Though Japan has not suffered greatly from a housing collapse or toxic assets, its economy has been – in some respects – hit harder by the global crisis than the US or EU, as several analysts have noted (e.g. Buiter 2009). Japan’s GDP contraction in the fourth quarter of 2008 was nearly double that of the US (12.1% vs. 6.3%).1
Japan’s contraction is almost entirely due to a steep fall in external demand – its drop in net exports accounts for 11.8 percentage points of the decline. In contrast, the US experienced only a decrease in net exports responsible for only 0.15 percentage points of the contraction, meaning that its 6.3% contraction in GDP was almost entirely due to decreased domestic demand. Why did Japan experience such an exceptionally serious decline in overseas demand?
Contraction in triangular trade
Until the economic crisis, there was vigorous “triangular trade”, in which the advanced economies of Asia, such as Japan, South Korea, and Taiwan, exported key components to developing countries, such as China, Thailand, or Vietnam, which assembled and exported the final products to the US (and Europe) in return for US Treasuries. The Asian developing countries specialised in relatively low value-added assembly and manufacturing processes, while the advanced countries concentrated on high-value added processes, such as the production of key components. However, the sudden decrease in US imports has very rapidly contracted this triangular trade. US goods imports fell during the last quarter of 2008 at an annual rate of 19.6%. And as developing countries’ exports to the US decreased, so did their imports from Asia’s advanced economies. Japan, South Korea, and Taiwan thus suffered a significant decline in exports.
A shift in the global demand structure
A large share of Japan’s exports consists of capital and durable consumer goods, such as cars, electric machinery, machine tools, and their components. It is very likely that investment in plant and equipment, as a consequence of what is known as the “acceleration principle,” suffered a particularly steep decline due to the global downturn. Simultaneously, households have been holding back purchases of durable consumer goods due to liquidity constraints and diminished expectations. Consumer demand has shifted from high-end items to inexpensive goods, as illustrated by the boost in sales experienced by firms such as Wal-Mart and Uniqlo. But Japan specialises in the production of high-end items (Fukao et al. 2003), so the disruption and shift in global demand appears to have hurt Japan’s exports particularly badly.
A lag in the timing of the recession
Trade flows brought the global recession to Japan relatively late, but – due to this lag – with greater severity. Until the fourth quarter of 2008, Japan’s exports had remained largely unchanged. When they did drop, the negative impact on GDP was particularly large. However, this also means that from the first quarter of 2009 onward, the negative impact of external demand on GDP growth is likely to diminish.
The impact of yen appreciation
As the global economic crisis escalated, the yen appreciated rapidly. The yen was at 106% of its 2007 average in September 2008 and reached 136% by January 2009. While this appreciation lowered net exports, such an effect typically operates with quite a lag, so it is unlikely that it greatly contributed to the fourth quarter export fall.
Examining the contraction of triangular trade
The third and fourth sources of contraction are well known mechanisms in open economy macroeconomics. However, the first two points concern new phenomena worthy of attention. Here, we attempt to estimate the impact of the contraction in triangular trade using the “Asian International Input-Output Tables” (“Asian I-O Tables”) of the Institute of Developing Economies.
The Asian I-O Tables have been compiled for the analysis of inter-industry linkages through trade in the ten economies of Japan, China, South Korea, Taiwan, Singapore, Thailand, Malaysia, Indonesia, the Philippines, and the US. Products of the same industry are treated as different goods if they are produced by different countries, and from these tables, it is possible to determine, for example, the input in China’s automotive industry of metal products made in Japan or South Korea.
In our calculation, we make the Keynesian assumption that final demand and its derived demand determine the level of production in each country. Based on this assumption, we can estimate – using the Asian I-O Tables – the effect of a decrease in triangular trade, for instance, a situation where a decline in US final demand brings about a decline in production in China and Vietnam, which in turn results in a decrease in Japan’s intermediate goods exports.
For the exercise here, we estimate the change in trade and production in response to a 1% decrease in US final demand for the products of all ten countries, including the US itself. The final demand of the nine other countries is assumed to remain unchanged. We also suppose that the value of trade of the ten countries with the rest of the world remains unchanged.2 Unfortunately, the latest available edition of the Asian I-O Tables is from 2000 and therefore somewhat dated. While this cannot be helped, we did use the Trade Industry Database published by the Research Institute of Economy, Trade and Industry and statistics from the US Bureau of Economic Analysis for 2007 in order to replicate as closely as possible the structure of US final demand by producer country and industry just before the global economic crisis.
The decrease in US final demand decreases Asia’s exports of final goods, decreases US intermediate input demand and thus Asia’s exports of intermediate inputs, and in turn decreases Asian production, thus reducing the trade of intermediate inputs amongst the Asian economies.
Figures 1 and 2 depict the impact of a 1% decrease in US final demand on net exports of intermediate goods by Japan and China, respectively.
These figures clearly show the effects of a contraction in triangular trade. In the case of Japan, the largest fall in net exports in intermediate goods is observed in exports to the US, but net exports of intermediate goods to the Asian countries, including China, also decrease. On the other hand, in the case of China, net exports of intermediate goods to the US also fall substantially, but because China’s imports of intermediate goods from other Asian countries, including Japan, decline3, net exports to these countries actually increase. This comparison shows that the impact of a contraction in triangular trade on Japan due to a decrease in US demand is greater than the drop in exports to the US alone, while the opposite is the case for China.
Next, Figure 3 shows the change in net exports by partner country for Japan and China including trade in both intermediate and final goods. The contraction in net exports of intermediate goods to the US is much greater for Japan than for China, simply because the value of Japan’s exports of intermediate goods to the US are much greater and because of the impact of the contraction in triangular trade. But because China exports considerably more final goods to the US than Japan does, it faces a much greater decrease in total net exports than Japan.
Finally, Figures 4 and 5 compare the impact on gross output and GDP, respectively, for the Asian countries. Reflecting the large decrease in exports, the drop in China’s gross output is 1.7 times as large as Japan’s. However, the difference in the impact on GDP is considerably smaller; the decrease in China’s GDP is only 1.3 times as large as Japan’s. This situation is also the result of triangular trade. China is relatively specialised in the processing and assembly of imported intermediate goods using cheap labour, and the products are then exported to the US. However, the value added ratio of this kind of production is typically low, and it is for this reason that the decrease in GDP is smaller than the decrease in gross output. Figure 5 also shows that, after China and Japan, the countries facing the greatest contraction in GDP are South Korea and Taiwan.
Using the “Asian International Input-Output Tables” of the Institute of Developing Economies, we estimated the impact of the contraction in triangular trade. Our results clearly show the effects of a contraction in triangular trade. In the case of Japan, not only net exports to the US but also net exports of intermediate goods to Asian countries decrease. China, on the other hand, experiences an increase in net exports to other Asian countries, due to its imports of intermediate goods. We also found that since the value added ratio of assembling activities in China is low, the decrease in GDP is much smaller than the decrease in gross output.
Since the deepening vertical division of labour is a process witnessed not only in East Asia but also in many other regions, such as Europe, our findings highlight the importance of estimating world I-O tables to analyse the international transmission of business fluctuations. From this viewpoint, the recently launched WIOD (World Input-Output Database) Project funded by the European Commission is a timely undertaking.
It is important to note that the discussion here considers only the effects of a contraction in triangular trade. Comparing actual US imports by partner country (on a US dollar basis) in 2007 and 2008, we find that imports from Japan fell by 19%, while those from China decreased only by 5%. This means that in order to sufficiently understand the large impact of the downturn in the US and worldwide on Japanese exports, it is necessary to also take into account other factors such as global demand for investment goods and durable consumer goods and a possible shift away from high-end goods.
Yet another aspect is that multinational corporations are the driving force behind triangular trade. They account for about half of China’s exports to the US and tend to be especially dependent on imports of key components. It is possible that the contraction in triangular trade has an especially large impact on Japan and Korea because it particularly affects multinational corporations’ production. Unfortunately, this is an issue that it is difficult to examine with the Asian I-O Tables, as these do not provide information that would make it possible to distinguish between multinational corporations and local independent firms in the analysis of inter-industry linkages.
Editor-in-Chief’s note: The Japanese version of this column appeared in Hitotsubashi University's Hi-Stat Vox No. 8.
1. Seasonally adjusted annualised rates. Second preliminary estimate for Japan; final estimate for the US.
2. The framework is based on Fukao and Yuan (2007), where we used the Asian I-O Tables to examine triangular trade, except that for our purposes here, we consider the impact of a drop in U.S. final demand.
3. Considered here, for the time being, is only the effect of the drop in net exports of intermediate goods to the United States. However, as will be discussed later, an additional contributing factor is the large drop in production in China through the decline in Chinese exports for US final demand use, which further reduces Chinese imports of intermediate goods.
Buiter, Willem (2009). “Green shoots: Grounds for cautious pessimism”, VoxEU.org, 29 April 2009
Fukao, and Tangjun Yuan (2007) “Is China Gaining from Triangular Trade? An Analysis Based on Asian International Input-Output Tables,” in Yosuke Noda and Masato Kuroko (eds.), Trade-Related Indices and Trade Structure, IDE Statistical Data Series No. 91, Institute of Developing Economies (in Japanese).
Fukao, Kyoji, Hikari Ishido and Keiko Ito (2003) “Vertical Intra-Industry Trade and Foreign Direct Investment in East Asia,” Journal of the Japanese and International Economies, Vol. 17, December 2003, pp. 468-506.