تولید در خارج و ارتقای مهارت توسط شرکت های تولیدی ژاپنی

We investigate the influence of offshore production by Japanese multinationals on domestic skill intensity. Identifying relationships based on within variation in a panel of 1070 firms, we find that additional foreign affiliate employment in low-income countries raises skill intensity. The positive effect of FDI on domestic skill intensity, however, diminishes as investment shifts towards high-income countries. Increases in affiliate employment in low-income countries also raise a firm’s reliance on finished goods purchases, suggesting that overseas employment affects domestic skill intensity because imports of final goods from foreign affiliates displace domestic production.

Over the past three decades, Japanese multinational enterprises (MNEs) have steadily increased their offshore manufacturing presence. They now possess substantial production capabilities abroad. Time magazine (April 22, 1996, p. 60) reports, ‘Currently, for example, Japan imports 23 times as many television sets as it exports. They are all assembled in Japanese-owned factories in places like Malaysia and Thailand…By the year 1998, Toyota expects that 65% of the cars it sells around the world will be made outside Japan’. Over the corresponding period, Japanese firms have also increased the share of the wage bill attributable to nonproduction workers, suggesting a demand shift away from workers with low skills. This paper investigates the relationship between the international production strategies of Japanese MNEs and skill upgrading in Japanese firms. We employ a 25-year panel data set for over 1000 Japanese manufacturing firms to investigate the effects of increases in foreign employment on the skill intensity of the domestic workforce. We use the nonproduction worker share of the wage bill and the firm-level average wage as proxies for skill intensity. The analysis explores differential effects on domestic skill intensity of investment in countries with different levels of per capita income. We also examine the electronics industry as a special case of extensive foreign production. We develop predictions about how FDI affects skill intensity of the parent firm at home based primarily on alternative depictions of MNEs described in the series of articles written by Markusen, Venables, and co-authors.1Markusen and Maskus (1999, p. 1) state that horizontal MNEs ‘replicate roughly the same activities in many locations’. Vertical MNEs, in contrast, ‘geographically fragment production into stages, typically on the basis of factor intensities, locating skilled-labor intensive activities in skill abundant countries and so forth’. FDI may have different effects on skill intensity at home depending on the type of investment and the income level of the host country. The relationship we observe between FDI and skill intensity will indicate whether horizontal or vertical investment characterizes our data. The paper contributes to a recent literature investigating the influence of globalization on the demand for skilled labour. Slaughter (2000) and Feenstra and Hanson, 1996a and Feenstra and Hanson, 1996b relate the nonproduction worker share of the wage bill of US industries to international activities. Slaughter demonstrates that the US industry data provide no support for a positive relationship between MNE activities and skill upgrading over the 1977–1994 period. Feenstra and Hanson, however, find that foreign outsourcing (defined by them as the substitution of imported inputs and finished goods for domestically produced goods) can account for 18.9–21.3% of the observed increase in the nonproduction worker share of the wage bill for their sample of 4-digit SIC industries in 1979–1990.2Feenstra and Hanson (1999) obtain somewhat stronger results. They use 4-digit industry data on the prices and quantities of outputs and inputs to relate estimated changes in factor prices to foreign outsourcing and the high-technology capital stock. They find that outsourcing can account for as much as three-fifths of the observed 0.59 annual increase in the nonproduction/production relative wage observed over 1979–1990. Our study adds to the existing literature in four ways. Firstly, our 25-year sample period includes the transformation of a number of Japanese firms into major multinational enterprises. Thus, by utilizing the large firm-level variation in our data we are able to identify a strong relationship between overseas employment and the use of nonproduction workers at home. Secondly, we find that this result depends on the income per capita of the host country: skill upgrading associated with foreign investment is most evident in low-income countries. Thirdly, the relationship between offshore employment and skill intensity is robust to using changes in firm-level average wages as an alternative proxy for skill upgrading. Finally, our results show that employment in low-income countries exhibits a strong positive correlation with parent firms’ purchases of finished goods. This information is consistent with the view that Japanese multinationals are outsourcing labour-intensive goods from their affiliates in low-income countries. In the next section, we describe alternative types of FDI and how their effect on skill upgrading may vary across host countries. Section 3 presents the Japanese firm-level data we employ in our analysis. The following section introduces our econometric specification and presents the estimation results. The concluding section summarizes how well the data support the alternative depictions of FDI and comments on the extent to which FDI has contributed to observed skill upgrading within firms in Japan.

The dramatic growth of overseas production by Japanese MNEs has the potential to explain the observed rise in the skill intensity of Japanese manufacturing. We use a large panel set of Japanese manufacturing firms to investigate the effects of offshore employment on skill composition in Japan. A fairly consistent set of results emerges across different specifications and samples. The rise in the share of nonproduction workers and the average wage can be explained in part by investment in production facilities in low-income countries. We also find that increases in scale and capital intensity tend to raise the production workers’ share of the wage bill. Our findings differ markedly from those of Slaughter (2000) who finds that the nonproduction worker share of wages in the US has an insignificant relationship with MNE activity and a positive one to scale and capital intensity. In regressions similar to his using first-differenced, industry-level data, we do not find significant FDI effects, although scale and capital intensity continue to be negative and significant. This suggests that one source of the difference between the two FDI estimates is our use of firm-level data. We note, in addition, that we consider production employment whereas Slaughter evaluates activity of all affiliates including distributors. This probably results in stronger skill upgrading effects. Our empirical results provide evidence consistent with vertical specialization. FDI in low-income countries appears to raise skill intensity at home but this effect falls as investment shifts towards high-income countries. For high enough host-country income levels, FDI can result in skill downgrading. This is consistent with low-skill activities being transferred to low-income countries and high-skill activities to high income countries. The positive relationship between investment in low-income countries and purchases of finished goods suggests that Japanese affiliates in these countries are assembling goods that are shipped back to the parent firm at home.