Explaining industrial growth in coastal China : economic reforms...and what else?

I 5 142 PSD Occasional Paper No. 2 July 1995 Explaining Industrial Growth in Coastal China Economic Reforms.. .and What Else? Ashoka Mody and Fang-Yi Wang I B I The World Bank Private Sector Development Department   Private Sector Development Department Occasional Paper No.2 Explaining Industrial Growth in Coastal China: Economic Reforms ... and What Else? The World Bank Ashoka Mody and Fang-Yi Wang July 1995 Elinor Berg, Michael Klein, Anjali Kumar, Jenny Lanjouw, Bart Verspagen, and especially Edward Glaeser and Paul Romer provided many helpful comments. The authors are grateful to Shahid Yusuf for his comments as well as help with initial financing and organization of this project. The project has been partly financed by the World Bank Research Conmnittee (RPO 677-50). The views expressed are those of the authors and should not be attributed to the World Bank or any of its affiliates. The World Bank Private Sector Development Department   Contents Abstract .......................,. ................... v 1 Introduction ......1,,,. . .... l 2 A Decomposition of Output Growth .5 3 Investigating the Correlates of Growth ..........................9., . ........,., 9 4 Correlates of Growth ..............,. ... . . . .. 13 5 Conclusions ............ ,,..... 23 References ..................... ....... 27 Appendix .. 31 Tables & Figure ..33 . .   Abstract In the 1980s, China experienced "an explosion of pent-up entrepreneurship" facilitated by wide-ranging, though often unorthodox, economic reforms. We use data on the output of 23 industrial sectors in seven coastal provinces/counties to study the correlates of growth. Reforms allowed greater play of economic incentives and increased the autonomy of state-owned- enterprises. Despite considerable market liberalization, however, the operation of market forces remains substantially circumscribed. To complement reforms for increasing allocative efficiency, China has pursued a long-term strategy for encouraging investments by specific new entrepreneurs. The "open door" policies and "Special Economic Zones" have successfully attracted investments from overseas Chinese to the south-eastern coast. Existing strengths and capabilities, especially human capital and infrastructure, also contributed to growth. Our results, moreover, illuminate the interplay between reforms and conditions conducive for growth for example, the power of foreign expertise is greatly enhanced by available human capital. Finally, China made judicious use of the advantages of backwardness by targeting areas that were less developed and less encumbered by the legacy of existing institutions. v   1 Introduction In the 1980s, China experienced "an explosion of pent-up entrepreneurship" facilitated by wide-ranging, though often unorthodox, economic reforms. 1 Growth in gross domestic product (GDP) jumped from 6.4 percent a year between 1965 and 1980 to 10.1 percent between 1980 and 1989. From 1985 to 1989, the years on which we focus, the pace of economic reforms was stepped up and performance was especially outstanding: GDP grew at 11.5 percent a year and industrial output, the principal engine of growth, grew at a yearly rate of 14.4 percent. Growth during the 1980s was not merely a consequence of more output to meet ambitious (and often wasteful) plan targets. Jefferson, Rawski, and Zheng (1990) find that total factor productivity (TFP), growing at an annual rate of 2.4 percent for state-owned enterprises and 4.6 percent for collectively-owned enterprises, accounted for 27 percent of output growth between 1980 and 1988. Comparing 1980-84 with 1984-88, TFP's contribution to output growth rose in both state-owned and collectively-owned enterprises, suggesting an increase in the role of technical change during the 1980s. In contrast, TFP made virtually no contribution to growth in the three decades before 1980 (Chow 1993). At the same time China became a substantial exporter of manufactured goods. Although exports had been increasing since the mid-1970s, China's share of world markets jumped dramatically between 1985 and 1989, particularly (but not exclusively) in light manufactured goods, such as textile yarn and fibers, travel goods and handbags, and clothing and accessories. Gains in industrial output were especially marked in the coastal region, where output growth during 1985-89 was significantly higher than that of other regions and was also substantially above its own growth rate in the previous five years (Table 1). Five coastal provinces (Guangdong, Fujian, Shandong, Jiangsu, and Zhejiang) were at the center of the "omiracle," registering growth rates of about 20 percent a year between 1985 and 1989. The performance of the three coastal counties (Beijing, Tianjin, and Shanghai), in contrast, was relatively lackluster. Throughout China, but especially in the coastal provinces, enterprises in the non-state sector were the star performers (Table 1). In the Chinese context, the non-state sector includes collective enterprises, which are typically owned by local governments that is, by governments below the provincial or county level whose officials have been a key source of domestic entrepreneurship (see Bateman and Mody 1991 and Oi 1992). I Tony Walker's (1993) apt metaphor rightly focuses the spotlight on China's entrepreneurs who include not just factory managers but also local government officials especially mayors of cities and counties. 1  2 Explaining Industrial Growth in Coastal China Four influences contributed to Chinese growth: a set of economic reforms to establish the basis for a market economy, a proactive approach to attract designated entrepreneurs, the existing stock of physical and human capital, and the momentum of growth itself, especially regional spillovers that amplified growth effects. To take the first of these influences, reforms allowed greater play of economic incentives and increased the autonomy of state-owned-enterprises. Though ultimately directed at both old and new economic participants, reform was focussed on improving resource use within the existing state-owned system of production and distribution. While results were not dramatic, better resource use was achieved (Byrd 1992; Tidrick and Chen 1987; Jefferson, Rawski, and Zheng 1990; and Jefferson and Xu 1990 and 1991). For example, Jefferson and Xu (1990) found that the behavior of increasingly autonomous managers conformed to the predictions of a profit- maximizing model, and in their later work, they found that allocative efficiency increased under price reform (Jefferson and Xu 1991). Despite considerable market liberalization, however, the operation of market forces remains substantially circumscribed. Reforms have been introduced only at a gradual pace. Moreover, the sequencing has been unorthodox, characterized often by abrupt cessation or even reversal of specific measures. In 1981-82, for instance, foreign exchange controls were strengthened, in 1985-86, certain import controls were reimposed, and in 1989, price reform was delayed (see Sung 1991 for details). Labor and capital markets are subject to numerous controls and rigidities. Lal (1990) argues that labor mobility was deliberately limited to ease the transition. Even product markets are constrained substantial output is channeled through state-owned trading corporations. Trade polices continue to evolve slowly from a closed to a more open 2 regime. To complement reforms for increasing allocative efficiency, China has pursued a long-term strategy for encouraging investments by specific new entrepreneurs. The "open door" policies and "Special Economic Zones" have successfully attracted investments from overseas Chinese to the south-eastern coast. The strategy was especially calculated to bring Hong Kong investors to Guangdong and, more recently, Taiwanese investors to Fujian. Guangdong and Fujian provinces were not only physically close to overseas Chinese communities but also were "relatively backward," and thus carried a less onerous burden of earlier state investments. At the same time, local governments were given greater autonomy to invest in new business ventures (e.g., the so- called "collectively-owned-enterprises") and in infrastructure. Existing strengths and capabilities have also contributed to growth. We consider enabling factors such as domestic human resources and infrastructure, which shape a growth-conducive environment (Hulten and Schwab 1991; Barro 1991; Dollar 1992) and structural features which facilitate knowledge diffusion, including the diversity of the industrial structure, degree of 2 The black-market premium on the exchange rate is often used to measure the opennes of an economy to foreign trade (see World Bank 1991 and Barro 1991). In March 1993, 8.2 yuan traded for one dollar on the free market, implying that the US dollar commanded a premium of 50 percent over the official rate (Oxford Analytica, March 2, 1993). The official and free- market exchange rates of the Indian rupee were, in contrast, much closer. From January 1, 1994, China has abolished the official exchange rate, though access to foreign exchange is expected to remain restricted for a number of years (Oxford Analytica, January 4, 1994). See also Panagariya (1993) on the "mysteries" of trade policy in Jiangsu, one of the more outward-oriented provinces.  Introduction 3 competition (and associated impetus to entrepreneurship), and social interactions (Jacobs 1969; Romer 1986; Porter 1990). In taking this extended approach, we hope to illuminate the inte.-play between reforms and conditions that are conducive for growth. For example, we are particularly interested in the interactions between policy measures to attract foreign expertise and the existing stock of human capital or the diversity of the existing industrial structure. Finally, growth, once initiated, has a dynamic of its own, creating spillovers that may or may not be anticipated. Given China's comparative advantage, it is not surprising that light (labor-intensive) industries have grown faster than the heavy industries that were especially favored earlier. Of special interest, however, is the changing identity of the most rapidly growing industrial sectors even in the short period examined. Surprisingly, these sectors coincide across the different provinces and counties, displaying a "wave-like" phenomenon of growth and suggesting interesting underlying patterns of regional coordination or diffusion. Thus, within the limitations of secondary data, our purpose is to direct attention to the elemental forces of entrepreneurship, human capital, and knowledge diffusion that complement the gains achieved from better resource allocation and less waste. Though this paper is neither a policy brief nor a commentary on the subtleties of China's economic reform, the focus on long- term growth determinants points to a set of policy measures that encourage new entry, attract foreign knowledge, create new investment in human capital and infrastructure, and make judicious use of the advantages of backwardness. We use data on the output of 23 industrial sectors in seven provinces/counties.3 We first decompose output growth into time-dependent, region-wide, and industry-specific components, as well as their interactions, to identify the proximate sources of growth (Section II). Section HI describes an approach to studying the correlates of growth and a set of explanatory variables and Section IV explores various partial correlations of industrial growth. (In view of the known fragility of such correlations, extensive sensitivity tests are also reported.) The concluding section draws lessons for other countries. 3 The 23 industrial sectors are: food processing, textiles, apparel, leather products, wood and funiture, papcr and printing, art products, plastic products, petroleum, chemicals, pharmaceutical products, chemical fibers, electricity and steam, rubber products, non-metal, ferrous, non-ferrous products, metal machinery, transportation equipment, electrical machinery, electronics, and telecommunications. The eight regions for which industrial output data are available include: the three coastal coanties Beijing, Tianjin, Shanghai, and five coastal provinces Guangdong, Fujian, Jiangsu, Zhejiang, and Shandong. However, Beijing, though considered in our output decomposition analysis, could not be included in the regression analysis because of incomplete availability of explanatory variables.   2 A Decomposition of Output Growth Did growth occur across the board or only in certain regions or industries? Within regions or industries, did growth vary substantially from year to year? Variance analysis allows us to quantitatively decompose output growth into time, region, and industry-specific effects and their interactions. Identifying the main sources of variance in the data through the decomposition analysis helps in a preliminary quantitative assessment of the different sources of growth. Growth in time period t, region r, and industry i, G(tri), is assumed to be the additive result of main and interaction effects. (1) G(tri) = ,u + (t) + 13(r) + y(i) + a(tr) + b(ti) + c(ri) + c(tri) Where , is a constant, (t), 13(r), and (i) are the main time, region, and industry effects, respectively, a(tr), b(ti), and c(ri) are the second-order interaction terms between any two main effects, and (tri) is the interaction term for the three main effects. Following Schankerman (1991), variance of output growth can therefore be expressed as: (2) Var (G(tri)) = Var ( (t)) + Var (B3(r)) + Var (y(i)) + Var (a(tr)) + Var (b(ti)) + Var (c(ri)) + Var (s(tri)) Table 2 presents the results derived by equating the expected values of these variance components to their observed values (see appendix for the derivation). The small variance of (t) 4.98 percent implies that during 1985 to 1989, time-varying factors had only a minor effect on growth. Thus, while the overall pace of reforms accelerated, the effect was not felt uniformly in all regions and industries. Purely regional effects, B(r) were also small 5.72 percent implying that across years and industrial sectors, there was no consistent ranking of regional growth. Together with their interaction, time and region effects, a(t), 13(r), and a(tr), explain 12 percent of the growth variation; thus reforms did not make manifest themselves primarily either through general coastal expansion or through growth in specific coastal provinces or counties. 5  6 Explaining Industrial Growth in Coastal China Industry-specific factors (y(i)) were small as well, accounting for 2.88 percent of the variation in output growth; hence, no industrial group grew uniformly rapidly or slowly. For example, the electronics and telecommunications sector grew only 5.6 percent in 1985-86, whereas in 1987-88, it rose by a remarkable 19.3 percent. The dominant source of variation in the data comes from the interaction of time and industry (b(ti)), which explains 48.14 percent of the total growth variance. Thus, output growth rates for specific sectors varied from year to year, but within a year were strongly correlated across regions. This effect captures a "wave" phenomenon evident even from a visual examination of the time pattern of sectoral growth rates (table 3). There was a dramatic shift overall to light industries. However, within this group, different subsectors led in different years; some of the most labor-intensive sectors, such as garments, achieved their peak spurt only very late. The wave phenomenon was not restricted to light industry. In 1985-86, the wave was evident in leather goods, pharmaceutical chemicals, chemical fibers, and metal products in most of the coastal region. In 1986-87, electronics and chemicals replaced leather and metal products. In 1987-88, paper and printing, transportation, electronics, and pharmaceutical chemicals expanded rapidly, only to lose their position to the apparel industry in 1988-89. Even within "miscellaneous light industries," the most rapidly growing group, the "waves" were quite striking: in any year, a particular industrial sector had the dominant growth performance in all regions. Such synchronization could be accounted for by a shift in buyers preferences, industry- wide technology improvements rapidly transmitted along the coast, and the diffusion of strategies among decision makers for promoting sectoral growth. It seems unlikely that shifts in preferences could explain annual increases or declines in the growth rates of sectors. Technology transmission is a better bet; formal and informal interactions among firms, and labor turnover (particularly through highly skilled managers and engineers), may have extended technology and skills learned in the "open" areas to the rest of the coastal region (Ho and Huenemann 1984, p.55); however, without micro-level data, such hypotheses cannot be adequately tested. Another intriguing possibility is that decision makers (whether in the Communist Party or in the administrative set-up) maintain close ties that lead to the rapid diffusion of development strategies along the coast (see Yusuf 1993 and the literature that he cites). This network of decision makers provides a grid for information flows leading to replication of sectoral-targeting strategies, among other things. The remaining 34.78 percent of the variance in output growth is attributable to the third- order interaction between time, region, and industry (E(tri)). We interpret this as a regional effect conditional on time-varying industry-specific factors. While the changing identity of high-growth sectors is a major source of growth variation, this third-order interaction indicates that growth in an industrial sector during a particular year is not uniform in every location. Regional differences in initial conditions, human capital endowment, and infrastructure availability, among others, may cause industries in some regions to grow faster than those in the others. Thus even though the "own effect" of regional differences (13(r)), and its interactions with time (a(tr)) or industry (c(ri)) do not explain much of the variation in growth, after adjusting for time-varying and industry- specific factors, significant regional effects remain.  A Decomposition of Output Growth 7 To recapitulate. Unconditional time, industry, and regional factors did not carry significant explanatory power for growth variation. Instead, about half the variation in growth during 1985-89 is associated with time-varying sectoral growth differences (b(ti)). Regional effects, on the other hand, emerge only after controlling for the time and industry effects.   3 Investigating the Correlates of Growth The annual growth rate of output in an industrial sector in a given region is our dependent variable. Various industry-specific and region-wide factors are the independent variables whose correlation with growth we seek to examine. The goal here is not test any specific model of growth but to describe its most robust partial correlates. Following Glaeser, Kallal, Scheinkman, and Shleifer (1992), we focus on growth itself rather than on increases in productivity. (Their dependent variable was employment growth; we use output growth.) While it is common to use increases in labor or total factor productivity or growth of per capita income as the dependent variable, this was not possible in our case since labor force data by industrial sector are not consistently available. We find, however, that the data on industrial output growth are so rich that considerable insights can be obtained even in the absence of information on labor and capital inputs. Indeed, if one believes that enterprise-level decisions to acquire or invest in labor or capital inputs are influenced by available knowledge, infrastructure, and human capital, and by the industrial organization, then not only productivity but also a considerable amount of output growth can be attributed to these factors. Our industry-specific variables degree of specialization and competition (the latter also interpretable as degree of entrepreneurship or relative firm size) are the same as those used by Glaeser, Kallal, Scheinkman, and Shleifer (1992). We add a set of regional variables to the regressions. The assumption is that having controlled for industry-specific characteristics, the effect of region-specific variables (such as infrastructure) will be similarly felt by all industries (for a similar assumption, see Waldmann and deLong 1990, who analyze growth across industries in different countries, and Stockman 1988). Limited degrees of freedom prevent our running separate regressions for each industrial sector. This poses a problem because growth has not been uniform across sectors, raising the possibility that independent variables have very different influences on the different sectors. Since a signficant feature of China's recent growth has been the rapidly growing share of light industrial sectors, our intermediate solution to this problem was to group sectors into light and heavy industries and reestimate the basic regression. Although the coefficients show interesting differences in magnitudes, we find the basic results unchanged and hence focus on the pooled results, noting the differences that do arise when light and heavy industries are considered separately. 9  10 Explaining Industrial Growth in Coastal China The base regression is estimated with time and industry dummies, but without regional dummies. When dummy variables for regions are added to the regression, we, in effect, remove from the data the variation due to differences in the levels of variables across regions. The coefficients thus obtained are weighted averages of "within"-region relationships, which are sometimes described as short-run effects. When region dummies are not included, we are able to compare across regions. Since interregional differences occur over a longer period of time than do variations within a region, dropping regional dummies, as we do in our principal regressions, captures the "long-run" effects. We also report the more interesting "short-run" estimates. Tables 4 and 5 summarize the definitions and descriptive statistics of the variables and data sources. Industry-Specific Variables A structural feature of an industry is its degree of regional specialization greater specialization is good if the relevant knowledge is best acquired within the industry, but is deleterious when diverse skills and information from other industries are desirable. Also of interest are the degree of entrepreneurship and competition, which can spur investment, although too much competition can lead to diminished investible surpluses. As will be noted, the existence of entrepreneurship or competition is inferred only indirectly from the size of firms in industry "i" in region "r" relative to its average in all seven regions. As in Glaeser, Kallal, Scheinkman, and Shleifer (1992), we calculate the following measures of specialization and entrepreneurship4: Sirt = (output in industry i in region r / total output in region r) (output in industry i in 7 regions / total output in 7 regions) Eirt = (number of firms / output) in region-industry ir (number of firms in industry i / output in industry i) in all 7 regions The time subscript indicates that these measures are different for each year. Sirt is the ratio of the share of industry i in region r to its average share across the seven regions. S greater than 1 implies that the industry commands a larger share of the region's output than the average share that industry enjoys in the seven regions. We interpret a rising Sirt for a region-industry as an indication of increasing specialization of that industry in that region. As S increases, knowledge flows will be increasingly restricted to sources within that industry. Learning from other industrial sectors is likely to be greater when S is low. Jacobs (1969), who contends ti .t information exchanges with different sectors are more productive than within a 4 The regression analysis, unlike the variance decomposition, is based on data only for seven regions (the five coastal provinces and two counties Tianjin and Shanghai); foreign investment data for Beijing were not available.  Investigating the Correlates of Growth 11 sector, predicts that the high-S industries will grow more slowly than the low-S ones; Porter (1990) makes the opposite prediction.5 Eirt is a possible measure of entrepreneurial strength, but could also measure the degree of competition, or merely relative size of firms if small firms are synonymous with more competition, or more firms imply existence of entrepreneurship, then the interpretations of the variable are indistinguishable. A high E for a region-industry implies more firms for a given output in that region relative to the average number of firms divided by output in the industry across all seven regions. A high E could, therefore, be interpreted as more entrepreneurship, greater competition, or smaller average firm size. In terms of effects on growth, an unresolved debate centers around whether competition or monopoly are more effective in encouraging innovation. Similarly, the effect of size on growth remains controversial. A control variable, GO, or growth of the industry outside the province/county, is also included in the regressions. By construction, GO is a time-varying region- and industry-specific variable. However, as a practical matter, since it captures across-the-board industrial growth, GO is close to being a time-varying industry-specific factor with little regional variation in a given time period. Much interest revolves around the interpretation of the channels through which GO influences growth. Region-Specific Variables The region-specific variables are the beginning of the year values of: GDP per capita (Y/N), secondary-school enrollment rates (SEC), foreign direct investment per person (FDI), road network (ROAD, proxied by the length of roads in the region divided by area), telecommunications availability (PHONE, telephone lines per capita), and congestion (DEN, measured by the population density). These growth-inducing factors are productive not only in their own right but can have important external or spillover effects. Lucas (1988) notes that human capital is "twice blessed": first because of its inherent productivity and second because interactions among well-educated people further increase efficiency. Shleifer (1990) suggests that good infrastructure provides the focal point for the development of agglomerations, which in turn create the environment for knowledge spillovers. Foreign investors bring in knowledge on international best practice in production technologies and also provide links to international markets. 5 Interpretations other than knowledge flows can also be used to "explain" the link between S and growth (for example, suitability of regional factor endow.ments to the sector may contribute to a positive relationship between S and growth).   4 Correlates of Growth Because results in this type of analysis are sensitive to the variables included (Levine and Renelt 1992), table 6 reports those results that appear robust to various specifications based on sensitivity tests (described at the end of this section). The time, region, and industry dummies are not reported in the tables presented, but their main patterns are worth noting. Consider first the time dummies. The size of the time dummy coefficients shows an upward trend through the period under consideration, and the coefficients are significantly different from the constant term for the base year. However, excluding the time dummies only reduces the R-square marginally, indicating their limited explanatory power. The statistical significance of the time dummies is sensitive to whether the observations are weighted by the population of the region; when observations are not weighted by population, the influence of the two counties, Shanghai and Tianjin, increases and the time dummies are not significantly different from zero, suggesting that the time effects were felt primarily in the five coastal provinces. The inference we draw from the increasing coefficients on the time dummies, consistent with the variance decomposition analysis, is that there were independent, though limited, time effects during this period. In other words, the gradual move toward a more market-oriented economy appears to have had some secular effects independent of the region and industrial sector.6 The coefficients of region dummies, using Shanghai as the base for comparison, are positive but not significantly different from zero (even at the 10 percent level of confidence). When we drop region-specific variables (education, infrastructure, foreign investment and initial per capita income) from the regression, the pattern of region dummies mirrors more closely the statistics of regional growth presented in table 1, with the regional coefficient for Guangdong higher than Fujian, followed by Jiangsu, Zhejiang, Tianjin and Shanghai. The exception in the regional order of growth is Shandong, whose coefficient indicates a higher growth rate than Guangdong's, although the F-test shows the two coefficients are not significantly different from 6 We also examined the time dummies after dropping the region-specific variables. Again, the estimated coefficients on the time dummnies are statistically different from the base year. More interestingly, the time dummies now reflect the 'stop- go" process noted earlier, tracking the downturn in 1986, a year of student unrest and economic uncertainty, resumption of growth in 1987 and 1988 when reforms resumed, and the slump in 1989 when the reform program was set-back. The implication, therefore, is that the stop-go process is basically reflected in the regional explanatory variables themselves once these factors are controlled for, growth seems to increase over time. 13  14 Explaining Industrial Growth in Coastal China each other. These results give some confidence that the regional variables, such as foreign investment (and accompanying know-how), domestic investment (especially in infrastructure), human capital, and initial per capita income levels are good explanatory variables for differences in regional growth. Industry dummies show no interesting pattern and including or excluding these dummies makes little difference in the results. The regressions that follow always include time and industry dummies. Finally, since we are pooling time series and cross-sectional data, we tested for serial correlation in growth rates. If observations in growth rates in four adjoining years in a specific industry in a particular region are not independent of each other, the standard errors will be biased and the inferences drawn will be stronger than warranted. Recall that we have three dimensions in our data: industry, region, and year. Our interest is in the correlation over time. We can, therefore, sort the data by region, then by industry, and finally by year; alternatively, we can sort by industry, followed by region and year. Both procedures ensure that the adjoining observations are for four successive years. In either case, the Durbin-Watson statistic for the base regression (with time and industry dummies, but without region dummies) is 1.87, implying that the autocorrelation problem is not serious. Specialization After controlling for other variables, industrial specialization (diversity) has largely a negative (positive) effect on growth, suggesting that knowledge flows across industries are more conducive to growth than such flows within an industrial sector (Jacobs 1969). Less-specialized industrial sectors gain from knowledge spillovers from other sectors. The short- and long-run effects are not very different (table 6). Recall that at S equal to 1, the output share of industry i in region r equals the average output share of industry i in the seven regions. A decline in S to 0.9 increases the growth rate by 0.5 percentage point (e.g., from 6 percent to 6.5 percent). However, the relationship between industrial specialization and growth does not appear to be linear. Beyond S equal to about 2, specialization (rather than diversity) enhances growth. The evidence, therefore, is also consistent with Porter's hypothesis on the benefits of knowledge flows within the same industry, although the degree of specialization must be large enough. Entrepreneurship, Competition, and/or Size of Firm The statistical significance of the coefficients for E and E2 is weak. However, the general thrust of the results is similar across various specifications and hence worth noting. For the most part, increasing firm size has a deleterious effect on growth. The largest (and possibly least competitive) firms are in Shanghai and Tianjin, and in the state-owned sector. Despite increased autonomy, these enterprises have, with some significant exceptions, been the most constrained in responding to changed conditions. In contrast, the newer and smaller firms, often promoted by enterprising local governments, have been more flexible; in part, this may be because they have fewer obligations to various constituents and, in part, because they have greater access to international sources of technical knowledge and to marketing channels.  Correlates of Growth 15 The data on hand does not allow us to answer an important and interesting question. Is the lack of dynamism of large firms due to their being state-owned or do most old firms lack the skills required for adjustment? Developing countries with much smaller state-owned sectors may have little basis for complacency: many of the privately owned large firms, ensconced within business houses, are liable to be as sclerotic as the most unresponsive Chinese state-owned enterprises. The value of the coefficients, however, also suggests that when the average size of firms in an industrial sector in a particular region is smaller than a third of the average firm size in all regions, growth in that region-industry suffers (possibly through excessive competition and/or diminished investible surpluses). Once again, short- and long-run effects differ little. Growth of the Industry Outside the Region Results show that the growth of an industrial sector in any region is powerfully influenced by the growth of the same industry in other regions during the same year (as represented by GO). On average, a 1 percent increase in the growth rate of an industrial sector outside the region is associated with 0.78 percent increase in the growth rate of that industry within the region. Unlike other variables, this variable not only passes the test of significance, it also accounts for 49 percent of the total sum of squares. This is another way of capturing the "wave" phenomenon noted in the variance decomposition exercise. The high t-statistics for the GO coefficient were also obtained by Glaeser, Kallal, Scheinkman, and Shleifer (1992), who interpreted the result as a demand effect exogenous growth in demand, in this view, conditions the growth of specific sectors irrespective of the region. However, as noted above, synchronization across regions can also occur as a result of technology diffusion or networking among decision makers. Two extensions of the basic regression were tried to gain further insight into the influences at work. First, we interacted the GO variable with the specialization variable. The results indicate that GO has a stronger effect in conditions of industrial diversity (table 7, column 1). In other words, the more a sector is specialized within a region the less it is affected by growth of that same industry outside the region. Thus not only can specialization limit growth by restricting opportunities of knowledge flows between sectors within the region, but highly specialized sectors can also suffer due to isolation from developments in other regions. In general, light industrial sectors tend to be less specialized than the heavier sectors. Light sectors can respond more quickly to external impulses. This finding is discussed further when we analyze light and heavy sectors separately. Second, in view of the policy attention accorded to Guangdong (and more recent!y to Fujian) and also given their physical proximity to Hong Kong and Taiwan, a question of interest is whether these regions were conduits of growth impulses. Surprising, these favored regions are better characterized as imitators than as leaders in the acquisition and diffusion of knowledge. When the variable GO was interacted with region dummies, the coefficients showed that Guangdong benefited most from growth outside the province and Fujian was third on the list, with Jiangsu in between (table 7, column 2). We then replaced the variable GO (which is growth in all outside regions) with growth in Guangdong (GGD) as an independent variable to isolate the effects that Guangdong may have had on growth in other regions (Guangdong itself was not  16 Explaining Industrial Growth in Coastal China included in this regression). Guangdong's growth does have a statistically significant impact on other regions but the magnitude of the effect is much smaller than when growth in all other regions is considered (table 7, column 3). Similar conclusions apply to Fujian. The imitation possibilities in Guangdong, and also in Fujian and Jiangsu, must be placed in context. Within the coastal region, these provinces have the greatest flexibility to respond to external stimulus. As other regions become more receptive to change, Guangdong and Fujian can be expected to have a greater spillover effects. Field surveys in Guangdong and Fujian show unambiguously that modern production techniques, including sophisticated methods of quality control, are being rapidly adopted in these provinces. As such experience accumulates, increasing labor mobility will complement existing administrative communication networks to diffuse the knowledge gained to other parts of China. Foreign Direct Investment A key element of economic reform in China has been the "open door" to foreign investment. Though triggered by government policy, growth in foreign investment has taken on a life of its own, reaching close to $20 billion in 1993. Many overseas Chinese have invested large amounts of capital and know-how, despite what, by Western standards, would be considered a great deal of uncertainty regarding property rights and enforcement of contractual obligations (see Yusuf 1993). Our results show that foreign direct investment has a strong impact on growth, particularly in the short run (column 1 in table 6). The short-run elasticity of growth with respect to foreign direct investment, calculated at the mean value of the foreign direct investment variable, is 0.10, indicating that a 10 percent increase in foreign investment can raise the growth rate by 1 percent. However, the apparent effect of foreign investment is influenced by trends in secondary- school enrollment rates, which as noted below, fell during this period of rapid growth. Hence "human capital" is seen to have a perverse effect on growth in the short run (see column 1, table 6). Since a change in school enrollment rates is not a good measure of change in the stock of human capital, the perverse effect is overstated, and to that extent, the positive effect of foreign investment is probably exaggerated in the short-run estimates. When the secondary-school enrollment rate is dropped, the coefficient for foreign investment falls by about half (see column 2, table 6). If we assume that there was little change in human capital within any region during the period under consideration, then the new estimate for foreign investment is closer to "being right," and hence the elasticity of the growth rate with respect to foreign investment is closer to 0.06.7 The effect of foreign investment declines in the long run (and hence is a less potent source of growth differences between regions) but still remains statistically significant and quantitatively important. The foreign investment coefficient decreases from about 4 to about 2, and the growth elasticity falls from 0.06 to 0.03. 7 When the secondary enrollment rates are dropped from the equation, the coefficient on foreign investment declines but other coefficients remain essentially unchanged (see column 2, Table 6).  Correlates of Growth 17 One interpretation of these results is that in the short run, foreign investment is the most mobile factor and hence is a dominant driver of growth. In the longer run, such variables as education and infrastructure respond to increased demand for complementary assets, and the contribution of foreign investment declines. Note that foreign investment here applies to the region as a whole, whereas the dependent variable is growth in a particular sector of the region. Thus foreign investment has a general influence on the growth of all sectors, not just the sector in which the investment is made. Such spillovers could arise from the flow of knowledge that accompanies the investment. This interpretation finds some support in the evidence that foreign investment has a stronger effect in more industrially diverse settings (smaller degree of sector specialization) table 9, column 1. There is also a complementary relationship between domestic human capital formation and foreign investment flows, as discussed below. Human Capital: Education versus Foreign Knowledge Measurement of the stock of knowledge available for productive use is a complex task even under normal conditions and is especially difficult in a dynamic situation when knowledge from many different sources is being utilized. Traditionally, secondary school enrollment rates have been used as proxies for domestic stock of knowledge, or domestic human capital, and serve well as long-run approximations. Using data for the only year available 1987 we compared enrollment rates to the more appropriate proxy, average years of schooling in the labor force, and found a very high correlation coefficient (0.965, significant at 99 percent) between the two indicators.8 If this finding applies to the whole period between 1985 and 1989, then secondary- school enrollment is a good surrogate for human capital endowment and our long-run estimates can be considered reasonably reliable. However, short-run changes in human capital are more difficult to measure. The extensive reforms that began in 1984 were accompanied by an actual fall in school enrollment rates in most of the coastal provinces/counties (table 8). This is not altogether surprising during a period of rapid growth and accompanying increases in demand for labor. Many of the new entrants to the labor force were young women who probably dropped out of school to take up newly available jobs. Over the short period under consideration, the stock of domestic human capital is unlikely to have changed as a consequence of such labor force responses, although unless the trend is reversed, human capital will deplete over time. The short-run, or "within," estimates show a negative coefficient for secondary education (first column in table 6), reflecting the cyclical shift out of education described above. The finding tells us little about the relationship between domestic human capital and growth in the short-run, since, as noted, changes in secondary enrollment rates greatly overstate the depletion of human capital. In the long run, that is, when the comparison is across regions, education has the 8 The labor force includes population in the age group 15 to 54. The average length of education is calculated as (16U + 12H + 9M + 6E + OI)/T, where U, H, M, E are the number of persons with university, high school education, middle high, and elernentary school education, respectively. I stands for illiterate. T is the total population in the working age group. The relevant data were obtained from the 1987 population census. Seven provinces/counties in the coastal region have data available Guangdong, Fujian, Zhejiang, Jiangsu, Shandong, Beijing, and Tianjin.  18 Explaining Industrial Growth in Coastal China expected positive effect on growth. If a linear relationship (column 4, table 6) is assumed between secondary-school enrollment rates and growth, a 5 percentage-point increase in secondary-school enrollment rates raises the growth rate by about 4 percentage points. However, returns to secondary education diminish beyond a point.9 The non-linear relationship (column 5) shows that when enrollment increases from 30 percent to 35 percent (that is, approximately from the Fujian enrollment rate to the Guangdong enrollment rate), growth rises by 5 percentage points. However, when enrollment increases from 55 to 60 percent, the increase in growth is only 3 percentage points. Thus, Tianjin gets a smaller bang from raising its enrollment rate than does Fujian; Shanghai, with an enrollment rate in the mid-60 percent range, gains even less. For Tianjin and Shanghai, it would appear that investments in roads and telephones, which are characterized by increasing returns, have a higher payoff. The effectiveness of education further increases when associated with foreign knowledge. Column 2 in table 9 shows that the interaction between school enrollment rates and foreign investment is significantly positive, suggesting mutual reinforcement between domestic human capital and foreign knowledge that accompanies the investment. Also, the coefficient on foreign investment become much weaker when the interaction term is introduced, implying that much of the power of foreign knowledge comes through the local human capital base. Complementarity between local and foreign knowledge arises because they serve quite different functions. Classroom education provides the basic reading and quantitative skills necessary to function effectively in a modem production enterprise, but can be ineffective if the economy is isolated from the global community (Romer 1992). Exposure to foreign knowledge breaks the isolation of the local economy and brings experienced-based practices that are rarely available in textbooks and are best communicated in a hands-on manner in a production setting. Though the results point to the importance of domestic human capital, at least in so far as it facilitates the absorption of efficient production practices, it is important to note that the levels of secondary education required are not very high, at least in the early stages. For instance, in 1985, secondary enrollment rates were 35 percent in Guangdong and 31 percent in Fujian."0 Growth in these provinces was based on investments from Hong Kong and Taiwan in small-scale, medium-technology, labor-intensive manufacturing. As the economy matures and becomes more sophisticated, however, demands for human capital are likely to increase. Without sufficient human resources, it is unlikely that more sophisticated foreign technology will be easily absorbed. Infrastructure Good infrastructure not only facilitates the flow of information but also provides the focal point for the development of agglomerations (Shleifer 1990). We consider two types of infrastructure: roads and telecommunications. Roads represent the traditional infrastructure and 9 The coefficient of the square of secondary enrollment rate in column 5 of Table 6 is negative but not statistically significant; however, we find that this result is sensitive to the specification and, in certain cases, the squared term is statistically significant. Thus we believe that the non-linearity needs to be taken seriously. 10 In 1985, the enrollment rates were 53 percent in Tianjin and 68 percent in Shanghai; since then they have fallen somewhat throughout the coastal region.  Correlates of Growth 19 their stock has grown only slowly (though this is likely to change as ambitious, privately built expressways come on stream). Phone lines, in contrast, have grown rapidly to meet the needs of the international trading community much, possibly all, of the new telecommunications investment uses modem digital technology. The results show that a network of roads has a positive effect on growth but is subject to diminishing returns in the short run (column 1, table 6). Roads are more productive in high- density areas (as reflected in the positive coefficient on the interaction term between roads and population density). In the long run, roads appear to enjoy increasing returns and density has little influence on their productivity (columns 4 and 5, table 6). The mean elasticity of growth with respect to the roads network, calculated at the mean values of the two variables, is -0.301 and 0.028 in the short- and long-run, respectively. Telecommunications growth shows an even stronger effect; telephones per 1,000 residents show increasing returns both in the short- and the long-run (columns 3 and 6, table 6). The short- and long-run elasticities are both approximately 0.10. The negative short-run elasticity of the road network could reflect indivisibilities and consequent scale economies (Weitzman 1970). Both planning and execution take a long time, and development must be based on ten- to twenty-year forecasts that take into account future population size and economic growth. Infrastructure may be more abundant than necessary in the short run as growth will not have reached the limit set by the targeted capacity. (Obviously, redundancy will be less of a problem where congestion or unsatisfied demands are severe.) The long-run increasing returns of infrastructure may be related to network effects. In the case of transportation, gains from an increase in route length rise as the longer routes interconnect with new areas and multiply the connections possible. As a result, economic opportunities increase, and higher growth occurs. Similar, and possibly more powerful, network effects work in telecommunications. The effectiveness of foreign knowhow accompanying investment flows also depends on the availability of infrastructure. Good infrastructure speeds up diffusion, increasing the number of adopters of innovations. This hypothesis is supported by our data, as is shown in the strong positive interaction between foreign investment and the roads network (see column 3, table 9). Initial Conditions The initial per capita income of a region turns out to be an important variable in explaining subsequent growth. When initial per capita income is not included in the regressions, the partial correlations between growth and the other variables change significantly; as noted below in our discussion on sensitivity tests, variables other than per capita income do not have a similar influence when added or dropped from the analysis. The strongly negative relationship between industrial growth rates in a region and the initial per capita income of the region suggests that growth is being influenced not just by "steady- state" factors but also by transitory influences. If steady-state growth had been achieved in the different industrial sectors and regions, both neoclassical and endogenous growth models predict that the initial levels of backwardness would have no influence on subsequent growth (Mankiw, Romer, and Weil 1992). Only when an economy is moving to a new steady state will initial levels  20 Explaining Industrial Growth in Coastal China of backwardness provide an additional impetus to growth. This seems particularly appropriate for coastal China, which has indeed been shaken-up and put on a new growth trajectory. Figure 1 shows a strong inverse relationship between rate of growth of industrial output during 1985-89 and the log of per capita GDP in 1985. In the terminology suggested by Mankiw, Romer, and Weil (1992), and Barro and Sala-i-Martin (1992), there is evidence of absolute convergence. In other words, even without controlling for other variables that may affect steady- state growth, the relatively backward provinces grew faster than the more advanced regions. For example, initial backwardness partly explains why Fujian grew so fast despite low educational attainments and limited infrastructure. Absolute convergence applies not only to industrial growth (as described in figure 1), but also to per capita GDP. Over the 1980s, the per capita GDP of the five (relatively poor) coastal provinces caught up with that of the three richer counties (the ratio of GDP per capita in the five provinces to that in the three counties rose from 0.23 in 1980 to 0.38 in 1988, see State Statistical Bureau 1990). Elsewhere, there was no absolute convergence. The per capita income of the coastal region was higher than in the rest of China when the reforms were launched, and the gap has increased over time. The region's GDP per capita was 50 percent higher than the average in the rest of the nation in 1980; it was 74 percent higher in 1988. These observations point to an interesting international parallel. In cross-country comparisons, absolute convergence is observed among advanced industrial countries but not among poor economies. Poor economies converge "conditionally", i.e., after controlling for education and investment rates. Within the group of industrial nations, the rate of conditional convergence is higher than the rate of absolute convergence, since the richer ones typically also have higher education and investment rates (see Mankiw, Romer, and Weil 1992). We have not investigated the possibility of conditional convergence outside the coastal region. However, not surprisingly, conditional convergence within the coastal region, as within the industrial economies, is more rapid than absolute convergence. The richer coastal regions also tend to have better education and infrastructure, and thus it may be supposed that they have higher steady-state growth rates. The fact that the poorer regions are growing faster despite their lack of endowments indicates that they are benefiting from their backwardness. The common interpretation of this catching-up phenomenon is that regions with low per capita income also have low capital per worker and so have a higher marginal product of capital than regions that are well-endowed with capital. Thus the poorer regions potentially attract new capital (along with new ideas). The evidence certainly supports this view: the poorer regions have attracted huge amounts of foreign capital and knowledge. But in addition, as discussed above, the more advanced regions have been burdened by an institutional set-up that has been a drag on growth. Light and Heavy Industries Thus far we have assumed that all industrial sectors respond to the explanatory variables in the same manner. Here we note some differences between light and heavy industrial sectors (table 10). Though the differences are of interest, the exercise also gives us confidence in the  Correlates of Growth 21 results reported so far the signs of the coefficients are almost identical and the key variables (barring education) continue to be statistically significant for both light and heavy industries. The estimated equation does a better job of explaining growth in light industries (R2 0.67) than in heavy industries (R2 = 0.49). Of special interest is the finding that the GO variable, which measures the degree of synchronization or diffusion across regions, has a higher coefficient for light industries. This is to be expected given the lower capital intensity and hence higher mobility of light industrial sectors. Guangdong, Fujian, and Jiangsu benefit specially from growth outside the region in both heavy and light industries; recall, that this difference is measured by interacting GO with regional dummies." In light industries, other regions also benefit strongly from the diffusion process, whereas the effect for heavy industries falls off in other regions and is not statistically different from zero for Shanghai and Tianjin. In both light and heavy industries, when growth in Guangdong is used as an explanatory variable (GGD), the partial correlation is positive and significant, but smaller in magnitude than the coefficient obtained for GO, implying again Guangdong is more an imitator than a leader. Foreign investment provides a bigger bang in light industries, although it has a significant coefficient for heavy industries. Similarly, infrastructure does more for light than for heavy industries. Education has a positive effect on growth in light industries, but the effect is not statistically different from zero. Thus, although formal education is important, its relationship with growth is noisy, and tacit knowledge based on experience (and channeled through foreign sources) appears to be a somewhat firmer predictor of growth. For heavy industries, we observe diminishing returns to education, as was seen above for all industries; within the range of observed secondary-school enrollment rates, this implies a positive, though declining, effect of education on growth. Diversity (or the lack of specialization) has a stronger association with growth among light industries, which is not surprising; skills are likely to be more mobile in such sectors. When all observations were pooled, we noted above that specialization was an aid to growth only beyond S=2. For light industries, the positive effects of specialization are felt at even higher levels of specialization (beyond S=2.5); in comparison, for heavy industries, specialization is conducive to growth after S=1.3. The implication is that specialized sectors, which have also grown rapidly, are principally in the heavy industry group. Sensitivity and Misspecification Our sensitivity analysis uses the methods of Belsley, Kuh, and Welsch (1980). We first dropped one observation at a time and found that no single observation influenced the coefficients significantly. This result could have been expected, given the large sample of 640 observations. We then dropped specific sets of observations, excluding from regressions a province, a year, an industry, a region-industry, a year-industry, and a year-region. The distributions of the coefficients show a very strong concentration around the mean value. We can therefore rule out the possibility of outliers driving our regression results. 11 These results are not presented to conserve space but can be provided on request.  22 Explaining Industrial Growth in Coastal China In the regressions reported, we have weighted the observations by the population of the region, which gives more weight to the provinces and less to the counties, reducing the influence of the counties in the regression results. To see how much the results are influenced by this weighting procedure, we also ran our basic regression by treating every observation equally (column 7, table 11). The results do not change qualitatively, except that diminishing returns to education are now more evident: this is as expected since the more educated counties that recorded relatively modest economic performance now have greater weight in the regression. Another type of sensitivity analysis was done by adding or dropping independent variables (table I1). Omitting secondary school enrollment rates has little effect on the sign and magnitude of the remaining coefficients (column 4). Similarly, the regression results are not sensitive to specifications that exclude an entire set of industry- or region-specific variables, as columns 5 and 6 demonstrate. If there is no serious misspecification problem, regional factors other than initial per capita income predict that the counties (Tianjin and Shanghai) should do especially well because they have had better than average access to foreign investment, education, and infrastructure. But instead growth in these counties was slow, possibly because of the significant presence of state- owned enterprises, which is not captured in the regressions. When we included the share of state- owned enterprises as an independent variable, it did not generate significant results since the share of these enterprises is correlated with per capita income (and also with the variable E, which is the inverse of average firm size). Thus, the relatively slow growth in recent years of the two richer regions reflects diminishing returns, which arise not merely from a technological source but also from the constraining effects of the institutional structure within which past industrialization occurred. The issue of causality is an open question. Identifying causality is difficult and our goal, as a first step, has been to identify the bundle of influences that coexist through a growth process. However, certain conventions were used throughout the paper that deal partly, and imperfectly, with the issue of endogeneity. The potential endogeneity of the industry-specific variables, competition and specialization, is addressed by using their lagged (beginning of period) values. The endogeneity of regional variables poses a less serious problem: first, beginning-of-period values are used in the regression, and second, our dependent variable is not growth in a region, but rather growth in a specific industry within the region. Infrastructure, education, and flows of foreign investment are likely to be influenced by overall regional growth rather than by the expansion of a particular industry.  5 Conclusions The basic facts are relatively clear. Growth rates in China's coastal provinces were substantially higher than in the interior. A remarkable shift to non-state enterprises and light industrial growth occurred all over China, but especially along the coast. Within the coastal region, the traditional industrial centers Beijing, Tianjin, and Shanghai although declared "open" quite early, grew at a much slower pace than the more "backward" coastal provinces, especially Guangdong and Fujian, which gained tremendously from the flow of investment from Hong Kong and Taiwan. Almost half of the variation in industrial growth along the coast is attributable to the synchronization in growth of particular industries across provincial and county boundaries. The identity of the most rapidly growing sectors changed from year to year across the entire region, possibly the result of informal information flows. What role did Guangdong and Fujian play in this diffusion process? These two provinces which have the best physical access to overseas Chinese knowhow and investment, are at present better at absorbing than at disseminating knowledge. Good education also differentiates good performers from poor performers over the long haul. When all sectors are considered, gains of even a few percentage points in secondary-school enrollment rates have an important effect on growth. When only light industries are considered, however, the relationship between growth and secondary-school enrollment is potentially important but imprecise. For heavy industries, education has diminishing returns, although the positive effects continue well into the range observed in the sample (as well as the range spanned by most middle-income countries). The role of secondary-school education, however, cannot be considered separately from knowledge acquired through international links. Secondary-school enrollment rates in Fujian province at 31 percent are close to the average for low income countries (World Bank 1991). Our results suggest that China's coastal provinces were able to exploit their educational attainments better than other low-income regions because the complementary effects of foreign knowledge enhanced the educational level of the workforce. Infrastructure investment, particularly in telecommunications but also in roads, yields increasing returns. There is some question whether infrastructure is a true enabling factor; while it accelerates output growth, it also responds to growth. Large infrastructure investments are occurring along the coast in the wake of the huge growth of the past several years. Thus although 23  24 Explaining Industrial Growth in Coastal China good infrastructure is valuable, conditions that enable externality-generating infrastructure investments to be put in place as demand emerges are equally important. Industrial diversity and competitive conditions spur growth, but the connections are not straightforward. A high degree of industrial diversity is associated with rapid growth, especially for light industries; but specialization tends to promote growth in heavy industries. Growth can be impeded by excessive competition as well as by too little competition. Finally, relative backwardness has been an important factor in growth. We observe not only conditional convergence (that is, convergence after controlling variables influencing steady- state growth) but also absolute convergence within the coastal region. Lessonsfrom Coastal China China has pursued a decentralized economic reform program. Particular reforms have been tried in specific regions sometimes with and sometimes without the blessing of the central government. Critics have noted that the lack of coordination and wasteful regional competition have resulted in damaging macroeconomic effects. However, the strong synchronization across regions suggests that there have been counteracting influences. Lessons and strategies from Guangdong and Fujian evidently move to other coastal areas, although these provinces benefit even more from regional information flows. The precise source of this synchronization cannot be discerned from the data at hand, but it is clear that a network of communication channels exists across the country. Such a network could reflect the links between the cadres of the Communist Party or could even predate the party, reflecting much older economic and social ties (Yusuf 1993). The network is a substitute for central direction; indeed, it is probably superior to central mandates. Imitation through network communication is likely to be less prone though not immune to major errors. The fundamental insight in this paper is that success has required a combination of centrally approved local experiments, local government entrepreneurship, and an effective network for diffusing success across different regions. An interesting aspect of the decentralization has been that the "easier" regions, that is, regions with relatively low per capita income and hence a large "catch-up" potential, were targeted early on. These regions were relatively unencumbered by state-owned enterprises, planning bureaucracies, and other mechanisms that guided output in the prereform era. Inceed, some of the counties in Guangdong province that experienced the most spectacular growth rates, such as Shenzen and the neighboring areas, were essentially agricultural communities (or even wastelands) 15 or 20 years ago. While the successes of the strategy have been evident, questions have been raised about policy reversals and set-backs and the consequent lack of government credibility (see Sung 1991 and Chen, Jefferson, and Singh 1992). Such credibility lapses are generally viewed as expensive, inasmuch as they create investor uncertainty and reduce investment. Yet investors, especially foreign investors, have scarcely been deterred. Foreign investment has almost been an independent, exogenous force, dampened occasionally by policy conditions, but largely oblivious of policy contradictions and reversals. At the same time, locally financed infrastructure and  Conclusions 25 human capital investments plus job-training within enterprises, have proceeded with vigor, fueing growth. We suggest two related possibilities: First, the credibility of government policies as a determinant of investment is overrated; it is likely that credibility and certainty derive from performance itself rather than from government actions; and second, investors may accept contradictions and reversals as a reflection of the government's response to evolving conditions. If this analysis of China's recent experience is approximately correct, what lessons does it hold for other countries? Decentralized experiments are valuable, but to have any chance of success, they require local governments that are entrepreneurial. Human capital and infrastructure aid the process of transformation. A steady flow of foreign investment and skills provides a strong legup. For wider impact, the lessons from decentralized experiments must flow to other regions. Mechanisms to ensure such information transfers are essential, but difficult to establish. In a complex reform process, simple and credible commitments may be desirable, but governments also need the ability to roll with the punches.   References Barro, Robert. 1991. "Economic Growth in a Cross-Section of Countries." Quarterly Journal of Economics 105: 407-443. Barro, R. and X. Sala-i-Martin, 1992. "Convergence." Journal of Political Economy 100: 223-25 1. Bateman, D. and A. Mody. 1991. "Growth in an Inefficient Economy: A Chinese Case-Study." World Bank, mimeographed. Beisley, D. A., E. Kuh, R. E. Welsch, 1980. Regression Diagnostics. New York: Wiley and Sons. Byrd, W., ed., 1992. Chinese Industrial Firms Under Reform. Washington, DC: The World Bank. Chen, K., G. H. Jefferson, and I. Singh, 1992. "Lessons from China's Economic Reform." Journal of Comparative Economics 16: 201-225. Chow, Gregory C. 1993. "Capital formation and Economic Growth in China." Quarterly Journal of Economics CVIII (3): 809-842. Dollar, D., 1992. "Exploiting the Advantages of Backwardness: The Importance of Education and Outward Orientation." World Bank mimeo. Glaeser, E. L., Kallal, H. D., Scheinkman, J. A. and Shleifer, A. 1992. "Growth in Cities." Journal of Political Economy 100 (6): 1126-1152. Ho, S.P.S., and R. Huenemann, 1984. China's Open Door Policy. University of British Columbia Press: Vancouver. Hulten, C. R. and R. M. Schwab. 1991. "Is There Too Little Public Capital? Infrastructure and Economic Growth." Paper presented in the American Enterprise Institute Conference 27  28 Explaining Industrial Growth in Coastal China "Infrastructure Needs and Policy Options for the 1990s". Jacobs, Jane. 1969. The Economy of Cities. New York: Vintage Books. Jefferson, G., T. Rawski, and Y. Zheng, 1990. "Growth, Efficiency and Convergence in China's State and Collective Industry." Research Paper Series 1, the World Bank. Jefferson, G., and W. Xu, 1990. "The Impact of Reform on Socialist Enterprises in Transition: Structure, Conduct and Performance in Chinese Industry." Research Paper Series 2, the World Bank. Jefferson, G. and W. Xu, 1991. "The Impact of China's Economic Reforms on Allocative Efficiency in Chinese Industry." Research Paper Series 4, the World Bank. Lal, Deepak. 1990. "The Fable of the Three Envelopes: The Analytics and Political Economy of the Reform of Chinese State Owned Enterprises." European Economic Review 34: 1213-1231. Levine, Ross and David Renelt. 1992. "A Sensitivity Analysis of Cross-Country Growth Regressions." American Economic Review 82 (4): 942-963. Lucas, R. 1988. "On the Mechanics of Economic Development." Journal of Monetary Economics 22: 3-42. Mankiw, N., P. Romer, and D. Weil, 1992. "A Contribution to the Empirics of Economic Growth." Quarterly Journal of Economics CVII: 407-38. (2, Jean C. 1992. "Fiscal Reform and the Economic Foundations of Local State Corporatism in China." World Politics 45 (1): 99-126. Panagariya, Arvind. 1993. "Unravelling the Mysteries of Jiangsu's Trade Regime." World Economv. Population Statistics of China, 1990, compiled by Yasuko Hayase and Seiko Kawamata, Institute of Developing Economies, Tokyo, Japan. Porter, M. E. 1990. The Competitive Advantage of Nations. New York, Free Press.  References ..... 29 Romer, P. M. 1986. "Increasing Returns and Long Run Growth." Journal of Political Economy 94: 1002-1037. Romer, P. M. 1992. "Two Strategies for Economic Development: Using Ideas vs. Producing Ideas. World Bank Annual Conference on Development Economics, 1992. Schankerman, Mark. 1991. "Revisions of Investment Plans and the Stock Market Rate of Return". National Bureau of Economic Research. Working Paper No. 3937. Shleifer, A. 1990. "Externalities and Economic Growth: Lessons from Recent Work." Background paper to the 1991 World Development Report, (BGP #3). Searle, S. R. 1971. Linear Models. New York: Wiley and Sons. State Statistical Bureau of PRC. China Statistical YearBook, various issues. State Statistical Bureau of PRC. 1990. Compendium of Historical Data, 1949-89. State Statistical Bureau of PRC. Statistical Yearbook of Guangdong, various issues. State Statistical Bureau of PRC. Statistical Yearbook of Fujian, various issues. Stockman, A.C. 1988. "Sectoral and National Aggregate Disturbances to Industrial Output in Seven European Countries." Journal of Monetary Economics 21, 387-409. Sung, Y. W. 1991. The China Hong-Kong Connection: the Key to China's Open-Door Policy. Cambridge. Cambridge University Press. Tidrick, G. and Chen, J. eds. 1987. China's Industrial Reform. Washington, DC. The World Bank. Waldmann, R. J. and De Long, J. B. 1990. "Interpreting Procyclical Productivity: Evidence from a Cross-Nation Cross-Industry Panel." Harvard Institute of Economic Research, Discussion Paper # 1495. Walker, Tony. 1993. "Hats Off to the Revolution." Financial Times March 30, 1993, p. 21. Weitzman, M. L. 1970. "Optimal Growth with Scale Economies in the Creation of Overhead Capital." Review of Economic Studies XXXVII: 555-570.  30 Explaining Industrial Growth in Coastal China World Bank. 1991. "The Challenge of Development: World Development Report". Oxford: Oxford University Press. Yusuf, Shahid. 1993. "The Rise of China's Nonstate Sector." World Bank, mimeographed.  Appendix 1 Standard variance components method assumes as the first approximation a growth equation of additive main and interaction effects with zero means and covariances. Let G(tri) denote growth in industry i in region r at time t. Variance of G(tri) can be decomposed to the main effects of time (a(t)), region (13(r)), industry (7(i)), and their respective interaction effects (a(tr), b(ti), c(ri), and e(tri)). Formally, G(tri) = p. + ca(t) + B(r) + T(i) + a(tr) + b(ti) + c(ri) + E(tri) where p, is constant, t= 1,...,n, r= 1,...,nr, i= 1,...,n1. With the assumption of zero covariance, Var(G) can be expressed as the follows: Var(G) = Var(u) + Var(B) + Var(r) + Var(a) + Var(b) + Var(c) +Var(E) or a (G) = 0". + (J26 + 0-2r + 0a2 + bT + 02, + 21 Variance components are estimated by equating observed values of variances to their expected values (Searle, 1971). Let N = ninrnt, Define To = F,,r,i G2(tri) E(T) nnn,A (4' + o. + oy2a + 02, + 02^ + O2b + 02c + 02 = N(,2 + + ++,, 2+ o + + 2c + u2,) T, = G...2/ninnt = (.ErE,G)2/ninrn, = (ninrnjz + 5.n,n,a + En,n,B + Ein,n,T + E,Ea + ,Eib + ,ic + ,riE)2/njnrn, E(TI) = Nj2 + nin,o2 + nin,or2 + n,nara2, + nLo2. + n,02b + n,o2, + o2, T2 = ,G. .2/ninr = ,(ErEjG)21nin, = ,(ninji + nin,c + rn1B + rinrr + niEra + n, b +E xc + EEje)I/njn, E(T2) = Nis2 + Noa2 + nin,og + n,nao2, + nnIa2, + n,n,o2b + n,oa2 + nao2, Similarly, let T3 = :rG 2/n,ni = r(EEjG /n,n 31  32 Explaining Industrial Growth in Coastal China E(T3) N,2 + nrn,a2 + Na&B + ntnra