan and East Asian countries. Japan’s imports of auto and parts fell almost as much as its exports.
Table 1c Change in Japanese imports by country and product (%).
Source: US data from Statistics of International Transaction, US Department of Commerce; Japanese data are from Trade Statistics, Ministry of Finance.
The “triad trade” and offshore sourcing
The changing source of US imports since 2000 are noteworthy. The facts reflect the internationalisation of Japan’s manufacturing, especially China’s growing role in Japanese firms’ value-added chains.
Table 2 shows that between 2000 and 2007, the US import share from East Asia hardly changed. This overall stability, however, hides a big shift within the region. China’s increase – from 8% to 16% – was offset by declines in the shares from Japan, the Asian newly industrialized economies (NIEs), and the largest ASEAN nations.
Table 2 Shifting source of US imports: China’s rise and Japan’s fall
Source: Statistics of International Transaction, Department of Commerce, US.
Note: * Other East Asia includes Indonesia, Korea, Malaysia, Philippines, Singapore, and Thailand.
Matching this shift in US import patterns is the shifting geographical distribution of Japan’s exports. In 2000, the share of Japanese exports was 30% to the US and only 6% to China. In 2007, these shares switched to 20% to the US and 15% to China. This trend has continued and the most recent data shows that China is Japan’s largest export market.
One of the driving forces behind the geographical reorientation of trade has been the development of a “trade triad” between the US, China, and Japan. Under triad trade, Japan exports parts and intermediate goods to China where they are processed into final goods and then exported to the US. This type of triad is typically organised by Japanese MNCs in China.1 According to the statistical facts presented in Ito, Tomiura and Wakasugi (2009) outsourcing to China exceeds 50% of all Japanese outsourcing. Moreover, offshore assembly of final goods in China has reached 20% of all the offshoring of Japanese firms.
Direct and indirect exports to the US
What this triad trade means is that Japanese multinationals export to the US both directly (from Japan) and indirectly (via China). Those goods coming directly from Japan tend to be high-end, knowledge- and skilled-labour-intensive. Lower-end goods, which are intensive in the unskilled labour, which is so abundant in China, tend to come to the US via China.
As far as Japan’s export collapse in concerned, the main point is that the drop in US imports from China hurt Japanese exports to China, thus magnifying the direct effect discussed above.
Variation of Japanese exports: Extensive and intensive margins
The so-called ‘new-new trade theory’ alerts us to the fact that total export changes can be driven by two very different adjustments – changes in the number of products exported, and changes in the amount of exports per product exported.
Recent research points out that these two margins may act very differently during gradual expansions and rapid contractions such as the 1997 Asian crisis.2 Bernard, Jensen, Redding, and Schott (2009) show that although an expanding extensive margin plays a very important role in gradual trade growth, most of the sudden reduction in trade during the 1997 crisis came from the intensive margin. (The chapter by Schott in this volume provides preliminary evidence that the same thing has happened in the US during the Great Trade Collapse.)
Following up this suggestion, we investigate the behaviour of the two margins in Japan’s recent trade collapse. We do this by using the well-known gravity relationship (which links bilateral trade to bilateral distance and their economic mass) extended to consider the extensive and the intensive margins.
Due to data constraints, we cannot follow the procedure that has been applied to US data. We do not have access to the number of exporting firms by destination country and the number of products for Japan. Therefore, we use total trade data disaggregated at the HS6 level (about 5000 products). Specifically, we decompose the total bilateral trade volume into the number of products (the extensive margin) and the average trade value per product (the intensive margin).
Details of the estimates are gathered in the appendix. Comparison of the results for Japanese exports to the US and Japanese exports to China suggests that changes in Japanese export structures from 1990 to 2007 are as follows:
- Japanese exporters to the US have narrowed the product range of the goods exported to the US;
- Japanese exporters have expanded the range of goods they exported to China, along with its rising income;
- The reduction of the range of exports to the US, with the expansion of the range of exports to China, coincides with the development of the trade triad due to the expansion of sourcing in China by Japanese multinationals.
In summary, our estimates confirm the notion introduced above that one of the reasons Japan’s exports were hit so hard is that exports to both the US and China were linked in an important way to US demand patterns. When US demand fell sharply for the type of goods in which Japan has a comparative advantage, Japan’s sales to both the US and China drop precipitously.
Extensive and intensive margins in the crisis
The estimated results of the variation of Japanese exports after 1990 provide a standard to evaluate the changes of Japanese exports after the financial crisis.
Japanese exports to the US
Specifically, Figures 1a and 1b show the extensive and intensive margins predicted by the estimated gravity equations (see appendix), comparing them to the actual Japanese exports to the US from 1990 to 2008.
Figure 1a Extensive margin, exports to China, actual vs. predicted
Source: Authors' calculations.
After 1990, the extensive margin (both actual and predicted) displays a downward trend, with the tendency becoming more marked from 2003. Observe that both the actual and predicted fell sharply with the US recession in 2008; there is some disparity between actual and predicted values during the crisis year, but not one which is particularly out of line with earlier, in-sample deviations. The results are quite different for the intensive margin. After 2003, when the US economy had recovered from the IT recession, the actual value of the intensive margin grew remarkably faster than the predicted value before 2007.
We find a marked deviation between predicted and actual when it comes to the extensive margin in the crisis year 2008. Although the average export value from Japan to the US showed the increase far exceeding a trend before the financial crisis, in 2008 it fell sharply and is now below the predicted trend.
The decline came most from the intensive margin
This means that the downturn of Japanese exports to the US, due to the demand contraction following the financial crisis, was far larger than expected according to recent historical experience. Importantly, this unexpectedly large drop came from the intensive margin, not the extensive margin. This is in line with the findings of Bernard, Jensen, Redding, and Schott (2009) in their investigation of US imports from crisis-stricken Asia during the 1997 crisis.
Figure 1b Intensive margin, exports to China, actual vs. predicted
Source: Authors' calculations.
With regard to the exchange rate’s role, before the financial crisis the yen depreciated, but more recently it has appreciated significantly. The estimation results in the Appendix to this paper show that the significant upward trend of the intensive margin after 2003, and the sharp drop after 2007, are related to the exchange rate changes of yen per dollar.
Japanese exports to China
Figures 2a and 2b show similar numbers for Japan’s exports to China, specifically the actual and predicted values of the extensive and intensive margins for exports to China.
Although the extensive margin for exports to China increased continuously between 1990 and 2005, it began to decrease slightly from 2006 onwards. The number of export goods from Japan, which has expanded through the international division of labour in production process with China, has not increased since 2006 (see Figure 2a).
The intensive margin, by contrast, has continued to increase from 1990 until now (Figure 2b). We do not find a significant difference between actual and predicted values in either the extensive or intensive margins in the case of exports to China. In the trend of the intensive margin of exports to China, we do not observe a sharp drop of the actual value far exceeding the predicted value that appeared for exports to the US.
This suggests that the recession since the financial crisis did not bring about a serious impact on Japanese exports to China.
Figure 2a Extensive margin, exports to US, actual vs. predicted
Source: Authors calculations.
Figure 2b Intensive margin, exports to US, actual vs. predicted
Source: Authors' calculations.
Conclusion: Towards sustainable international trade
Before the financial crisis, Japanese exporters narrowed the range of export goods to the US and specialised in high-end goods – goods with high-income elasticities; products like automobiles and capital goods. During the process, they formed a trade triad between the US, China, and Japan. To China, they have widened the range of goods exported through the development of an international division of labour with China.
This development was market driven and thus probably cost-saving. Nevertheless, it made Japanese exports to the US more vulnerable to sudden demand shifts for high-end durable goods, such as the one we have recently observed. This trend is a major part of the reason that Japanese exports to the US dropped so sharply. In short, the financial-crisis shock to US demand came after a gradual structural change of Japanese exports.
This helps to explain why the average value of Japanese exports to the US behaved so differently from those of China. In the short-run, this matching of Japanese exports to the demand of the US – Japan’s largest export destination – was one of the factors that led to such a serious recession of the Japanese economy following the financial crisis. The flip-side of this is that the ongoing recovery of the US economy should produce a speedy revival of Japanese exports to the US.
From a long-run perspective, we have two features of the world economy to note:
- Global imbalances in the world economy, in particular in the US and China; and
- Significant changes of the world distribution of purchasing power.
It is doubtful that international trade can sustain its rapid expansion in the face of huge trade imbalances, especially the US current account deficit and the Chinese current account surplus.
The share of nations with per capita GDPs ranging from $2,000 to $10,000 has dramatically increased, from 16% in 1996 to 40% in 2006. The high rate of growth in these economies (which includes China and East Asia) is likely to be one of the engines that will help world trade to recover – especially given the US’s huge current account deficit. In short, trade expansion between OECD and middle-income countries is essential for the sustainable development of international trade and the recovery of the world economy. The inclusion of middle-income countries into the talks for multilateral trade liberalization is thus more important than ever before, and the ongoing WTO talks (the Doha Development Agenda, DDA) are a key part of this. The successful completion of DDA, with the full involvement of middle-income countries, must therefore be given the highest priority.
Bernard, A., I. Eaton, B. Jensen, and S. Kortum, 2003, ”Plants and Productivity in International Trade, American Economic Review, 93(4), 1268-1290.
Bernard, A., S. Redding, and P. Schott, 2006, “Multi-product Firms and Product Switching,” NBER Working Paper, 12293.
Bernard, A., S. Redding, and P. Schott, 2007, “Firms in International Trade,” Journal of Economic Perspectives, 21(3), 105-130.
Bernard, A., B. Jensen, S. Redding, and P. Schott, 2009, “The Margins of US Trade,” NBER Working Paper 14662.
Chaney, T, 2008, “Distorted Gravity: the Intensive and Extensive Margins of International Trade,” American Economic Review, 98(4), 1707-1721.
Dean, J, M. Lovely, and J. Mora, 2009, “Decomposing China-Japan-US Trade: Vertical Specialization, Ownership, and Organizational Form,” a paper presented at Conference on The PRC, Japan and the US: Deeper Integration, Asian Development Bank Institute, Tokyo.
Ito, B, E. Tomiura, and R. Wakasugi, 2009, “Offshoring by Japanese Firms: A Comparison of Destinations,” Harvard Asia Quarterly, 12 (1), 14-19.
Melitz, M., 2003, “The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity,” Econometrica, 71, 1695-1725.
Wakasugi, R., B. Ito, and E. Tomiura, 2008, “Offshoring and Trade in East Asia: A Statistical Analysis,” Asian Economic Papers, 7(3), 101-124.
Appendix: Details of the estimates
In our empirical work we define the extensive and intensive margins using the following identity:
Here V is the total exports to partner-i in year-t; and N is the number of HS6 product categories exported to the same partner and in the same year.
The coefficients are estimated on the basis of the gravity equation in which the size of trade partners determines the trade volume. For the estimation, the extensive margin and the intensive margin are dependent variables; GDP, the exchange rate, and institutional factors such as WTO membership are included as the explanatory variables. The estimating equations are:
where Vi,t is the Japanese export to country i in the year t, Ni,t is the number of exported goods from Japan to the country i which are counted on the basis of HS 6 digits product categories, GDPi,t is GDP of country i in the year t, GDPJ,t is the GDP of Japan in the year t, EXi,t is the nominal exchange rate expressed by yen per the currency of country i in the year t , WTO dummy captures China’s WTO accession with it equally one if China is the member of WTO in the year t. As the estimation is on time series data, the country-pair specific factors such as language and distance are subsumed in the constant term. The estimation is conducted for Japanese exports to the US and China for 18 years (1990 to 2007).3
Exports to the US
The results are shown in Table A1. The first column shows the results for the standard gravity equation where the dependent variable is the value of export. The sign and size of the coefficients are roughly in line with priors. (Note that since we use only bilateral trade between the US and Japan, we cannot estimate the impact of distance.)
The second column shows our results for the extensive margin. Interestingly, the income elasticity is negative for the extensive margin (i.e. number of products exported) while the income elasticity for the intensive margin (average export per product) is positive. The negative income effect for the extensive margin of Japanese exports to the US is somewhat unexpected. It probably reflects development of the triad trade noted above whereby the number of exported goods from Japan decreased along with the increase of the US demand during the years from 1990 to 2007.4 In line with priors, the income elasticity to the intensive margin shows a large and positive coefficient.
Taken together, these results suggest that the export strategy of Japanese firms has been to concentrate on a narrowing range of product when it comes to their exports to the US. As for the effects of exchange rate changes on extensive and intensive margins, we find that the yen depreciation has a positive effect to increase both extensive and intensive margins at a high statistical significance.
Table A1 Extensive & intensive margins, Japan to US, 1990-2007
Export to China
Our results for bilateral exports to China show a very different trend; both the extensive and intensive margins increased with Chinese income (Table A2). That is, the Japanese exporters increased the number of exported goods and also raised the average export value according to the expansion of the market demand in China. It is notable that the yen depreciation vis-à-vis the yuan also has a significantly positive effect on the increase of intensive margin.
Table A2 Extensive & intensive margins, Japan to China, 1990-2007
Note: The figures in parenthesis are t-statistics; * and ** indicate 5% and 1% significance.
2 Also see recent theoretical and empirical studies of international trade including Melitz (2003), Bernard, Eaton, Jensen, and Kortum (2003), Bernard, Redding, and Schott (2006) and Chaney (2008) highlighted the number of exported goods and the number of destination countries. Bernard, Jensen, Redding, and Schott (2007) presented the intensive margin of average import or export value per firm-products increased.
3 For collection of data and estimation in this section, I acknowledge Tomoyuki Iida, a graduate student of Keio University, for his excellent assistance.
4 The notion is strengthen by the estimate for the Japan-China extensive margin over the same period (Table 4 second column).