direct and indirect impact - CiteSeerX

THE DIRECT AND INDIRECT IMPACT OF FOREIGN DIRECT INVESTMENT ON PRODUCTIVITY GROWTH IN INDIAN MANUFACTURING

Rashmi Banga

•

ABSTRACT: Studies have found the impact of foreign direct investments (FDI) on productivity growth to be firm-industry-host economy specific. However, the impact of FDI may differ with the existing market conditions in the economy and also with respect to the source of the FDI. The paper examines the impact of FDI on productivity growth of the Indian industry in the period of low market demand conditions, i.e., 1995-96 to 1999-2000. It also compares empirically the spillover effects of Japanese and U.S. FDI on the total factor productivity growth both at the industry and at the firm level. Panel data estimation techniques are employed.

JEL CODES: F23, L22, L60, O33 KEY WORDS: Productivity Spillovers, Japanese and U.S. FDI, Total Factor Productivity Growth, Technological Gap

1. INTRODUCTION

The Indian government in the post-liberalisation period has slowly but steadily •

I am extremely grateful to Prof. B. N. Goldar (Institute of Economic Growth), Dr. A. Bhattacharjea (Delhi School of Economics) and Prof. N.S. Siddharthan (Institute of Economic Growth) for their valuable suggestions and comments, which have helped me to improve my work substantially. E- mail: [email protected]

1

tried to facilitate the inflow of foreign direct investment (FDI) into different sectors of the economy. FDI was sought because it augments investible resources and, more important, it improves technological standards, efficiency, and competitiveness of domestic industry. FDI influences the productivity growth of an industry directly through infusing new capital in the industry and by improving the average skills and efficiency levels of the industry. It is also associated with bringing in of "relatively" latest technology into the industry since markets for technology are imperfect, which makes the transaction costs for sales of technology by the MNCs to outsiders high (Buckley and Casson, 1976, Caves 1996, Teece 1981). Indirectly, FDI is expected to influence the productivity levels in an industry through external effects or "spillovers". Caves (1974a) emphasised three potential spillover benefits as additional gains from FDI. These were improvements in allocative efficiency; higher level of technical or X-efficiency; and diffusion of technology and knowledge to local firms. Other positive effects of FDI emphasised by various studies on productivity of domestic firms are via competitive pressures which eventually force the domestic firms to become more effic ient; learning by doing and diffusion of knowledge through demonstration effects, labour turnover or reverse engineering. (e.g., Globerman 1979, Blomstrom and Wolf 1994, Djankov and Hoekman 2000). The indirect effects of FDI, however, may not always be positive. There are empirical studies suggesting that the effects of foreign presence are not always beneficial for local firms (Kokko 1994, Aitken, Hansen and Harrison, 1994, Haddad and Harrison, 1993). At one level, the new entry increases competitive conditions, which should induce local firms to replace inefficient technologies and organizational practices through imports of capital goods and R&D effort thereby increasing overall industrial productivity. However, if the new entrants cut into the market of local firms then it could result in decline in average industrial productivity as local firms are compelled operate at sub optimal scales (Kokko 1994, Kokko et al., 1996, Haddad and Harrison 1993).

The literature on the impact of FDI suggests that the exact nature of the impact of foreign presence depends on many factors. These include the technological levels

2

prevailing in the industry, the learning capabilities of the firms and the absorptive capacity of the host economy, which determines the rate of technical diffusion of the technology (Kokko 1994, Haddad and Harrison 1993, Aitken and Harrison 1999, Kokko et al., 1996). These factors appear to be firm-industry-host economy specific in nature. In the case of the Indian economy Goldar (1995) and Kathuria (2000, 2001) have studied the impact of FDI on productivity growth. The period of analysis of these studies is 1987-88 to 1989-90 and 1989-1990 to 1994-95 respectively. However, as yet few studies exist which cover the second half of the post-liberalisation period, i.e., 1995-96 to 1999-2000.

The present study differs from the existing studies with respect to its scope. It tries to take into account factors other than those that are firm-industry-host economy specific. Firstly, it focuses on the impact of market demand conditions prevailing in the economy during the period of analysis. If the period of analysis is one of low state of demand in the economy, FDI may have a depressing impact on the productivity growth both at the firm and at the aggregate level. A low state of demand in the economy would imply that foreign firms have negative productivity spillovers. This happens because with the entry of foreign firms in depressed demand conditions the local firms will lose their market shares and operate at the sub optimal levels of output. This lowers the TFPG at the aggregate levels since the number of domestic firms in an industry is always far greater than the foreign firms.

Another aspect studied in the paper is the influence of the source-country of the FDI. Hardly any of the existing studies on spillovers of FDI have tried to disaggregate the impact of FDI from different source countries. Foreign direct investments come from different sources. These sources of FDI are likely to have different levels of technology, different modes of transferring technology and may follow different output and investment strategies. They sometimes may also have different industrial patterns. It is therefore possible that they have differential impact on the productivit y of domestic firms. The two source countries selected for the analysis are Japan and U.S. It is generally believed that Japanese firms behave differently from other firms, either because of their protected domestic base or because they have different financial and institutional structures. (Graham and Krugman 1995). The Japanese FDI are largely undertaken by small and medium sized 3

firms while large firms undertake US FDI. This implies that the level of technology at which they operate would differ. The two source countries are also expected to differ with respect to their corporate governance, financial structures and output & investment strategies. While short-term profits are important for US firms, it is the long- term profits that the Japanese firms aim at. Moreover, the FDI from these two sources may also choose different industrial pattern. US FDI in manufacturing is usually undertaken in most technologically sophisticated industries with not yet standardised products that are more capital-intensive in nature while Japanese FDI generally enter industries that are less capital-intensive producing standardised products that are less technology-intensive. Further, the mode of transfer of technology by the Japanese firms is termed as orderly transfer of standardised production that differs from the American "reverse-order" transfer of technology (Kojima 1978).

Given these differences in the nature of the two FDI it is possible that the productivity spillovers from the two sources may also differ.1 The paper attempts to analyse the impact of Japanese and US FDI along with the impact of aggregate FDI on the productivity growth in the Indian manufacturing. The analysis is carried out at the industry level using a panel data for the period 1995-96 to 1999-2000 covering 74 three-digit level industries. At the firm level the productivity spillovers are examined for the period 1993-94 to 1999-2000 covering 153 domestic firms in 26 three-digit level industries. The paper is organised into seven sections. Section 2 discusses the literature related to the direct and indirect impact of FDI. The analytical framework and the hypotheses of the study are presented in Section 3; Section 4 describes the data and the sample, construction of variables and methodology used in the paper; Sections 5 and 6 presents the results at the aggregate and firm levels respectively. Section 7 concludes the study.

2. LITERATURE RELATED TO THE IMPACT OF FDI 1

Most of the studies have analysed the impact of aggregate FDI on the productivity of the industrial sector in the host economy. Very few studies have considered FDI from a particular source . One such study has been done by Okamoto, (1999) where the impact of Japanese FDI on productivity of US manufacturing sectors is studied.

4

Literature related to the direct impact of FDI has emphasised that FDIs, due to the resources associated with it and the attributes embedded in it are expected to provide a package of tangible and intangible wealth-creating assets. These assets become available directly for use in productive activities in the host countries and are further amplified by externalities and spillovers that strengthen the resource base and production capabilities in developing economies. The financial capital generated, transmitted and invested by the foreign firms is one of the principal contributions of FDI's to a country's output or productivity growth. This financial capital is generated internally since not all of their profits are distributed to shareholders as dividends; some are retained and reinvested, adding to firm's capital stock. The very pres ence of FDI in an industry is also expected to improve the average productivity and skill levels of the industry since MNCs are associated with higher efficiency levels due to their ownership and internalisational advantages (Caves 1974b, Dunning 1973).

Another direct impact of FDI, which goes far in promoting productivity growth in industry, is that FDI leads to an access to technology particularly through imports of capital goods. Such technology is exported through FDI to wholly owned foreign affiliates and joint ventures. This way the foreign firms maintain their competitive advantage by transferring their most recent technology to their affiliates, while selling or licensing older technology to others. For developing countries, therefore, FDI may be the only way to gain access to latest or "relatively" latest technology.

The empirical evidence on the impact of FDI is however, mixed. Some studies e.g., Caves (1996), Globerman (1979), Blomstrom and Wolf (1994), Djankov and Hoekman (2000) have found that FDI have a positive or weak positive effect on the productivity levels. On the other hand, there are others e.g., Kokko (1994), Kokko, et al., (1996), Aitken, and Harrison (1999) and Haddad and Harrison, (1993), who have found that foreign firms have negative effects on the productivity performance of the domestically owned firms.

5

Harrison (1994) suggests that in imperfect competitive markets entry by foreign investors cause domestic incumbents to lose their market shares, impeding their ability to attain scale economies. Gagelmann (2000), in his study using firm level Polish data concludes that higher foreign presence within an industry affects local firms negatively. However, the effects differ between groups of firms and groups of industries. FDI has a negative effect on the performance of the most productive local firms in high competition industries. By contrast, the effect on the least productive firms in low competition industries is positive. Grether (1999) in his study for Mexican manufacturing firms found that foreign capital is found to have a positive influence on productive efficiency at the plant level but contrarily to crossindustry studies it does not lead to significant spillovers at the sector level. Tsou and Liu (1998) used Taiwan manufacturing census data and concluded that in sectors where technology gap between foreign plants and locally owned plants is low a positive and statistically significant spillover effects occur. However, if the technology gap is large, there is no strong sign of spillovers. The above studies are just a few in the vast literature on impact of FDI. However, these studies together do establish the fact that impact of FDI is industryfirm-host economy specific. For the Indian economy, Goldar (1995) and Kathuria (2000, 2001) have studied the impact of FDI on productivity growth of Indian firms. The study by Goldar examines the impact of technology acquisition via FDI on TFPG for the period 1987-88 to 1989-90. His results do not reveal any strong positive effect of the technology acquisition accompanying FDI on productivity growth. The study of Kathuria indicates that there exists positive spillovers from the presence of foreignowned firms, but the nature and type of spillovers vary depending upon the industries to which the firms belong and also on the R&D capabilities of the firms.

However, as mentioned earlier these studies have not taken into account the impact of market demand conditions while studying the impact of FDI on productivity growth. Their results therefore may be period specific and valid for a state of relatively high or low level of market demand. Again, hardly any of the studies in the literature have tried to disaggregate the spillovers of FDI from different sources and compare them. This paper is an attempt in this direction. It tries to bring out the intersource differences in the impact of FDI on productivity levels in the Indian 6

manufacturing along with the impact of low demand conditions on the productivity growth.

3. THE ANALYTICAL FRAMEWORK AND THE HYPOTHESES

The Intangible asset theory shows that foreign firms possess superior technological and organizational practices in comparison to local firms in developing economies. The entry of foreign firms and implications for industrial productivity operate at several levels. At one level, the new entry increases competitive conditions, which should induce local firms to replace inefficient technologies and organizational practices through imports of capital goods and R&D efforts and in turn increase overall industrial productivity. However, if the market expands at a lower rate than increase in capacity due to new entry, new entrants, i.e., the foreign firms in this case, can cut into the market shares of local firms. In such a situation it could result in decline in average industrial productivity as local firms are compelled to operate at sub optimal scales. On the other hand, if the number of firms increase under the increasing demand conditions without any loss of scale, the increase in the number of firms could result in external economies at the industry level which shifts cost curves down for all the firms (Rotenberg and Saloner, 2000). Another implication of the entry of foreign firms on productivity is spillover effects. If local firms, through deliberate effort or spillover, bring into use superior practices it would improve overall industrial productivity (Grossman and Helpman 1991, Branstetter 2000, Kokko 1994). Local firms would be able to internalize these spillovers and absorb them effectively if they make technological efforts in terms of investing in R&D and adapting imported technologies efficiently. (Kathuria 2000)

Apart from the market conditions, the source country of the FDI can also have a differential impact on the productivity growth of the domestic firms. The debate about whether significant differences exist between Japanese and US FDI has a long if not altogether distinguished history. Many of the arguments made by early commentators, such as Kojima (1978, 1986) distinguish between the trade-enhancing

7

nature of Japanese FDI and the trade-undermining characteristic of US FDI. Ozawa (1979) emphasized the importance of relative factor endowments in driving Japanese FDI. However, these studies have not withstood the test of time and empirical examination (for criticisms see Hill 1988; Hill, 1990; Ramstetter, 1987).

Further, the presence of Japanese FDI in export-oriented manufacturing in different parts of Asia in the last decade appears to be little different from US FDI in its motivations. Along with this the importance of global trends to which all firms must respond regardless of home base, inspires caution in any attempt to make apriori comparisons between Japanese FDI and US FDI. However, the characteristics of Jap anese corporations displayed in their domestic operations are regarded as unique (Aoki 1988; Womack, Jones and Roos 1991) and which, if they are changing at all, are doing so only slowly (Yamamura 1990; Yamamura 1994). This suggests that it is likely that the operations of Japanese FDI will continue to differ from those of US FDI. Some of these unique dimensions of Japanese corporations are intercorporate relations, and the relations between corporations and the home government. Some of the unique production techniques like just-in-time sourcing, total quality management, FMEI, quality control circles, ringi system etc. if replicated in overseas affiliates, may benefit the host economy. Others, such as the keiretsu relations that link assemblers and supplie rs may not lead to any potential benefits, in fact may lead to negative influences in the host economy. HYPOTHESES:

Based on the analytical framework and literature on the characteristics of Japanese FDI that are unique in nature, we derive some hypotheses regarding the impact of aggregate FDI, Japanese FDI and US FDI on the productivity growth in Indian manufacturing, both at the aggregate and the firm level.

HYPOTHESIS 1: FDI is expected to have a negative impact on total factor productivity growth of the industries if there exists low market demand conditions. 8

The impact of FDI at the industry level will compose of both direct as well as indirect effects of the FDI. It is expected that if the market demand conditions are low, i.e., if the perio d of analysis is one of the slow down of industrial growth rates then the indirect effect of the FDI will be negative. This is expected because if the market expands at a low rate than increase in capacity due to new entry of foreign firms then foreign fir ms will cut into the market shares of local firms. This could result in a decline in the average industrial productivity as local firms operate at sub optimal scales. HYPOTHESIS 2: Japanese FDI are expected to have larger positive spillover effects on total factor productivity growth of domestic firms as compared to US FDI

This hypothesis is derived on the basis of the unique characteristics of Japanese FDI that differentiates them from US FDI. Some of these differences lead to differences in the type of technology brought in by the Japanese FDI and US FDI. The two also differ with respect to their mode of technology transfer and to some extent in their industrial pattern. The unique characteristics of Japan's transfer of technology, as pointed out in the literature by many studies, is that it is an orderly transfer of technology of standardised production which begins in those industries where technical gap between the providing and receiving countries is the lowest. This should make transfer of technology easier and its spread effects larger. On the other hand, US FDI in manufacturing is expected to be undertaken in the most technologically sophisticated industries, with not yet standardised products that are more capital than labour intensive and rank on the top of the comparative advantage of the home economy. Kojima calls this "reverse-order" transfer of technology. This implies a large gap between the existing technology of the host developing country. This leads to the creation of "enclaves" and makes spillovers of the transfer of technology to the domestic firms difficult.

Furthermore, the mode of technology transfers starkly differs between Japanese and US FDI. It is often quoted that Japanese production models draws its strength more from human related dimensions of engineering techniques, workplace practices and corporate culture than through R&D or technology imports of capital 9

goods or other disembodied technology imports. Organisational innovations and management practices have in recent times come to be accepted as one of the major aspects of technological development for enhancing productivity and growth.

Apart from the technology related differences Japanese and US firms also differ with respect to their corporate financial structures (Banga 2002) and corporate governance. This makes them follow different output and investment strategies that might have differential impact on the productivity growth in the host economy. For example, Japanese firms do not follow aggressive output strategies that aim at shortterm profitability and so may not engage in price wars. They are therefore less likely, as compared to the US firms, to cut into the market shares of the local firms at a time of low market demand conditions. US firms, on the other hand, are more likely to engage in aggressive output strategies and compete with the domestic firms for the market share at a time when market demand conditions are low since the share of management in the ownership is low. This makes short-term profitability and returns to share holders an important objective for the management.

In the specific case of India, the industrial pattern of Japanese and US FDI also differ to some extent. US FDI is concentrated in Chemicals and Pharmaceuticals industries, which involve process technology. This makes their technology less susceptible to the reverse engineering or imitation unlike the technologies in industries like Auto-Ancillary and Electrical where Japanese have a stronghold. These differential characteristics of the two FDI sources make Japanese FDI more susceptible to leakage or spillovers as compared to their oversea competitors. The Japanese firms are therefore expected to have larger positive spillover effects as compared to the U.S. firms after controlling for the industry-specific and firm-specific effects.

HYPOTHESIS 3: The initial productivity gap between domestic firms and Japanese firms will be positively related to the TFPG of domestic firms and therefore support

10

the "catch-up" theory, however the initial productivity gap between domestic firms and US firms may not be positively associated.

There are conflicting views on the relationship between the rate of diffusion and the gap between the level of technology in MNC’S home country and the country hosting FDI. In Findlay (1978) and Wang and Blomstrom (1992) the rate at which new technology is diffused is an increasing function of the gap i.e., the larger the gap is the greater the potential for “catch-up”. On the other hand, some studies argue that spillovers are negatively related to the technology gap between the relatively “backward” host country and the “advanced” home country, due to the fact that the superior technology may not be appropriate for the backward country.

Taking an evolutionary approach, Perez (1983) argued that the ability of indigenous firms to “catch-up” depends on their level of technical competence. This competence is characterized by “path-dependency” and the absorptive capacity of indigenous firms that depends on their past accumula tion of technology. The influence of these multiple factors suggests a non-linear relationship between spillovers and the size of the technology gap. Spillovers increase as the technology gap widens up to a certain critical level. Beyond this level spillovers begin to fall because technical competence by indigenous firms will be too low to exploit fully the technical opportunities arising from foreign presence and at still higher level may become negligible or even negative.

It is observed that US FDI in manufacturing is usually undertaken in most technologically sophisticated industries and by large firms. We therefore expect the initial technological gap between US and Indian firms may not be within the critical levels. However, Japanese FDI generally enters industries that are less capitalintensive producing standardised products that are less technology-intensive and it is undertaken mainly by small and medium sized firms. The technological gap between Japanese and Indian firms is therefore expected to be within the critical level. The "catch-up" theory can be expected to hold for the Indian firms vis -à-vis Japanese firms but not vis -à-vis US firms.

11

4 . DATA, SAMPLE, VARAIBLES AND METHODOLOGY: INDUSTRY LEVEL AND FIRM LEVEL ANALYSIS

DATA AND SAMPLE

In order to estimate the productivity growth rates at both industry and the firm level, we have collected data from Capitaline Package. This is supplemented with data taken from various issues of Annual of Survey of Industries (ASI), various issues of National Accounts Statistics and some publications of Ministry of Industry. Industry level analysis uses a balanced panel of 74 three-digit level industries for the years 1995-96 to 1999-2000. Firm level analysis is based on data of 153 firms for the year 1993-94 to 1999-2000 across 25 three-digit level industries. At the industry level all available manufacturing industries in the Capitaline have been selected and have been aggregated to match the ASI industrial classification. As regards the criteria used for selecting industries at the firm level analysis, only those industries are selected where at least 15% of the total equity invested in the industry is from Japan and/or US. Not much spillover effect is expected in the industries where this is less than 15% of foreign equity. A matching of industries from Capitaline to ASI is done so as to construct data on variables that are not provided by the Capitaline. All the variables used in the panel data estimation of productivity are measured in constant prices of 1993-94. Deflation of output and inputs has been done with help of suitably constructed deflators.

The earlier studies estimating production function for the Indian manufacturing have used the wholesale price indices to deflate the series on output and inputs of the firms to arrive at the constant prices. However, we have used the actual prices of the major outputs and inputs of the firms to arrive at indices for deflating output and input series of the firms. This is considerable methodological improvement over earlier studies. The construction of variables is taken up next for discussion.

VARIABLES

12

There are two sets of variables used for the analysis: (a) variables for the estimation of production function for deriving productivity estimates and (b) variables used in the regression analysis explaining variations in productivity growth. These are discussed in that order.

(A) VARIABLES FOR PRODUCTION FUNCTION ES TIMATES:

OUTPUT AT INDUSTRY LEVEL: The data on output of the industries (sales plus net addition to inventory) are collected from Capitaline package. These are deflated by the wholesale price indices of output for that year available from CMIE data.

OUTPUT AT FIRM LEVEL: The Capitaline package provides data on output of the firms In most of the earlier studies on Indian manufacturing the wholesale price indices are used as deflators for output. However, Capitaline package provides the data on the major outputs of the firms along with their prices. Weighted output price indices are constructed using the prices of two major outputs of the firms. The value (price*quantity) of the output is used as the weights in the series.

INTERMEDIATE INPUTS AT INDUSTRY LEVEL: The data on raw materials used in industries are deflated using the weighted price indices, which have been constructed using wholesale price indices, the weights being taken from an input-output matrix. The input coefficients of ten major inputs along with the coefficients for energy have been used as weights to construct the input deflator for raw materials.

INTERMEDIATE INPUTS AT FIRM LEVEL: Capitaline package also provides data on the major inputs used by the firms along with their prices. The total raw-materials consumed by the firms is deflated by the weighted input price series, which is constructed using the actual prices of the inputs The total cost of the inputs is used as the weights

LABOUR AT INDUSTRY LEVEL:

13

Capitaline package does not give data on employment. The data on employment is constructed using the total emoluments and number of employee data from matched industries in ASI. Using this data the wage-rate in the industries is estimated. Since data is available only till 1997-98, the growth rate of cost of living index for the years 1997-98 to 1999-2000 is applied to the wage-rate from ASI. Data on cost of living index are collected from CMIE. The data on the total employee cost is collected from the Capitaline and using the wage-rate, the figures for total number of employees are arrived at.

LABOUR AT FIRM LEVEL: The data on total employee cost of the firms is collected from the Capitaline and the series on number of employees is constructed using the wage-rate in corresponding industries estimated from ASI.

CAPITAL SERIES AT INDUSTRY LEVEL: Separate series on Plants and machinery and Land and Buildings in industries are collected from Capitaline and separate deflators are constructed for the two series. The perpetual inventory method is followed. To arrive at the capital stock in the base year, i.e., 1995-96, a revaluation factor is estimated using the age-distribution of capital in the past 15 years. Data on fixed capital and depreciation are taken from ASI for matching industries. Using these data, gross investment of different years are corrected for price change to obtain value of capital stock at 1993-94 prices for 199596. Comparing this to the gross value of fixed assets the revaluation factor is computed. This revaluation factor is applied to the benchmark figures to get benchmark capital stock. Then, deflated gross investments in the consecutive years are added. The series on plant and machinery is deflated by the wholesale price index for industrial machinery, collected from CMIE data and the series on construction is deflated by using deflators constructed by applying weights to different components of construction, i.e., bricks, cement, etc.

CAPITAL SERIES AT FIRM LEVEL: The methodology used to estimate capital is that used by Srivastava (1996). However, the deflators used for deflating different series of capital are more disaggregated. Capital stock is taken to consist of Plants and Machinery, Land & 14

Building and other Fixed Assets. Two separate series of capital are constructed i.e., one for Plants and Machinery along with other fixed assets and the other for Land & Building. These are deflated separately to arrive at estimates of capital stock in the base year i.e., 1993-94 for each firm. Data on Gross capital formation in plants & machinery and construction at current and constant prices are collected from NAS and an implicit deflator is arrived at. Applying this implicit deflator, capital stock in the year 1993-94 is estimated. However, since in the base year the firm's asset mix is valued at historic cost, the value of capital at replacement cost for the current year is arrived at by revaluating the base year of capital. Implicit deflators are constructed for last 15 years in case of plants and machinery of the firms and for last 25 years or the date of incorporation of the firm for construction in the firms.

A revaluation factor (as used by Srivastava) is then applied to each series to obtain capital stock at replacement costs at current prices. Deflating these values we arrive at capital stock in real terms for the base year. Subsequent years investment is then added i.e., Gross fixed assets t - Gross fixed assets t -1 to the capital stock existing at every time period using the perpetual inventory method. The capital stock series is hence arrived at for the firms. FUEL AND POWER AT FIRM LEVEL: Energy is an important input in firms’ output. Capitaline provides data on expenditure on fuel and power. Weighted price indices are constructed to deflate the expenditure on fuel and power. Wholesale price indices for electricity for industrial purposes and furnace oil from CMIE publications are used. Weights used are the firms’ expenditure on oil and power. Since the data for some of the series is available only till the year 1997-98 extrapolation has been used to arrive at the index numbers for recent years.

(b) VARIABLES USED IN REGRESSION ANALYSIS

Earlier studies have found that the impact of FDI on productivity growth is firm-industry specific. Therefore three different sets of variables have been used, i.e., the spill variables, 15

the industry-specific variables and the firm-specific variables. The spill variables show the extent of presence of foreign firms in the industry. The aggregate analysis uses the industryspecific variables along with the spill variable. While in the firm-level analysis uses firmspecific variables along with the spill variables. An attempt is made to control for the industry-specific effects at the firm level also.

Many studies have used the sales of foreign firms to total sales in the industry as a spill variable. However, this is the market share of foreign firms, which may not truly capture the extent of presence of foreign firms in an industry. Wholly owned subsidiaries of foreign firms may have larger spillovers but in large industries may have smaller market share. Also, the market share of foreign firms is expected to be some negatively associated with the TFPG of the domestic firms in a low market demand condition therefore the spill variable used in the study is the proportion of foreign equity invested to total equity invested in the industry. Foreign equity invested by Japanese and US firms to total equity invested is taken as a spill from Japanese and US firms respectively. The type of technology imported by Japanese and US firms may differ. Therefore, interaction terms are considered i.e., technology imported by the Japanese firms* proportion of Japanese equity in the industry and technology imported by the U.S. firms* proportion of U.S. equity in the industry. The variables representing spillover effects are:

SPILL-VARIABLES:

a) Foreign equity as a ratio of total equity invested in the industry (FOREQ) b) Japanese Equity as a proportion of total equity invested in the industry (JE) c) US Equity as a proportion of total equity invested in the industry (USE) d) Technology imports by Japanese firms*Japanese equity in the industry (TECHM*J) e) Technology imports by US firms*US equity in the industry (TECHM*US)

The productivity growth of the industry is found to be related to some industryspecific variables like capital intensity in the industry; R&D intensity of the industry, imports of knowledge capital goods in the industry; outward -orientation of the industry and

16

policy regulations controlling the industry. To analyse the impact of FDI on the productivity growth of the industries it becomes important to control for these variables. INDUSTRY-SPECIFIC VARIABLES: a) Capital imports by the industry (IMPCAP) b) R&D intensity of the industry (R&D) c) Export-Intensity of the industry (XI) d) Import of knowledge capital in the industry, i.e., the payments for disembodied technology (IMPDT) e) Effective Rate of Protection in the industry (ERP) f) Capital-Labour Ratio in the industry (K/L) g) Skill intensity, i.e., number of high-salaried employees / total number of employees h) Size of the Industry, to tal fixed assets / number of firms in the industry i) Concentration Ratio of the industry. (CR4)

The productivity growth of the firms is a dependent on both industry-specific as well as firm-specific variables like size of the firms, age of the firms, R&D intensity of the firms, etc. To control for firms -specific variables the following variables are considered:

FIRM-SPECIFIC VARIABLES: a) Size of the firm i.e., log of sales of the firms (SIZEF) b) R&D Intensity of the firm i.e., R&D expenditure/sales (R&DF ) c) Export Intensity of the firms (XI F ) d) Capital-Intensity of the firm (KIF ) e) Gap between initial productivity levels of domestic firms and average productivity levels of foreign firms (GAP) f) Gap between initial productivity levels of domestic firms and average productivity levels of Japanese firms (GAPJ) g) Gap between initial productivity levels of domestic firms and average productivity levels of US firms (GAPUS)

The data on most of the variables are collected from Capitaline. The data on foreign equity invested for the years 1993-94 to 1995-96 has been constructed using 17

ratio of the dividends paid in foreign exchange by the firms to total dividends paid. This may also include the dividends paid to foreign institutional investors. However, it is not expected to be large for this period. The ERP series estimated by NCAER for the years 1995-96 to 1998-99 has been used.

METHODOLOGY The methodology adopted to estimate the productivity growth at industry as well as at the firm level is the "time-variant firm specific" technical efficiency approach, first introduced by Cornwell, Schmidt and Sickles (1990). This methodology for estimating TFPG has also been used by Srivastava (1996) and Kathuria (2000) for estimating TFPG at the industry and the firm level respectively. Three inputs based Cobb-Douglas production function is estimated at the industry level and four inputs based Cobb-Douglas production function is estimated at the firm level. For industry level analysis different production functions have been estimated for capital-intensive industries and labour-intensive industries (See Appendix 1). It was felt that estimating the same production function for all industries, irrespective of their capital intensities, may not be right and may affect the TFPG comparisons across industries

The impact of FDI at the aggregate level is estimated using the panel data estimation techniques. The choice of the model is based on the LM test, which tests whether ordinary least squares is appropriate or not, and on the Hausman specification test, to choose between random-or-fixed effects framework. These two tests suggest that a random-effects model is most appropriate. Fixed-effects estimation assumes that firm productivity growth is constant overtime. The problem with the assumption is that we want to examine changes in productivity arising from increased competition. The random-effects model avoids this assumption, but assumes that productivity shocks at the firm level are uncorrelated overtime.

At the firm level an average of the seven years is taken and linear regressions are run. This is done so as to have a longer period of analysis that smoothes out the

18

impact of the year-to-year fluctuations in demand. Also, the spill-variables at the firm level comprise of foreign equity existing at the industry level and it is felt that its impact on TFPG of the domestic firms may be felt with different lags by different firms.

The impact of FDI on TFPG in the industry is expected to vary according to the prevailing market demand conditions in the economy and the source country of the FDI. To capture this impact it becomes important to control for other variables that may affect the TFPG of the industry. TFPG is therefore taken as a function of the following variables.

TFPG of ind ustry = f ( FDI, industry specific variables, policy variables) + error term

While at the firm level to study the impact of FDI on TFPG of the firm we control for industry-specific as well as firm-specific effects. The spill variables show the impact of FDI on the TFPG of the firms.

TFPG of firm = f (spill variables; firm-specific variables; and industry-specific variables) + error term

The period of analysis considered at the aggregate industry level is that of an ind ustrial slow down of the Indian economy. It is therefore expected to lead to depressed market conditions. While, at the firm level averages are considered for the period 1993-94 to 1999-2000. This leads to on an average a stable market condition for the entire period.

5. IMPACT OF FDI: INDUSTRY LEVEL ANALYSIS DIRECT IMPACT 19

It was only in 1990s that India experienced a significant inflow of FDI (see Table 1). Foreign direct investment during the decade has been about USD 18 billion, and a much larger quantum stands approved. This rise in FDI inflows is expected to increase the productivity growth of the industry directly through the increased capital invested in the industry and by improving the average skills and efficiency levels of the industry. It may also lead to an absolute increase in productivity enhancing expenditures like expenditures on R&D, import of better technology, etc in the industry if such expenditures are higher in foreign firms. However, indirect effects of FDI also accompany these direct effects. The net impact of FDI therefore may not always be positive on the productivity growth. An analysis of the impact of FDI at the aggregate level will encompass both direct and as well as indirect effects of FDI.

EMPIRICAL RESULTS: INDUSTRY-LEVEL ANALYSIS

To analyse the impact of FDI at the aggregate levels we first compare the average levels of TFPG, skill-intensities, R&D intensities, export-intensities and import intensities in three categories of industries for the period 1995-96 to 1999-2000. These categories of industries are all industries taken together, industries with FDI, i.e., industries where at least five percent of the total equity invested is foreign, and in industries without FDI, i.e., where this proportion is less than five percent (Table 2). The results show that the average TFPG of all industries during this period was 1.5%. TFPG in industries with FDI is found to be higher than the industries without FDI. However, not much of a difference is seen in the average skill intensities of the industries and the average export-intensities between the two groups of industries. The R&D intensities and the imports of embodied and disembodied technology is found to be higher in the industries with FDI.

Since the period of analysis, i.e., 1995-96 to 1999-2000 is one of an overall slow down of the economy following a negative agricultural growth due to adverse weather (FY96), the Asian financial crisis during 1997-98, the international sanctions following India's nuclear tests in 1998 and considerable volatility in world commodity and oil prices during 1998-2000; the TFPG for the different categories of industries has also been compared for different years separately (Table 2). It is found that the

20

TFPF for all industries declined from 2.4% to 0.07% over the period 1995-96 to 199798 and thereby improved. However, it is interesting to note that industries with FDI enjoyed a high growth rate of TFP in the year 1995-96 (4.1%) but this steadily declined in the following years and reached to 0.2% in the year 1999-2000. The TFPG in the industries without FDI improved steadily during this period and it increased from -0.02 in 1995-96 to 2.7% in the year 1999-2000.

To analyse this trend in the TFP growth rates industries with FDI have been further divided into industries with very high proportion of foreign equity invested as a proportion of total equity invested (i.e., >25%) and industries with less than 25% of foreign equity invested (Table 3). The average TFPG in the industries with high amount of FDI was found to be -0.01% per annum in this period while the industries with FDI between 5% to 25% was found have a high growth rate of 4.00%. This implies that industries where the share of FDI is very high experienced a decline in the TFPG. Two possible explanations for this can be put forward. Firstly, it has been argued that FDI largely enters industries that have high concentration ratios and where the domestic firms enjoy monopoly profits due to barriers to entry for the new firms. 2 However, foreign firms with their access to large and inexpensive capital resources, better technology and higher skill levels break through these barriers and reduce the monopoly profits leading to a decline in the productivity levels of the domestic firms in these industry.

Secondly, it can also be argued that the time of entry of FDI in the industry can have an important impact on the productivity growth of the industries. At a time when the market is expanding entry of foreign firms in the industry may lead to an improvement in the TFPG of the industry since the domestic firms under the competitive pressures may be forced to improve their efficiency levels. However, if the market is not expanding or in the period of recession foreign firms may negatively affect the market shares of the domestic firms, which are unable to produce optimal levels of the output. This can lead to an overall decline in the TFPG of the industry3 . The period 1995-96 to 1999-2000, witnessed a cyclical downturn of industrial growth 2

See Caves 1974 (b) A study by Catherine Y Co (2001) shows that the entry of FDI affects the industry performance positively for industries with "low" levels of concentration. Beyond a "critical" level, the competitive

3

21

that started in 1996-97 and was prolonged by the twin effects of the East Asian crisis and post-Pokhran sanctions. The recovery in the industrial output growth started only during the year 1999-2000. 4 During this period of industrial recession we therefore find industries with higher share of FDI are the ones with lower TFPG.

The empirical results of panel data analysis using ordinary least squares and a random -effect specification are presented in Table 4. The TFPG of industry is taken as the dependent variable and the principal explanatory variables are FOREQ, USE and JE, after controlling for the industry-specific variables. Due to high correlation between FOREQ and JE and USE separate equations are estimated separately. The results for FOREQ are presented in columns (1) and (2). The estimated coefficients of FDI is negative and statistically significant for both the specifications, suggesting that, as predicted, the impact of FDI on TFPG is negative if the market is expanding at a lower rate. The greater the foreign participation in an industry at a time when low demand conditions prevail the greater will be the loss of market-share for the local firms and result in sub optimal levels of production and an overall decline in the industrial productivity. The estimated equation controls for a number of industryspecific effects like the capital-intensity of the industry; the existing skill levels in the industry; the extent of import of technology in the industry; the existing market structure in the industry; the size of the industry; the export-intensity of the industry and the existing policy regime in the industry.

The capital intensity of the industry along with the import of disembodied technology into the industry and the existing market structure in the industry also to have significant impact on the TFPG. The TFPG is found to be higher for capital intensive industries and for industries that have higher imports of disembodied technology. The concentration ratio of the industry is negatively related to the growth of productivity. 5

The OLS and random-effect specifications estimated for JE and USE are presented in column (3) and (4) of Table 4. The results show a difference in the effects of FDI predominates.See also Kokko 1994 and Haddad and Harrison 1993. 4 See Table 1 5 The R&D variable was dropped from the regression as it including it gave a very low explanatory

22

impact of Japanese FDI and US FDI on the TFPG of the industry. Japanese FDI has a significant positive impact while US FDI has a negative impact on the productivity growth. These differences arise probably because the two source- countries come with different levels of technology, have different modes of transferring technolo gy and follow different output and investment strategies. Japanese firms are less likely to adopt aggressive output strategies like price wars to capture market share in the host economy since they aim at long-term profitability. They are also internationally vertically integrated, i.e., their "up-stream" and "down-stream" production processes are located in different countries. This implies that the reason for them to enter an industry need not necessarily be the domestic market. However, the US FDI are more internationally horizontally integrated and more domestic market-oriented. In a low state of demand they would not necessarily follow aggressive output strategies that can adversely affect the TFPG in local firms.

6. EMPIRRICAL RESULTS: FIRM-LEVEL ANALYSIS

To examine the productivity spillovers from different sources of FDI a firm level analysis is undertaken. Table 5 compares Japanese, US and domestic firms. This comparison is undertaken with respect to their TFPG, export intensities, import intensities and R&D intensities. The period of analysis is now extended to seven years, i.e., 1993-94 to 1999-2000. This has been done in order to include the periods of industrial growth in the analysis so as to capture those indirect effects of FDI that are not period specific. The total number of the firms considered is 153; out of which around 51% of firms are domestic firms. The results show that the TFPG has been highest for the Japanese firms during this period. The TFPG of Indian firms has been found to be higher than that of US firms. This shows that all foreign firms may not be alike in their operations. The R&D intensities are higher for the Indian firms. This is expected since much of the R&D activities by the foreign firms are done in their home countries. The imports of disembodied technology and embodied technology are higher for the Indian firms as compared to the US firms. This is indicative of the efforts made by the Indian firms to "catch-up". power to the regression analysis.

23

Table 6 presents the results of the firm level analysis of productivity spillovers of FDI on the TFPG of domestic firms. Though a number of variables have been used, the equations with the best fit are presented.

The results show that FOREQ as an aggregate has no significant impact on the TFPG of domestic firms. However, Japanese firms have a positive spillover effect on the TFPG of domestic firms. Spillovers from US firms are not significant. This supports our hypothesis with respect to positive spillovers from Japanese firms. To capture the impact of the technology brought by Japanese and US firms interaction terms are used for the imports of technology by the foreign firms with the proportion of equity invested by them. The results suggest that technology imports of Japanese firms favorably affect the TFPG of the domestic firms. One of the reasons for this could be the difference in the level of technology brought in by the Japanese and US firms. Japanese FDI are largely undertaken by small and medium sized firms and their mode of transfer of technology is such that it facilitates transfer of technology more easily.

In specific case of Indian manufacturing more it is found that US FDI concentrates in Pharmaceuticals, Electronics, Chemicals -Organic, Personal Care, Dry Cells and Electric Equipment; which are technologically advanced sectors as compared to Japanese FDI which are prominently present in industries such as AutoAncillary, Automobiles-LCVs/ HCVs, Glass-Sheet/ Float and Rubber-Retreading. Given the type of industries where US FDI have entered this seems to be true since the chemicals and pharmaceutical industries involve process technology which makes their technology less susceptible to reverse engineering or imitation unlike autoancillary and electronic technology which is the stronghold of Japanese FDI in India.

Taking the initial gap in the productivity levels of the domestic firms and the average productivity levels of foreign firms in the industry, we find that gap has no significant effect on the TFPG of the domestic firms (Table 7). However, if the initial gap between the productivity levels of domestic firms is compared to the average productivity levels of Japanese and US firms separately, we find that the gap with the Japanese firms has a significant positive impact on the TFPG of the domestic firms. 24

This supports the “catch-up” theory. The gap with the US firms does not have any significant effect on the TFPG of the domestic firms.

The result with respect to the initial gap is in line with the results of Blomstrom & Wolff (1994) and Kokko et al (1996). While Blomstrom found that domestic producers’ productivity increases more rapidly and the gap from competing foreign producers’ productivity grows narrower, the larger the initial productivity gap, Kokko found positive and statistically significant spillover effects only in a sub sample of locally owned plants with moderate technology gaps vis-à-vis foreign firms. It also supports our hypothesis with respect to the effect of gap on the TFPG of domestic firms. This gap differs with respect to the US FDI and Japanese FDI and Indian firms are more likely to “catch-up” with the Japanese firms as compared to the US firms.

The overall results at the firm-level and at the industry level support our hypotheses and it can be said that the source of FDI and the prevailing state of market demand also have an impact on the TFPG of domestic firms, independent of firmspecific and industry-specific effects. Japanese FDI in the Indian manufacturing has positively contributed to the TFPG of the domestic firms though the same cannot be said for the US FDI.

7. CONCLUSION

25

India was one of the lowest recipients of FDI among developing countries till the end of 1970s. During 1970s cumulative inflows of FDI was about US$454 million or 0.20% of gross domestic investment (GDI). Many factors contributed to a lower level of FDI. One obvious factor was the restriction in foreign shareholdings of equity, which was limited to the maximum of 40% under the FERA. Although the absolute value of FDI rose sharply in 1980s in comparison with the earlier decade its share in GDI remained constant. While India is not yet anywhere near ASEAN countries and far too behind China in attracting FDI, it has done remarkably well in recent years as compared with its own past performance. It now becomes important to assess the impact of FDI on the growth of the industry. An important source of growth in industry is growth in productivity. FDI is expected to have both direct as well as indirect effects on the growth of productivity.

The paper contributes to the literature on productivity spillovers from FDI in two respects. It attempts to analyse the effect of the prevailing market demand conditions in the economy on the direction of impact of FDI on total factor productivity growth. Secondly, it examines the importance of the source-country of FDI in determining the impact of FDI on productivity growth given the firm and industry specific effects. It also tries to differentiate the productivity spillovers effects according to the source country.

Both aggregate level and firm level analysis have been undertaken. A panel data analysis for the period 1995-6 to 1999-2000 is carried out at the aggregate level analysis while the period of analysis at the firm level is 1993-94 to 1999-2000. The results show that if the markets are expanding at a slow pace, the foreign firms may capture substantial market shares and force the local firms to operate at sub optimal scales leading to an overall decline in the productivity growth. The spillovers from the FDI may also depend on the source of FDI. Japanese FDI are found to have larger spillovers as compared to US FDI. This is so because of the type of technology brought in by them and also because a low productivity gap exist between Japanese and Indian firms as compared to US and Indian firms

26

REFERENCES

Aitken, B., and Harrison A., 1999. Do domestic firms benefit from FDI? Evidence from Venezuela.. American Economic Review 89 (3), 605--618. Aitken, B., G.D. Hanson and A. Harrison., 1994. Spillover, Foreign Investment and Export Behaviour, NBER Working Paper No. 4967, December. Aoki, M.,1988. Information, Incentives and Bargaining in the Japanese Economy. Cambridge: Cambridge University Press. Banga. R. 2002, Corporate Financial Structures and Financial Performance of Japanese, US and Indian Firms in the Indian Manufacturing, Journal of International Economic Studies, No. 16, forthcoming. Basant, R., and Fikkert, B., 1996. The Effects of R&D, Foreign Technology Purchase and Technology Spillovers on Productivity in Indian Firms. The Review of Economics and Statistics 78 (2), 187--99. Blomstrom,M., and Wolf, E., 1994. Multinational Corporations and Productivity Convergence in Mexico, in: W. Bahmol, R. Nelson,and E. Wolff (Eds.), Convergence of Productivity: Cross National Studies and Historical Evidence, New York, Oxford University Press, p. 243--259. Branstetter, L.G., 2000, Is Foreign Direct Investment a Channel of Knowledge Spillovers? Evidence from Japan’s FDI in the United States’ Dept of Economics, University of California, Davis. Buckley, P.J. and M. Casson (1976), The Future of the Multinational Enterprise, London; Macmillan. Caves, Richard E., 1974a, The distinctive nature of the multinational enterprise, in: J.H. Dunning, (Eds.), Economic Analysis and the Multinational Enterprise, London, Allen and Unwin, pp. 13--30. Caves, Richard E.1974b, Causes of Direct Investment : Foreign Firms Shares in Canadian and United Kingdom Manufacturing Industries . Review of Economics and Statistics, 56 August: 207-293. Caves, R.E., 1996, Multinational enterprise and economic analysis, Cambridge University Press, Cambridge. Catherine Y Co, 2001, Trade, Foreign Direct Investment and Industry Performance, International Journal of Industrial Organisation, 19, 163-183

27

Cornwell, C., Schmidt , P., and Sickles, R.C., 1990. Production Frontier with CrossSection and Time-Series Variation in Efficiency Level. Journal of Econometrics 46 (1-2), 317--32 Djankov, S., and Hoekman, B., 2000. Foreign Investment and Productivity Growth in Czech Enterprises. The World Bank Economic Review Vol.14, No. 1, 49--64. Findlay, R., 1978. Related Backwardness, Direst Foreign Investment and the Transfer of Technology: A Simple Dynamic Mode., Quarterly Journal of Economics 92 (February), 1--6. Dunning John H.1973. The Determinants of International Production. Oxford Economic Papers, 25, November: 289-336. Gagelmann, K.Z., 2000. Productivity Spillovers from Foreign Direct Investment in Poland. Economic Systems Vol.24, (3), 223--256. Globerman,S., 1979. Foreign Direct Investment and Spillover Efficiency Benefits in Canadian Manufacturing Industries. Canadian Journal of Economics 12, 42--56. Goldar, B.N., 1995. Technology Acquisition and Productivity Growth: A study of Industrial Firms in India, Working Paper Series, Institute of Economic Growth, Delhi University Enclave, September. Graham, Edward M., and. Krugman P.R., 1995. Foreign Direct Investment in the United States. Washington, D.C. Grether, J.M., 1999. Determinants of Technological Diffusion in Mexican Manufacturing: A Plant-Level Analysis. World Development Vol. 27, 1287--1298. Grossman, G and E. Helpman, 1991, Innovation and Growth in the Global Economy, The MIT Press, Cambridge. Haddad, M., and Harrison, A., 1993, Are there Positive Spillovers From Foreign Direct Investment? Evidence from Panel Data for Morocco. Journal of Development Economics, October. Harrison, A., 1994. Productivity, Imperfect Competition and Trade Reform: Theory and Evidence. Journal of International Economics 36, 53--73. Hill, H., 1988. Foreign Investment and Industrialization in Indonesia. Singapore. Oxford University Press. Hill, H., 1990. Foreign Investment and East Asian Economic Development. AsianPacific Economic Literature 4,2 (September): 21--58. Kathuria, V., 2000. Productivity Spillovers from Technology Transfer to Indian Manufacturing Firms. Journal of International Development 12(3), 343--69. 28

Kathuria, V., 2001. Foreign Firms, Technology Transfer and Knowledge Spillovers to Indian Manufacturing firms: a Stochastic Frontier Analysis., Applied Economics, 33, 625-642. Kojima, K., 1978. Direct Foreign Investment: A Japanese Model of Multinational Business Operations. London: Croom Helm. Kojima, K., 1986. Japanese-Style Direct Foreign Investment. Japanese Economic Studies 14,3 (Spring): 52--82. Kokko, A., 1994. Technology, Market Characteristics and Spillovers. Journal of Development Economics 43, 279--293. Kokko, A., Tansini, R., and Zejan, M.C., 1996. Local Technological Capability and Productivity Spillovers from FDI in the Uruguayan Manufacturing Sector. Journal of Development Studies 32 (4), April, 602--11. Okamoto, Y., 1999. Multinationals, Production Efficiency, and Spillover Effects: The Case of the US Auto Parts Industry, Weltwirtschaftliches Archiv, Vol. 135, No. 2. Ozawa, T., 1979. Multinationalism: Japanese Style. Princeton, N.J.: Princeton University Press. Perez, C., 1983. Structural Change and the Assimilation of new technologies in the Economic and Social Systems. Futures 15 (4). Ramstetter, Eric D. 1987. The Impact of Direct Foreign Investment on Host Country Trade and Output: A Study of Japanese and US Direct Foreign Investment in Korea, Taiwan and Thailand. Direct Foreign Investment and Export Promotion: Policies and Experience in Asia. Ed. Seiji Naya et al. Honolulu: East-West Resource Systems Institute, 223-57. Rotenberg, J.J and G. Saloner, 2000, `Competition and Human Capital Accumulation: A Theory of Interregional Specialization and Trade’ Regional Science and Urban Economics, 30, 373-404. Srivastava, V. 1996. Liberalisation, Productivity and Competition: A Panel Study of Indian Manufacturing, Delhi: Oxford University Press. Teece, D.J., 1981. The Market for Know-How and the Efficient International Transfer of Technology, the Annals of the American Academy of Political and Social Science, Vol. 458, 81-96. Tsou, M.W. and Liu, J.T., 1998, The Spillover Effect from FDI: Empirical Evidence from Taiwan Manufacturing Industries, Taiwan Economic Review, June, 155-81.

29

Wang and Blomstrom, M. , 1992. Foreign Investment and Technology Transfer: A Simple Model, European Economic Review, 36, 137-155. Womack, James P., Daniel T. Jones, and Daniel Roos., 1991. The Machine That Changed the World: The Story of Lean Production--How Japan's Secret Weapon in the Global Auto Wars Will Revolutionize Western Industry. New York: Harper Collins. Yamamura, Kozo., 1990. Will Japan's Economic Structure Change? Confessions of a Former Optimist. Japan's Economic Structure: Should It Change? Ed. Kozo Yamamura. Seattle: Society for Japanese Studies, 13-64. Yamamura, Kozo., 1994. Comment. Does Ownership Matter? Japanese Multinationals in Europe. Eds. Mark Mason and Dennis Encarnation. Oxford: Clarendon Press, 87-90

TABLE 1:FDI INFLOWS AND GROWTH RATES OF MANUFACTURING SECTOR YEAR

INDEX OF INDUSTRIAL PRODUCTION

DIRECT

GROWTH RATES

INVESTMENT

MANUFACTURING SECTOR (BASE: 1993-94)weight: 79.36

1994-95

9.1

INFLOW (Rs. Crores) 4126

30

1995-96

13.8

7172

1996-97

6.7

10015

1997-98

6.6

13,220

1998-99

4.4

10,358

1999-2000

9.2

9338

SOURCE: CSO, GOI: HANDBOOK OF INDUSTRIAL POLICY AND STATITICS 2000& CIER’S INDUSTRIAL DATA BOOK 1998

TABLE 2:COMPARISON OF INDUSTRIES WITH AND WITHOUT FDI TOTAL NUMBER OF INDUSTRIES NUMBER OF INDUSTRIES WITH FDI

74 58

NUMBER OF INDUSTRIES WITHOUT FDI 16 ALL INDUSTRIES WITH INDUSTRIES FDI MEA MEAN σ σ N 1. TFPG 0.015 0.15 0.018 0.16 2. FOREIGN EQUITY 0.07 0.11 0.10 0.12 3. SKILL INTENSITY 0.28 0.08 0.28 0.07 4. EXPORT INTENSITY 0.11 0.14 0.12 0.15 5. R&D INTENSITY 0.001 0.002 0.009 0.002 6. IMPORT OF 0.001 0.002 0.002 0.002 DISEMBODIED TECHNOLOGY 7. IMPORT OF 0.14 0.13 0.15 0.13 EMBODIED TECHNOLOGY TFPG 1995-96 0.024 0.15 0.041 0.11 1996-97 0.016 0.07 0.027 0.06 1997-98 0.007 0.09 0.007 0.07 1998-99 0.012 0.15 0.005 0.13 1999-2000 0.013 0.24 0.002 0.19

INDUSTRIES WITHOUT FDI MEAN σ 0.004 0.0007 0.28 0.10 0.001 0.0007

0.19 0.001 0.08 0.13 0.002 0.002

0.10

0.09

-0.02 -0.01 0.002 0.016 0.027

0.20 0.07 0.11 0.15 0.19

TABLE 3:COMPARISON OF TFPG OF INDUSTRIES WITH HIGH FDI AND LOW FDI TOTAL NUMBER OF INDUSTRIES 58 NUMBER OF INDUSTRIES WITH HIGH FDI (>25%) NUMBER OF INDUSTRIES FDI BETWEEN (5-25%) INDUSTRIES WITH INDUSTRIES WITH >25%FEP FDI

31

10 48 INDUSTRIES FDI