w PS /c/ 3 /' POLICY RESEARCH WORKING PAPER 1634 Japanese Multinationals The intangible assets conducive to foreign in Asia investment derive not from research but from marketing networks and production Capabilities and Motivations skills. The factors most affecting foreign investment Susmita Dasgupta in Asia include the human Ashoka Mody capital of the hosts, the Sarbajit Sinha behavior of competitors, and whether earnings can Le repatriated. The greater and more diverse the investm-nent, the greater the interest in low wages. The World Bank East Asia and Pacific Regional Office Office of the Regional Vice President August 1996 I POLICY RESEARCH WORKING IPAPFR 1634 Summary findings Using a specially designed survey, Dasgupta, Mody, and oriented, suggesting that Asian investment is not driven Sinha identify the characteristics of Japanese firms likely by trade barriers, unlike investment in the United States to invest worldwide and in key Asian countties and and Europe. country characteristics associated with Japanese Firms investing in Asia look to the human capital of investment in Asia. the hosts, though interest in low wages expands the Investment abroad is related negatively to research and greater and more diverse the investment. development (R&D) undertaken but positively to export The behavior of competitors is a strong guide to the propensity, indicating that the intangible assets direction of investment. conducive to foreign investment derive not from research Whether earnings can be repatriated is the factor most but from marketing networks and production skills. conducive to foreign investment, although its importance Among the foreign investors, those investing in Asia declines as a firm gains experience in a country. are less prone to do R&D. Tiiey are also less export- This paper - a product of the Office of the Regional Vice President, East Asia and Pacific Regional Office - is based on a survey conducted through the Japanese Ministry of International Trade and Industry in March 1993. Copies of this paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Susmita Dasgupta, room N10-007, telephone 202-473-2679, fax 202-522-3230, lnternet address sdasgupta@worldbank.org. August 1996. (37 pages). The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors anid shouldk be used and cited accordingly. The findings, interpretations, and conclusions are the authors' own and should not be attributed to the W%orld Bank, its Executive Board of Directors, or any of its member countries. Produced by the Policy Research Dissemination Center JAPANESE MULTINATIONALS IN ASIA: CAPABILITIES AND MOTIVATIONS Susmita Dasgupta, Ashoka Mody, and Sarbajit Sinha The World Bank Washington D.C. This paper is based on a survey conducted through the Japanese Ministry of International Trade and Industry in March 1993. We are grateful to Osamu Kawaguchi for his help in designing and implementing the survey and to Vinod Thomas and Ramagopal Agarwala for suggesting that we undertake this endevour. I Introduction Using a specially designed survey, this paper identifies: (a) the characteristics of Japanese firms likely to undertake foreign investments-worldwide and in key Asian countries; and (b) country characteristics associated with Japanese investments in Asia. The survey-based, firm- level analysis here contrasts with the macro approach that correlates country characteristics with aggregate flows into that country (see Wheeler and Mody 1992). It contrasts also with the industry-level analysis where industry characteristics such as the degree of competition, the level of entry barriers, and the degree of technological sophistication are examined as determinants of foreign investment (e.g., Kogut and Chang 1991). The firm-level perspective is important because much of the theorizing on decision to invest abroad derive from the interplay of firm capacity and motives to invest abroad (see Caves 1982 for a literature review). This paper confirms certain findings obtained from the more aggregative studies, but highlights the significant importance of firm characteristics (or capabilities) in the decision to undertake foreign investment. The data permits us to distinguish between the capabilities of firms-proxied by their size, export propensity, and research and development (R&D) expenditures-and their motives for investing abroad. The motives, in turn, arise from operating cost conditions in Japan (due for example to increasing wage costs or an appreciating yen) and from the pursuit of lower production costs and investment incentives offered by host countries. The procedure followed is to examine the various motivating factors after controlling for firm characteristics; in certain instances, where interesting results were obtained, we also present the interactions between capabilities and motivations. 2 In studying motives, survey-based analyses are typically limited to determining the subjective preferences of investors ("do you value low wages? or "does country X have a favorable foreign investment policy?"). Such analyses are intended to elicit the priority accorded by investors to the cost and policy characteristics of alternative investment locations. While the subjective perceptions are valuable, it is also necessary to determine if firms in fact act according to the priorities they state in such surveys, i.e., do they put their money where they say they would? In this project, we did ask firms to rank the factors they considered important in their decisions to invest outside Japan and in specific Asian countries. However, we also asked them for their: (1) share of foreign investment in total investment; (2) share of foreign investment undertaken in Asia; and (3) their likelihood of investing in specific Asian countries in the following three years. Thus the information available allows us both to rank their stated preferences and to conduct an econometric analysis that identifies partial correlations between investments undertaken and firm and (perceived) country characteristics. Analysis of the actual investments undertaken help not merely in validating the stated preferences but also in drawing more complete explanations of investment behavior. High costs of Japanese labor exert a general push towards investing abroad. However, we find that the variation in the investment undertaken depends much more on differences in firm attributes. Also, our respondents do state a strong preference for low wage locations. But perception of low wages in specific Asian countries do not seem to have been a determining factor thus far in deternining investment levels in those countries-in fact, low wages have actually been a disincentive where they were associated with low labor quality. The new Japanese investment 3 that is expected to flow into Asian economies is being guided more by low wages, though considerations of labor quality continue to be very important, confirming the importance of developing country domestic human capital in attracting foreign investment (Lucas 1990). Another important finding of the paper is the importance accorded by investors to actions they perceive are being taken by other investors. Strategic rivalry, or the importance of staking out early positions in growing markets, clearly is a key influence in the foreign investment decision-tnaking process, creating the possibility of cascading effects (Vernon 1993). We begin in the next section by describing the firms in this survey. Next, we examine the firm characteristics and Japanese cost conditions leading to investment out of Japan, irrespective of the country of destination. This is followed by an analysis of the past investment decisions in Asia. Finally, we describe the expected flow of Japanese investment and its distribution within Asia, where we focus on China, Malaysia, Thailand, Indonesia, the Philippines, India, and Vietnam. 4 The investors: some descriptive statistics The survey questionnaire, designed by the authors, was mailed by the Japanese Ministry of Trade and Industry (MITI) to several hundred Japanese firms of which 173 returned usable responses in March 1993. The sample thus obtained cannot be treated as representative of all Japanese firms-we do not know the characteristics of firms who did not respond. There is, however, sufficient heterogeneity amongst the respondents to permit a statistical analysis of their foreign investment behavior; also, as described below, the foreign investment pattern of this sample mirrors that of all Japanese manufacturing firms. The firms in our sample are relatively large (table 1). The average annual sales are 330 billion yen (over $3 billion). For the purpose of this analysis, we have found it useful to sort the firms by size and then divide them up into five equal groups. The average size of the largest fifth is 1.4 trillion yen (about $14 billion)-the largest firm is truly large, with sales of $70 billion. The smallest fifth of the firms in our sample has average annual sales of around $40 million-the smallest firm has sales of about $2 million. This is also a set of firms that is prone to making significant foreign investments (table 1)Yl In the three years prior to the survey, over a fifth of their investment was undertaken outside Japan. The very interesting feature, though, of the sample is the strong tendency of the smallest firms to investment abroad: in the three years prior to the survey, about 28 percent of 1' According to a survey conducted by the Japan External Trade Organization (Jetro), one- fifth of production by Japanese firms is located overseas and this share is expected to grow to a third. The shift abroad is often described as the "hollowing" out of Japanese manufacturing. U.S. firms produce 25 percent of their output abroad while German firms are the least international with less than 20 percent of their output produced at foreign locations (Dawkins 1996). 5 their investment was undertaken abroad. The next three larger groups of firms have a lower propensity to invest abroad. Only the largest fifth of the firms invest a higher proportion-35 percent-outside Japan. Thus, the smallest and the largest firms are amongst the trailblazers of Japanese foreign investment. Small and large firms are being driven to invest by different concerns. As may be expected, small firms have low research and development (R&D) ratios (expenditures divided by total sales), with this ratio rising steadily as the size of the firm increases (table 1).' Thus, we infer that smaller firms are technologically less sophisticated than larger firms. Two possibilities exist: smaller firms are principally in low technology sectors or, within particular sectors, small firms perform tasks that are technologically less demanding than those of larger firms. From our survey, we find that there is a slightly greater tendency for larger firms to be in the continuous process sectors, such as food and chemicals. However, all sizes of firms are represented in electrical machinery, general machinery, and transport equipment sectors, where much of the sample is concentrated. The implication is that the smaller firms are generally at the lower end of the product quality range in each sector. As with R&D, the ratio of exports (from Japan) to worldwide sales rises with size of investor. However, the smaller firms are clearly not shy of exports: indeed, our data shows that from their foreign investment locations smaller firms have a higher export propensity than larger firms. As such, it appears, primafacie, that smaller firms have invested abroad in response to To reduce the burden on survey respondents, we did not ask them to specify their exact R&D ratio. Instead, we asked them to state the range of their R&D expenditure to total sales ratio: zero, less than 1 percent, 1 to 3 percent, 3 to 5 percent, and greater than 5 percent. 6 high and rising costs of production in Japan, making it increasingly difficult for them to competitively sell Japanese-made products and inducing a shift to lower cost locations. Thus, given the lower technological sophistication of small firms, they can be expected to focus on low wage locations; however, their export requirements from the foreign locations also suggest that high labor quality is also likely to be of value to them to ensure quality standards in export markets. Which Japanese firms invest abroad and why? To explore further the characteristics of Japanese foreign investors and why they invest abroad, we investigated the determinants of the share of all foreign investment in the firm's total investment (in the three years prior to the survey). The share of foreign investment was regressed on firm characteristics (size, R&D, export propensity) and the firm's perceptions of cost conditions in Japan. We also controlled for sectoral characteristics through the use of sector dummy variables. Three features of the econometric estimation methodology are worth highlighting. First, a number of firms report zero foreign investment. These firms could "desire" a "negative" foreign investment; however, all observed values are censored at zero. Thus the data contain a potential non-linearity at the point where zero foreign investment becomes the preferred objective. Forcing a linear regression through the data would bias the results since it would require that the slopes with respect to the explanatory variables be the same for foreign investment less than and greater than zero, even though they are clearly observed to be different. 7 This requires us to use the so-called "Tobit" regression to prevent bias in estimates (see Maddala 1983). Second, not all firms reported the share of foreign investment in their overall investment. A natural question that arises, therefore, is whether the responses are missing in a random or a systematic manner. If firms that did not respond vary systematically from firms that did, the estimates obtained will be biased when the regression analysis is performed with the reduced number of observations. To guard against such a selection bias, we followed a two-step procedur6. We estimated a "reporting" equation to predict the probability that an enterprise reports its foreign investment share, as requested in the questionnaire. A binary choice (choosing to report or not to report) probit model was estimated for all 173 observations as a function of firm characteristics. From this estimation, a "correction" term was obtained. The correction term, which reflects the firm's features that lead it to report or not to report, was then added to the Tobit regression. The coefficient on this correction term provides a measure of the selection bias. In our estimates, the coefficient on the correction term was always statistically insignificant and hence we cannot reject the null hypothesis that the non-reporting is indeed random and hence there is no selection bias (see Maddala 1983, chapter 9 and Heckman 1979). Finally, no attempt is made to establish or claim causality from the chosen firm characteristics to the share of foreign investment. It is possible that export shares and R&D are influenced by the extent of foreign investment. Any credible attempt at establishing causality would require time-series data on the firm characteristics. As such, the results here should be viewed as partial correlations (i.e., correlations between the share of foreign investment and the individual firm characterisitcs, controlling for other variables included in the regression). This 8 same limitation and interpretation applies to industry level analyses where, for example, an industry's foreign investment and level of R&D may be jointly determined (example, Kogut and Chang 1991). Firm Characteristics. Controlling for other factors, larger Japanese firms have a greater foreign presence (table 2). Note here the difference between the ordinary least squares estimates (OLS) and the Tobit estimates. The size variable is positive but not significant at conventional levels in the OLS estimates, whereas the significance is strong in the Tobit estimates. The relevance of the Tobit analysis here is clear: forcing a linear relationship through observations where foreign investment is zero leads to biased results. (For other variables too, the statistical significance of the results improves when the Tobit method is used.) The size effect, though statistically significant for the Tobit regressions, is not large in terms of magnitude. An increase in firm size (worldwide sales) of 1 trillion yen (or $10 billion) leads to an increase in the foreign investment share by 0.7 percentage points.!' In contrast to firm size, more R&D is associated with a lower foreign investment propensity. Recall that larger the size of firm, the more R&D it undertakes. These results, therefore, imply that while larger firms are more prone to invest abroad, within groups of similar-sized firms the propensity is dampened if firms are engaged in greater R&D (and hence in sophisticated product or process technologies). The result also reflects the relatively high propensity of small firms with little R&D to invest abroad. This effect is stronger than is For the United States, Horst (1972) also finds increased size to be a predictor of the likelihood that a firm will invest abroad; however, he finds a much stronger positive effect than we do here. As the descriptive statistics show, small Japanese firms do undertake significant foreign investment. 9 apparent from the regressions. We asked firms only to provide the range of their R&D ratio. Hence, the R&D variable in the regression is measured with error, biasing its coefficient towards zero. The fact that it is still significantly different from zero, implies a strong effect. This finding-that a firm's foreign investment is inversely related to its R&D to sales ratio-may appear at odds with the result that an industry's foreign investment is typically positively related to the extent of R&D undertaken in the home country (Caves 1982, Cantwell 1989, and Kogut and Chang 1991). Such a relationship is thought to reflect advantages from home technological capability. Three comments are in order. First, consistent with our result, the finding also exists that R&D capability abroad causes investment to flow abroad in search of new technologies and rising home technological capability limits foreign investment (Kogut and Chang 1991). Second, comparing foreign investment and home R&D across industries must be distinguished from an analysis that focuses more heavily on a within-industry analysis. Since we use firm-level data, we capture within-industry variation, suggesting that even where high home R&D and foreign investment are positively related at the industry level, more investment abroad is undertaken by firms that do a smaller amount of research either to search new technologies or to seek lower cost production sites for less sophisticated products. This is a reminder also of the substantial variation in technological sophistication that remains even within narrowly defined industrial sectors. Finally, R&D generates only one element of technological capability (see discussion in Swedenborg 1979, where high R&D Swedish firms are found to prefer exports while those undertaking less R&D invest abroad). Production efficiency based on factory-floor learning may often be a more substantial element of a firm's technology capital. A firm's export share may better reflect its more inclusive technology capacity. 10 Greater export orientation is strongly conducive to foreign investment. A one percentage point increase in the export ratio (exports as a percent of total sales) leads to a 0.7 percentage point increase in the share of foreign investment. As noted in the descriptive statistics (table 1) large firm size is associated with greater export intensity (as it is with more R&D); as such, the result here indicates that within groups of similar-sized firns a greater interest in export markets is associated with greater foreign investment from Japan. The direction of causality is difficult to determine from a single survey: it could be that firms with traditionally high export propensities (and significant technological capabilities) are being forced to invest abroad (due to increased trade barriers) or it could also be that firms undertaking investment abroad export (especially intermediate goods) to their foreign subsidiaries. What conclusions follow on Japanese investor capabilities? A variety of intangible assets are thought to create the advantage for the potential foreign investor-these assets include product or process knowledge, brand names, and marketing or distribution channels. The interesting result from our survey is that R&D does not seem to be the source of asset driving Japanese foreign investment-on the contrary, firms doing greater R&D tend to stay at home. It could be that brand names are the source of Japanese advantage; certainly, the big electronics firms derive much of their market power from their established market positions. However, that does not explain what competitive advantage allows the smaller firms to successfully invest abroad. A likely explanation, supported indirectly by our survey, is that small firms have access to Japanese marketing channels-these channels are used not just to export from the subsidiaries to Japan but also to other parts of the world. Exports constitute about a third of the output from 11 the foreign subsidiary and so create a need for strong export channels. Moreover, these same distribution channels are perhaps of equal (or greater) importance in importing high quality and competitively priced inputs. In addition, while the R&D ratios of the investing firms may be low, the production capabilities of the firms are likely to be quite high. Japanese production and management techniques are widely believed to give them a competitive edge over producers in other parts of the world. These techniques are widely diffused among Japanese firms, even ones that do little R&D of their own.4' Sectoral Distribution. As noted above, the distribution of firms across industrial sector shows a heavy concentration in the areas of electrical and heavy machinery and automobiles. In doing these regression, we added dummy variables for the different sectors to examine if firms in particular sectors have particularly strong tendency to invest abroad. These dummy variables are not reported in the regressions. However, the results are of some interest. The dummy variables show no statistical significance. In other words, once the variables described above are accounted for, sectoral differences do not seem important in driving foreign investment. Specifically, the size, R&D ratio, and export propensity of the firm are more important predictors of its propensity to invest abroad than is its sectoral identity. This, once again, suggests the importance of firm-specific advantages in determining the flow of Japanese foreign investment. For a recent survey of the literature relating Japanese foreign investment to its "organizational prowess", including factory-level practices and long-term inter-firm networks, see Caves (1993). 12 Japanese cost conditions. Is Japanese foreign investment sensitive to cost conditions at home? Here we consider both the stated weight placed on specific cost factors when operating in Japan as well as the influence of these perceptions on the amount of investment undertaken. The question asked was: "How important have the factors listed below been in influencing your decision to invest abroad?" Respondents were asked to place a check mark (/) in any one of the boxes on the scale provided. Not Very Important Important Appreciation of the Yen High labor costs in Japan High capital costs in Japan Note first from table 2 that the relative importance of these Japanese cost factors does not explain the differences in the share of foreign investment. Firms that consider a high yen and high Japanese capital costs as important factors leading to foreign investment do invest more abroad, but the standard errors of the coefficients are high and the statistical significance of the coefficients is therefore low. For labor costs, the standard errors are even higher. Are these results a reflection of low importance accorded by respondents to Japanese cost conditions? A look at the stated perceptions is helpful (table 3). All firms report labor costs to be of, considerable importance in creating an incentive to invest abroad: the average importance (on a scale of 1 to 7) is 5.40, with a high of 5.85 for small firms and about 5.2 for firms in the larger groups. Thus, most Japanese investors think that Japanese labor costs are high and view foreign investment favorably as means of lowering production costs. However, 13 the amount of foreign investment actually undertaken does not correlate with the perceived severity of the labor costs faced in Japan. Similarly, yen appreciation and high capital costs do not appear to seriously influence the extent of foreign investment undertaken. Thus, in the absence of other identifying factors, the conclusion is that while these cost factors act to push investment abroad, the extent of a specific firm's international investment depends more on its capabilities to exploit opportunities in foreign locations. Investment in Asia Our focus now shifts from the extent of foreign investment to the allocation of that investment across alternative regions. Thus the variable we examine is the share of Japanese foreign investment in Asia. Indirect inferences are also possible regarding firm characteristics that lead to investment in other major locations-the United States and Europe. Asia's share of foreign investment by the sample firms is 35 percent. However, there is a strong inverse relationship between the size of a firm and its share of investment in Asia (table 1). The smallest firms, on average, undertake 60 percent of their foreign investment in Asia; the share falls to about 20 percent for the largest firms. Presumably, smaller firms have fewer opportunities for diversification. To examine this relationship more closely, we regress the share of investment in Asia against firm characteristics and the strength of various preferences stated by firms." Once again, as above, we control for bias due to non-reporting by some firms. However, here a Tobit regression was not required since all reported shares of investment in Asia were greater than zero. 14 Firm and Industry Characteristics. The regression results confirm that, all else equal, smaller firms place a larger proportion of their foreign investment in Asia (table 4). Unlike for aggregate investment flows, we do not now see any impact of firm's R&D on its location decision in Asia. It could have been expected that a greater share of investment in Asia is associated with lower R&D intensity and that technologically sophisticated firms invest to a greater degree in the the United States and Europe. However, any effect of technological sophistication appears subsumed by size of investor, since small firms also tend to do less R&D. This speculation draws support from our later discussion on future plans of the surveyed firms where we. find that in coming years large firms are more likely to make investment in Asia than small firms, and where R&D does show up as a negative influence on Asian investment.t' A feature of some interest (and robustness) is the negative relationship between the export propensity of the investing firm and the share of its foreign investment in Asia. This, it will be noted, is the opposite of the relationship between export propensity and the flow of all foreign investment (measured as a share of the firm's aggregate investment). The implication also is that investment outside Asia (mainly in Europe and the United States) is undertaken by firms with high export intensity. A possible interpretation of this finding is that the Japanese firms investing in Europe and the United States have long exported to those destinations and undertook major investments in these countries during the second half of the 1980s and the early 1990s to produce locally under the threat that their exports would be restricted; the survey captures a This result is consistent with the Kogut and Chang (1991) finding that a Japanese industrial sector is more likely to undertake investment in the United States the higher the level of R&D in the industry. Thus, although high R&D firms are in general not likely to undertake foreign investment, if they do so, the investment is more likely to be in the more industrialized nations. 15 snapshot of the tendency of these highly export-oriented firms to increase their production facilities in western industrialized nations. In contrast, firms investing in Asia are not doing so under the threat of barriers to exports in that region. Rather in seeking to participate in the rapid growth elsewhere in Asia, they are also choosing production locations that offer the possibility of efficient production and low costs of inputs. Thus, while perceptions of high Japanese costs do not differentiate foreign investors, the high negative perception of domestic production costs creates the basis for seeking low cost sites, provided (as we shall see) no sacrifice is entailed in production quality. We next explore these production cost and quality features in some detail. Labor costs. While there is a general presumption from the discussion above that high costs of labor in Japan are a major reason for driving Japanese investment abroad and there are reasons to believe that investment in Asia specifically seeks low wage labor, the survey points to a more nuanced analyses of investment flows to Asia. For example, perceptions of Japanese labor costs was not a significant explanatory variable for investment into Asia. Thus, as before, while Japanese firms consider labor costs in Japan to be onerous, there is little variation in this perception and so it has limited ability to explain the variation in foreign investment undertaken. To approach the matter from a different perspective, we also asked how important (on a scale from 1 to 7) were low Asian wages in their decision to invest in Asia. Firms did report that low wages were important, the average response being 5.6. However, though low Asian wages are attractive to all investors, regression results presented in table 4 indicate that the perceptions of Asian wages by Japanese investors also do not differentiate the investment 16 decisions. If anything, it appears that firms potentially attracted by low Asian wages tend actually to place a smaller share of their investment in Asia.2' Rather, firms require high quality labor-not merely cheap labor. Our results show that it is the variations in the demand for high quality labor that influence the investment decision. Firms that think highly of the labor quality in the Asian region are the ones who undertake substantial investment in that region-greater the perception of labor quality in Asia, greater is the share of investment in Asia. The important message from these findings is that while low wages may be desirable, perceptions of labor quality are key to attracting foreign investment. These results support the hypothesis proposed by Robert Lucas (1990) that the lack of complementary human capital inputs lowers the productivity of physical capital in developing countries and hence limits the flow of foreign investment to these countries. However, the measure of human capital is not necessarily the levels of educational attainment in that country. For example, our respondents find Thai labor quality to be higher than in the Philippines although the secondary enrollment rates in the Philippines are much higher. Thus the labor quality of interest to foreign investors is related more to industrial experience rather than to formal educational achievements. This does create the possibility of a self-reinforcing condition where an industrially literate labor force attracts foreign investment and such investment further enhances the quality of the labor force. Those not in this loop are in danger of being excluded from the benefits of international capital flows. 2' Once again, there is a similarity between Japanese and Swedish foreign investors. Swedenborg (1979) reports that Swedish firms are found to invest in high wage locations and interprets this to indicate a preference for high skill labor required for the relatively sophisticated production operations undertaken by Swedish multinationals. 17 Our data does identify differences between firms with regard to their attitude towards low wages. Small firms place a significant premium on low wages. This is seen by the interaction of labor cost perception variable with a dummy for the small 20 percent of the firms. Here the coefficient is positive and significant. Importantly, though small firms also place a premium on labor quality. Capital costs. The regression result shows that the severity of perceived Japanese capital costs correlates strongly with the share of investment undertaken in Asia. The surveyed investors, on average, do not consider the capital cost disadvantage in Japan to be high. However, the perceptions of the firm in this regard vary much more and this variation is, in turn, reflected in the variation in foreign investment undertaken. We find further that perceptions of high capital costs in Japan are positively and significantly (at the 5 percent level) correlated with the goal to raise capital from the country in which the investment is being made. The question asked was: how important (on a scale of 1 to 7) is availability of local financing in your decision to invest in Asia. The correlation between this variable and the severity of perceived capital costs in Japan (also measured on a scale of 1 to 7) was 0.29, which is significant at the 5 percent level. Thus, capital costs (and/or the easy availability of finance) appears to be an important factor determining the choice of investment location. To conclude: we found in the previous section that costs in Japan relative to those prevailing abroad (particularly high Japanese labor costs) play virtually no role in the decision to invest abroad-rather the extent of investment undertaken depends much more on firm-specific attributes, In this section, we have found that relative costs are a much more significant factor 18 in determining the allocation of investment across competing locations. However, costs need to be interpreted here in the broader sense of costs of doing business. Thus, low wage costs do not by themselves constitute low production costs-the quality of labor is critical. Capital costs and availability of local financing are also important considerations in the location decision. These findings closely match the findings from the aggregate data (Mody and Srinivasan 1996). Country Policy Characteristics. In choosing their production locations, what country policies do Japanese investors look for? Limitations on repatriation of earnings is considered a serious disincentive by Japanese investors planning investments in Asia (on a scale of 1 to 7, with 7 being the most severe disincentive ranking, limits on repatriation rank 5.62). Moreover, this perception of disincentive is strongly correlated with a low share of foreign investment going to Asia. Note, however, the absence of any other country policy measure as correlate of the share of Asian investment in aggregate foreign investment. Of particular interest is the absence of ownership restrictions as influencing Asian investments. Once again we find that the raw perceptions and regression results give seemingly different results. However, it is more appropriate to view these as reflecting different perspectives on investor preferences. We do find that' government restrictions on foreign ownership are strongly resented by Japanese investors. Requirements to export are similarly considered a major disincentive: on a scale of 1 (low disincentive) to 7 (high disincentive), restriction of ownership to less than 50 percent of firm equity is 1 rated at 5.7. The requirement to export more than 50 percent of output is rated at 4.9. 19 However, most countries in which the Japanese firms invest have limited or no ownership restrictions on foreign investors. This results in foreign investors owning a large share of the equity of their venture abroad. For example, investors responding to our survey note that, on average, they owned about 45 percent of the venture in the Philippines and 60 percent of the venture in China. For Thailand, Indonesia, and Malaysia, the average share of equity lay in between 50 and 60 percent. Only in Vietnam and India was the share of equity owned on the low side-below 25 percent. One can interpret these findings in the following manner. Japanese firms screen out those countries with the most onerous ownership restrictions and export obligations. In countries where they do invest, they do view ownership restrictions (and local content requirements) as a disincentive if those restrictions are in place; however, for the most part, these are countries with low testrictions or countries in the process of dismantling restrictions. The results also point to a warning. If, for some reason, the restrictions were reintroduced, then firms with the greatest investment in the country would be the most seriously affected, creating the possibility of large investment outflows. The implication also is that concerns about repatriation of earnings are both strong and current-unlike ownership restrictions which are in practice being phased out, repatriation of earnings is not thought to be a concern that can be dismissed as practically unimportant. Thus, firms that view repatriation of earnings as a serious problem in Asia do, in fact, lower their Asian investments. 20 Future plans of Japanese investors in Asia Currently, Malaysia and Thailand are the most favored locations, followed closely by China. Indonesia is next. Philippines, India, and Vietnam have attracted very little Japanese investment. See table 5, which gives the number of firms that have investments in each of the countries. However, as these Japanese firms plan substantial new investments in East Asia, a shift in direction is discernible.9' In this section, we describe the determinants of expected Japanese investments in Asia. A question was asked to determine the likelihood of the firm's investment in each of the seven Asian countries in the three years following the survey. For each country, the firm was asked to rank the likelihood from 1 (very unlikely) to 7 (very likely). Table 5 compares this likelihood with current presence in the countries considered. There is clearly increasing interest in China, Indonesia, and Vietnam. From an already strong position, China emerges as a strikingly popular likely destination. Japanese investors maintain a solid interest in Thailand and Malaysia. India and the Philippines are currently positioned relatively poorly and there are no indications that they will gain significant Japanese investment in the near future. As in previous sections, we present now a more detailed analysis of future investment preferences. Our dependent variable is the likelihood of investing in a specific country in the A similar shift is evident in the aggregate as investments in Europe and the U.S. have grown at a slower pace in recent years while Asia and Latin America have become increasingly favored locations (Dawkins 1996). 21 next three years. This set of investigations also pernits us to examine the effect of country specific attributes on the plans of investors.2' The regression presented in the first column in table 6 assumes that planned investment bears the same relationship to investor characteristics and country features for all host countries. Since this is unlikely to be the case, we also repeat the regression for three country groups: (1) China, Vietnam, and Indonesia; (2) India and the Philippines; and (3) Malaysia and Thailand. Any grouping runs the risk of being arbitrary.. Our reason for this particular division was that the first group is experiencing the most growth in investors; the second group has had low investment and is not attracting much investor attention either; the third group, Malaysia and Thailand, has had significant investment in the past and is continuing to retain strong investor interest. Firmn characteristics. The likelihood of investment in the countries under consideration increases as the size of the firm increases, this being true for all countries and for each of the three groups. This suggests that unlike in the past when small firms were more focussed on Asia, the larger firms are displaying strong interest with the possibility of increased share of their investment in that region. Consistent with previous results, once size is controlled for, then firm R&D has a negative influence on planned investment. Finally, the export ratio of the firm is also negatively correlated with its planned investment in the Asian countries under consideration. This implies a continuation of the past investment pattern: those serving the 2' In this set of regressions, there is no "correction" term for missing observations since almost all firms reported their likelihood of investing in different countries. 22 Japanese market are more actively seeking Asian investment locations than those who have a significant export presence. Role of competitors. A most interesting finding is that foreign investment decisions are very closely related to those that competitors are expected to take. We asked our respondents how likely it was that their competitors would invest in each of the seven countries. The regression results show a strong partial correlation between the firm's plans and its expectations regarding the behavior of its competitors: if a firm expects that its competitors are very likely to invest in China, the firm itself considers it very likely that it would invest in China. Of interest here is the difference between the three country groups. The influence of competitors is most powerful for the China/Indonesia/Vietnam group. The coefficient of 0.33 could crudely be interpreted as indicating that a one percent increase in the likelihood of competitors investing in these countries will have the effect of raising the respondent firm's likelihood of investing in those countries by 0.33 percent. (This elasticity interpretation assumes that the scales represent logarithmic preferences.) At the other extreme, the coefficient for India and the Philippines has a value of 0.16, suggesting that while investors are influenced by the behavior of other investors in these countries, the rush to India and the Philippines is much less. Finally, the coefficient for Malaysia and Thailand is 0.28, which is a strong indication that these countries are not close to saturation in terms of the flow of new investmnent, as some observers are inclined to believe. Our findings here support Raymond Vernon's observations on the importance of strategic considerations in driving foreign investment.1°' IO/ Such rivalry evidently extends beyond Japanese investors. A German investor recently summarized well the phenomenon: "We simply cannot sit back and let the Japanese take over another market unchallenged" (Financial Times, March 28, 1993). It is very likely 23 Host countrEy characteristics. Respondents to our survey reported that their most important market was the domestic market of the country they were investing in, with importance measured on a scale from 1 (not important) to 7 (very important). Here both the stated perceptions and the regression results point in the same direction. The raw score for importance of the domestic market was 5.5 (for all other markets, the score was less than 5.0). Moreover, the pooled regression results, as well as the results for different groups, show that perception of the size of the host country's domestic market exerts a positive influence on planned investment in that country. Availability of parts and equipment is an important factor in guiding future investment, though the results are not significant at conventional levels. The effect appears most pronounced for Malaysia and Thailand, and for the pooled regression. The implication is that investors are sensitive to perceptions of availability of parts and equipment when making investment decisions within Malaysia and Thailand, and when choosing between the three groups. Low labor cost and high labor quality are also important factors. Notice that while labor quality continues to be important, low wages emerges as a strong factor here, in contrast to the case where we explained the share of Asian investment in all foreign investment. The finding could reflect a change from the past when Japanese investment was only in its early stages; with major new investment planned, wages may play a more key role in determining investment locations: However, one difference between the results here and in the previous section is that here we are considering investments in specific countries whereas earlier we were investigating determinants of investments in all these countries. The implication could be that when taken as that Japanese and other investors feel similarly. 24 a group, low wages in these countries are not the main feature attracting Japanese investment, but choice within the group of countries is influenced by perceived wage levels. Finally, favorable FDI policy is a desirable country characteristic. Here we are unable to distinguish between different aspects of FDI policy. Our earlier discussion (in the previous section) was based on generally desirable properties of FDI policy. In this section, our explanatory variables refer to specific countries and attempting to elicit views on specific aspects of FDI policy for every country was determined to be unworkable. However, we did try one variation on the basic regression to further examine the role of FDI policy. We interacted the country perception of FDI policy with the presence or absence of the firm in that country. The result shows that policy has less influence on likelihood of investment if the firm is already present in the country under consideration. The implication is that perception of FDI policy is more relevant as a potential barrier to entry rather than as an impediment to expansion of existing firms that, presumably, have learnt to work within the system. 25 Conclusions and discussion of findings The payoff from this micro-level analysis of Japanese investment has been two-fold. First, we have been able to identify key firm characteristics that influence decisions to invest abroad. Second, we were able to relate stated preferences to actual investments undertaken and to likely future investments. Firm characteristics. We find that firm characteristics have a very significant influence in determining both aggregate foreign investment flows and their allocation across competing locations. In particular, firm characteristics are the dominant discernible influence on the decision to invest abroad. Perceptions of cost conditions in Japan-generally negative, especially with regard to labor costs-may well influence the extent of foreign investment but since the perceptions are held with some uniformity across firms, the ability to distinguish their influence is limited. Thus while there may be a general push based on rising domestic costs, the differential response to this push indicates that those with favorable inherent capabilities are best able to exploit opportunities for establishing production facilities abroad. The precise nature of foreign investment capabilities can only be indirectly inferred from the data. Firm size is important in determining investment abroad-large firms, all else equal, invest more abroad. An important finding was the dampening effect of high R&D on foreign investment. Technologically sophisticated Japanese firms are under less competitive threat than other Japanese firms and hence have less need to move production closer to customers or seek cheaper locations. If not R&D, then what are the firm-specific advantages that enable Japanese firms to be successful foreign investors? An important clue may lie in the strong positive relationship between foreign investment and export propensity. Large size, low R&D, and high 26 export intensity combine to suggest that Japanese foreign investors exploit their access to marketing channels (for international trade) and their superior factory-level production techniques. In considering the allocation of foreign investment to Asia, the results show that in the past smaller firms have had a greater share of their investment in Asia but looking ahead, large firms have stronger expectations about investments in Asia, implying that they could increase the share of their foreign investment in that region. The share of investment in Asia is inversely related to export propensity, suggesting that investments undertaken by Japanese firms outside Asia (especially in Europe and the United States) to leap over trade barriers while, in contrast, Asian investments are being driven more by perceptions of market growth in that region. It is possible that the growing Asian markets could equally be served from production locations in Japan. However, the attraction of Asia is augmented by favorable production conditions in some countries in the region. The race with competitors is indicative of the presssure to establish early presence in the markets to be served. Market Orientation. According to this survey, foreign subsidiaries of Japanese firms export about one-third of their output from their operations in Asia. However, our analysis of the expected trends in the coming few years suggests that the domestic market will continue to be a major attraction for foreign investors in all countries. Indeed, except for exports to Japan from the China-Indonesia-Vietnam group, expectations of exports play no part in explaining the likelihood of a firm making an investment in a particular country. What are the factors that can increase Japanese investor interest in exports from their host locations? The survey shows that firms that export a larger share of their output from their 27 subsidiary in the host country, tend to import a larger share of their inputs. Where imports are being restricted by trade barriers and local content rules, the implication of our survey would be removal of such barriers would permit greater exports. The survey also indicates that exports from host countries depend upon the availability of a pool of trained manpower, specially technicians and supervisors. Human capital development. The survey results highlight the strong Japanese investor preference for operating in conditions where the human capital is well developed. We saw this in different ways: the share of investment in Asia was higher for firms looking for high labor quality and the likelihood of investing in a country in the coming years was also influenced by its labor quality. In addition, the size of operation of the subsidiary (measured by the number of employees) is correlated with the labor quality in that country and, as noted, the share of exports in sales is related to easy availability of technicians and supervisors. This is strong confirmation of the Lucas (1990) hypothesis on the need for complementary human assets for the flow of financial capital. However, it also raises questions about the developmental role of foreign investment. It is widely presumed that foreign investment can be an effective means for bringing "best practice" to a country and for widely disseminating these practices through labor turnover from the foreign enterprises or through training to local suppliers. We were unable to verify the extent to which such diffusion occurs through on-the-job training, Japanese practices -such as job rotation that leads to multiple-skilling, or the imposition of quality standards on suppliers leading to an indirect transmission of knowledge. 28 What we do see, however, is that Japanese investors demand a minimum labor quality-which includes availability of technicians and supervisors and, hence, is a pretty stiff minimum. Where such a high quality environment exists, Japanese production techniques are effective, and lead to further enhancement of skills. Where the relevant skills are not available, the probability of attracting Japanese investors, and the ensuing further development of human skills, is less likely. This, of course, raises the difficult question of how such minimum skills are to be identified and developed. Since perceptions of labor quality are only loosely related (and sometimes quite contrary) to conventional educational measures, measures undertaken to encourage firm-level training would have a higher pay-off than secondary school enrollment. An implication could be to explicitly require foreign investors to undertake significant training programs, as apparently was the case in Singapore. FDI Policy. The findings strongly support investor preferences for favorable FDI policies. For the three groups of countries considered, perception of favorable FDI policy leads to greater investment in the country. There is, of course, a difficult question of causality here. Firms that invest a lot in a particular country are likely to be more familiar with the workings of policy and could view it as less intimidating than firms that have no (or limited presence) and hence limited experience in the country. If this is indeed the case, then the FDI policy would have more of an influence on entry than on continued expansion in the country. The evidence presented suggests there is in fact some basis for thinking so. The implication is that better information and dissemination of information on FDI policy may have dividends, such as through using existing investors more actively to convey their experiences to potential investors at carefully organized symposia. 29 Alternatively, specialized and proactive agencies that "hold the hands" of potential investors could also be a useful instrument in attracting foreign investment. These agencies do not have to be a part of the government-non-governmental organizations and even private enterprises could play an important role. The example of " shelter operations" in northern Mexico is instructive. Despite the favorable maquiladora policies that allowed foreign investors easy- imports of equipment and materials, foreign firms still find themselves hesitant to make the first investment. About two decades ago, a largely private initiative led to the formation of shelters, which provide the foreign firm with all local services (including dealing with the government, hiring workers, and renting space). This has proved to be a lucrative business and has fostered foreign investment in the region. The role of the government was in financing some of the early industrial estates that became the homes for some of the shelters; much industrial estate development since has also been privately financed. FDI policy covers a diverse range of initiatives and it is necessary to peel open its components. Where it was possible to do so, we found that repatriation of earnings was a very serious concern to investors, not just in terms of their perceptions but also in terms of how it influenced their investment decisions. That this should be the case is easily understandable. A clear implication is that greater freedom to transmit earned profits has to be a high policy priority. Other dimensions of FDI policy, although perceived by investors as important, seemed to have less influence on actual investment decisions. Ownership restrictions and export requirements for example were perceived as serious disincentives but had no clear effect on investments undertaken. One explanation we offered is that ownership restrictions are, in practice, not very strong in most East Asian economies. Thus, while investors would view them 30 as disincentives if they were in place, their practical lack of restrictiveness makes them not very relevant in explaining investment decisions. Export requirements and local content rules seem to go together. If export requirements are present, it is important, as discussed above, that local content rules be relaxed-firms that need to export require international quality inputs at international prices. Thus, a policy that imposes export targets without allowing virtually free imports is going to be a serious deterrent to investors. Alternatively, if local content rules are in place, foreign entrepreneurs will have to be allowed to sell a substantial proportion of their output in the domestic market. 31 REFERENCES Cantwell, John. 1989. Technical Innovations in Multinational Corporations. London: Basil Blackwell. Caves, Richard. 1982. Multinational Enterprise and Economic Analysis. New York: Cambridge University Press. Caves, Richard. 1993. "Japanese investment in the United States: Lessons for the Economic analysis of foreign investment." World Economy 16 (3): 279-300. Dawkins, William. 1996. "Japan shifts foreign investment focus." Financial Times February 13: 8. Heckman, James. 1979. "Sample selection bias as a specification error. " Econometrica 47: 153- 161. Horst, Thomas. 1972. "Firm and industry determinants of the decision to invest abroad: an empirical study." Review of Economics and Statistics 54 (August): 258-266. Kogut, Bruce and Sea Jin Chang. 1991. "Technological capabilities and Japanese foreign direct investment in the United States." Review of Economics and Statistics 73 (3): 401-413. Lucas, Robert E. 1990. "Why capital doesn't flow from rich to poor countries?" American Economic Review 80 (2): 92-96. Maddala, G.S. 1983. Limited Dependent and Qualitative Variables in Econometrics. New York: Cambridge University Press. Mody, Ashoka and Krishna Srinivasan. 1996. "U.S. and Japanese invesrors: do they march to the same tune?" Draft. Swedenborg, Birgitta. 1979. "The multinational operations of Swedish firms: an analysis of determinants and effects." Stockholm: The Industrial Institute for Economic and Social Research. Vernon, Raymond. 1993. "Where are multinationals headed?" In Kenneth A. Froot. ed. Foreign Direct Investment. Chicago: The University of Chicago Press. Wheeler, David and Ashoka Mody. 1992. Journal of International Economics 33- 57-76. 32 Table 1: Characteristics of the Sample Firms (Mean values) Worldwide Share of R&D Exports to Investment Sales Foreign Expenditure Worldwide share in (billion Investment to Sales (%) Asia (%) yen) in all Worldwide Investment Sales (%) (%) 0.2 - 8.6 4.1 28.1 1 - 3 8 56.4 (33) (33) (17) (24) (31) (8) 8.7 - 37.8 20.5 15.8 1 - 3 12 65.0 (32) (32) (20) (31) (32) (4) 37.9 - 94.6 63.9 17.3 1 - 3 13 57.0 (32) (32) (18) (30) (32) (10) 94.7 - 300.0 190.0 20.0 3 - 5 13 26.9 (33) (33) (22) (30) (32) (16) 300.1 - 7450.0 1414.7 35.1 3 - 5 17 20.4 (31) (31) (10) (30) (31) (21) ALL FIRMS 329.0 21.9 1 - 3 13 35.2 (161) (161) (94) (155) (158) (62) 1. Numbers of Respondents in parentheses 2. Respondents were asked to give the range of their R&D investment rather than an exact number. 33 Table 2: The decision to invest abroad (Dependent Variable: Share of Foreign Investment in Total Investment, 1990-92) Variable Model 1 Model I Model 2 Model 2 (OLS) (Tobit) (OLS) (Tobit) Worldwide Sales .0006 .0007* .0007 .0008 (1.36) (2.22) (1.53) (2.70) R&D Ratio -6.41** -6.77* -6.59 ** -7.02 ** (-1.96) (3.68) (-1.93) (3.82) Export Ratio 0.73* 0.72* 0.70* 0.70* (3.40) (9.68) (3.16) (8.88) Appreciation of 1.91 2.24 Yen (.84) (.88) High Labor Cost -0.021 -0.06 in Japan (.0087) (-.001) High Capital 1.12 .41 Cost in Japan (.56) (.04) Error Correction -15.97 -13.64 -11.98 -11.56 Term (-0.93) (.54) (-.65) (.36) Constant 44.62* 41.64** 28.88 28.54 (2.18) (3.59) (1.16) (1.21) R-Squared .25 .27 Log likelihood -260.75 -259.66 No. of 62 62 62 62 Observations T-statistics in parentheses for OLS estimates and chi-square statistics for Tobit estimates. * significant at 5%, ** significant at 10% 34 Table 3: Perception of disadvantage due to high Japanese costs (Mean Responses on a 1 - 7 Scale, from favorable to very unfavorable) Firm Size (billion Appreciation of Yen High Labor Cost in High Capital Cost yen) Japan in Japan 0.2 - 8.6 4.52 5.85 4.26 (33) (25) (27) (23) 8.7 - 37.8 4.82 5.55 4.44 (32) (28) (29) (27) 37.9 - 94.6 4.43 5.10 4.10 (32) (30) (30) (30) 94.7 - 300.0 4.90 5.28 4.19 '(33) (31) (32) (32) 300.1 - 7450.0 4.73 5.19 4.57 (31) (30) (31) (30) ALL FIRMS 4.70 5.40 4.29 (161) (153) (159) (151) Number of Respondents in parentheses. 35 Table 4: Determinants of allocation of investment to Asia (Dependent Variable: Asian share of past foreign investment) Variable OLS OLS with Selectivity Correction Worldwide Sales -.0009* -.0009* (-2.14) (-2.19) Export/Worldwide -54.39** -51.54 ** Sales (-1.98) (-1.87) High Capital Cost 7.25* 8.04 * in Japan (2.85) (3.06) Low Labor Cost in -5.53 -6.21 ** Host Country (-1.67) (-1.85) Good Labor Quality 13.14* 12.45 * in Host Country (3.70) (3.47) Restriction on -11.24 * -11.08 * Repatriation of (-3.31) (-3.27) Earnings _ Correction Terrn 3.30 (1.14) Constant 37.38 37.11 (1.33) (1.32) R-squared .49 .50 No. of Observations 50 50 * significant at 5%, ** significant at 10%-t-values in parentheses. 36 Table 5: Characteristics of Likely Future Investors by Country Variable China India Indo- Malaysia Phili- Thai- Vietnam nesia ppines land Number of 162 155 155 155 155 156 158 Firms Likely to Invest in next 3 years Likelihood of 4.10 1.74 2.86 2.87 2.03 3.18 2.55 Future Investment (1-7 Scale) Of those likely 33 5 28 40 11 47 2 to invest, the number who are already present Total Number 34 5 31 43 11 52 2 Currently Present 37 Table 6: Determinants of Expected Investment in Asia (Dependent Variable: Likelihood of Future Investment) Variable ALL China, India and Malaysia and COUNTRIES Indonesia and Philippines Thailand Vietnam Worldwide 0.00003* 0.00003* 0.00005* 0.00002** Sales (4.574) (3.131) (4.246) (1.714) R&D Ratio -.185** -0.229 -0.162 -0.083 (-1.926) (-1.563) (-1.031) (-0.453) Export/Sales -2.108* -2.008** -0.096 -3.911 * (-3.063) (-1.909) (-.088) (-2.864) Investment by 0.285* 0.317* 0.161* 0.308* Competitors (6.847) (5.078) (2.092) (3.987) Domestic 0.179* 0.121* 0.134** 0.157* Market (5.048) (2.436) (1.924) (1.988) Labor Cost 0.195* 0.155* 0.059 0.282* (4.670) (2.212) (0.915) (3.071) Labor Quality 0.079** 0.046 0.056 0.079 (1.703) (0.681) (0.711) (0.791) Equipment and 0.007 -0.072 0.126 0.135 Parts (0.13) (-0.931) (1.228) (1.373) FDI Policy 0. 121 * 0.280* 0.077 -0.144 (2.132) (3.511) (0.704) (-1.271) Constant -0.463 0.171 0.057 -0.191 (-0.997) (0.225) (0.081) (-0.204) Adjusted R 0.38 0.38 0.36 0.29 Squared * significant at 5%, ** at 10%, t- values in parentheses Policy Research Working Paper Series Contact Title Author Date for paper WPS1611 Economic Analysis for iealth Jeffrey S HammTer May 1996 C. Bernardo Projects 37699 WPS1612 Stock Market and Investment The Cherian Samuel May 1996 C. Samuel Signaling Role of the MarKet 30802 WPS1613 Does Public Capital Crown Out Luis Serven May 1996 E. Khine Private Capital? Evidence from India 37471 WPS1614 Growth, Globalization, and Gains Thonmas W Hertel May 1996 A. Kitson-Walters from the Uruguay Round Christian F Bach 323947 Betina Dimaranan VVill Martin WPS1615 Issues in Measuring and Modeling Martin Ravallion June 1996 P. Sader Poverty 33902 WPS1616 Transient Poverty in Rural China Jyotsna Jalar June 1996 P. Sader Martin Ravall!on 33902 WPS1617 Why is Unemployment l ow in the Simon Commander June 1996 L. Alsegaf Former Soviet Union? Enterprise Andie, Tolstopiatenko 36442 Restructuring and the Structure of Compensation WPS1618 Analytical Aspects of the Debt Stijn Claessens June 1996 R. Velasquez Problems of Heavily Indebted Enri^a Detragiache 39290 Poor Countries Ravi Kanbur Petei Wickham WPS1619 Capital Flows to Latin America Sara Calvo June 1996 M Gomez Is There Evidence of Contagion Carmen Reinhart 38451 Effects? WPS1620 Bank Insolvencies: Cross-country Gerard Caprio Jr July 1996 B. Moore Experience 38526 WPS1621 The Sustainability of African Debt Daniel Cohen July 1996 S. King-Watson 33730 WPS1622 Capital Control Liberalization dnd Ross Levine July 196 P Sintim-Aboagye Stock Market Development Sara Zervos 38526 WPS1623 Environmental Degradation and Deon Filmer July 1996 S. Fallon the Demand for Children Searching I ant Pritchett 38009 for the Vicious Circle WPS1624 Structural Adjustment, Ownership Luca Barbone July 1996 C. Pelegrin Transformation, and Size in Polish Domenico Marchetti Jr 85067 Industry Stefano Paternostro Policy Research Working Paper Series Contact Title Author Date for paper WPS1625 Restructuring and Taxation in Simon Commander July 1996 L. Alsegaf Transition Economies Andrei Tolstopiatenko 36442 WPS1626 Partners or Predators? The Impact Jeffrey D. Lewis July 1996 N. Mensah of Regional Trade Liberalization on Sherman Robinson Q4-058 Indonesia WPS1627 Tradable Water Rights: A Property Paul Holden July 1996 P. Mendez Rights Approach to Resolving Water Mateen Thobani 38893 Shortages and Promoting Investment WPS1628 Globalization A New Role for Shigeru Otsubo July 1996 J Queen Developing Countries in an Integrating 33740 World WPS1629 Form of Ownership and Financial Fabio Schiantarelli July 1996 P. Sintim-Aboagye Constraints Alessandro Sembenelli 38526 WPS1630 Water Pollution Abatement by Susmita Dasgupta August 1996 S Dasgupta Chinese Industry. Cost Estimates Maiinul Huq 32679 and Policy Implications David Wheeler Chonghua Zhang WPS1631 Bank Regulation and the Network Patrick Honohan August 1996 P Infante Paradigm Policy Implications for Dimitri Vittas 37642 Developing and Transition Economies WPS1632 Evaluating Bolivia's Choices for Sarath Rajapatirana August 1996 L. Schunk Trade Integration 31779 WPS1633 Essentials for Sustainable Urban Jorge M Rebelo August 1996 A. Turner Transport in Brazil's Large 30933 Metropolitan Areas WPS1634 Japanese Multinationals in Asia: Susmita Dasgupta August 1996 S. Dasgupta Capabilities and Motivations Ashoka Mody 32679 Sarbajit Sinha