Week-6

Oxford Development Studies, Vol. 30, No. 3, 2002

Trade Policy, Equipment Investment and Growth in India

KUNAL SEN

ABSTRACT The relationship between trade policy and economic performance is one of the oldest controversies in economic development. In this paper, we examine an alternative mechanism through which trade reforms may impact on economic growth to those commonly discussed in the literature. This mechanism builds on the link between equipment investment and growth that has been observed in cross-country data. We argue that that in countries which have had highly restrictive trade policies with respect to capital goods, liberalization measures that speciŽ cally target capital goods imports may bring about a fall in the relative price of capital goods, leading to an increase in the rate of investment in equipment. Quantifying the link between trade policy, equipment investment and economic growth in the Indian case, we Ž nd strong support for this mechanism.

1. Introduction

The relationship between trade policy and economic performance is one of the oldest controversies in economic development. In recent years, there has been a revival of interest in the debate on trade and growth due in part to current widespread trade liberalization in developing countries and in part to developments in economic theory– most notably, the endogenous growth theories pioneered by Romer (1986) and Lucas (1988). The endogenous growth theories identify several mechanisms by which trade reforms may have a sustained impact on economic growth. First, trade liberalization increases the variety of goods, and raises productivity by providing higher quality intermediate and capital goods. Second, trade liberalization leads to the exploitation of scale economies as Ž rms in the reforming economy expand into world markets. Finally, trade reforms may lead to greater technological progress in the reforming economy as Ž rms are able increasingly to capture new ideas being generated in the rest of the world.

While the endogenous growth theories provide new and important insights on the dynamic effects of trade on growth, the theoretical literature does not yield an unam- biguous conclusion on whether trade reforms have a positive impact on economic growth (Rodrik, 1988; Tybout, 1992). Furthermore, systematic attempts at quantiŽ cation have failed to single out trade policy as a major factor in economic growth (Rodrik, 1992). Both the ambiguity in the theoretical literature and the weak empirical evidence have led trade liberalization sceptics to argue that “the effect of trade

Kunal Sen, School of Development Studies, University of East Anglia, Norwich NR4 7TJ, UK. This paper has beneŽ ted considerably from the helpful and detailed comments received from the Editors of the journal, the late George Peters and Sanjaya Lall. The usual disclaimer applies.

ISSN 1360-0818 print/ISSN 1469-9966 online/02/030317-15 Ó 2002 International Development Centre, Oxford DOI: 10.1080/1360081022000012725

318 K. Sen

liberalisation on growth is, at best, very tenuous, and at worst, doubtful” (Edwards, 1998, p. 383).

A separate line of enquiry in the growth literature has explored the link between equipment (or machinery) investment and economic growth, arguing that equipment investment is strongly and robustly correlated with long-run growth rates in cross-coun- try data (De Long & Summers, 1991, 1992, 1993). In particular, for developing countries, the magnitude of the estimated returns to equipment investment is extremely high–well over 50%–and much higher than the estimated returns to investment in structures (Temple, 1998). Furthermore, there is a strong negative relationship be- tween the relative price of equipment and economic growth (Jones, 1994). Thus, policies that decrease the relative price of equipment will encourage the accumulation of equipment capital, leading to favourable effects on economic growth in developing countries.

In this paper, we propose an alternative mechanism by which trade policy may impact on economic growth to those that have been discussed in the endogenous growth literature. This mechanism builds on the link between equipment investment and growth observed in cross-country data. We argue that in countries which have had highly restrictive trade policies with respect to capital goods, liberalization measures that speciŽ cally target capital goods imports may bring about a fall in the relative price of capital goods, leading to an increase in the rate of investment in equipment. This provides a simple mechanism by which trade policy can have an unambiguous positive and sustained impact on economic growth. We attempt to quantify the link between trade policy, equipment investment and economic growth by examining the Indian case.

As is well known, since independence India has had one of the most highly restrictive trade regimes with respect to capital goods in the world. Since 1977–78, there has been a slow but steady liberalization of capital and intermediate goods imports, culminating in the comprehensive reforms of 1991. In this paper, we examine whether trade reforms enacted since the late 1970s have had an appreciable positive impact on the rate of equipment investment and whether this has contributed to higher economic growth in India.

The rest of the paper is divided into six sections. In the next section, we provide a brief overview of trade policies in India since independence. In Section 3, we set the stage for the empirical analysis by attempting to identify patterns in investment behaviour in India during the period of our study. Section 4 presents the conceptual framework and Section 5 the empirical results. In Section 6, we examine an alternative explanation of the observed rise in the equipment investment rate in the post-1991 period: the role of domestic deregulation of industrial policy. Section 7 concludes.

2. Trade Policy in India

The import and exchange rate regime that Indian policy-makers followed since inde- pendence was aimed at the comprehensive, direct control over foreign exchange utilization, with an overwhelming reliance on quotas rather than tariffs (Bhagwati & Srinivasan, 1975). The allocation of import licences re ected two major criteria: (1) “essentiality”; and (2) “indigenous non-availability”. Thus, imports, in terms of both magnitude and composition, were permitted only if the Ž rm in question certiŽ ed to the government that they were “essential” (as inputs or equipment for production). At the same time, the government had to clear the imports from the viewpoint of indigenous availability: if it could be shown that there was domestic production of the goods

Trade Policy in India 319

demanded, then imports were not permitted (regardless of cost and quality consider- ations).

Nearly all imports were subject to discretionary import licensing or were “canalized” by government monopoly trading organizations. The only exceptions were commodi- ties listed in the Open General Licence (OGL) category. Capital goods were divided into a restricted category and the OGL category. While import licences were required for restricted capital goods, those in the OGL category could be imported without a licence subject to several conditions. Intermediate goods were also classiŽ ed into the banned, restricted and limited permissible categories plus an OGL category. As the names suggest, the Ž rst three lists were in order of import licensing stringency. OGL imports of intermediate goods were also governed by the “actual user” condition. The import of consumer goods was, however, banned (except those that were considered “essential” and could only be imported by the designated government canalizing agencies).

Beginning with the export—import policy of 1977–78, there was a slow but sustained relaxation of import controls. Several capital goods that were not allowed to be imported without an import licence were shifted to the OGL category. The number of capital goods on the OGL list increased from 79 in 1976 to 1170 in April 1988. These changes were made with the intention of allowing domestic industries to modernize. Moreover, during the 1980s the import licensing of capital goods in the restricted list was administered with less stringency (Pursell, 1992). As a consequence, the import penetration ratio in the capital goods sector increased from 11% in 1976–77 to 18% in 1985–86 (Goldar & Renganathan, 1990). In the case of intermediate goods too, there was a steady shift of items from the restricted and limited permissible categories to the OGL category. However, in practice a capital or an intermediate good was placed in the OGL list only if it was not domestically produced. Thus, import liberalization during this period may have led to some degree of competition to established producers of intermediate and capital goods in India (though in several instances, the goods that were allowed to be imported were imperfect substitutes of domestically produced goods). On the other hand, there was an increase in tariff rates across all commodities and, in particular, on capital goods. By 1987/88, the unweighted average of tariffs on manufactured goods was 147%, with most tariff lines for manufac- turing clustered around a range of 140–160%.

The pace of the trade reforms–in particular, the shift from quantitative import controls to a protective system based on tariffs–initiated in the mid-1970s was consid- erably quickened by the government (led by Rajiv Gandhi) that came into power in November 1985. Restrictions on the import of capital goods were further eased to encourage technological modernization. Also, beginning in the mid-1980s, there was a renewed emphasis on export promotion. The number and value of incentives offered to exporters were increased and their administration streamlined. The allotment of REP licences–tradable import entitlements awarded to exporters on a product-speciŽ c basis– became increasingly generous (Agrawal et al., 1995). Finally, the duty exemption scheme for imported inputs was extended to cover all imported inputs for both direct and indirect exporters.

In 1991, as a part of the comprehensive economic reform programme, there was a signiŽ cant liberalization of the trade regime with respect to capital goods. Import licensing was virtually abolished with respect to most machinery and equipment and manufactured intermediate goods (Ahluwalia, 1999). There was also a signiŽ cant cut in tariff rates, with the peak rate reduced from 300 to 150% and the peak duty on capital goods cut to 80%.1 Import-weighted custom duty rates fell from an average of

320 K. Sen

97% in 1990–91 to 29% in 1995–96. There was, however, little change in trade policy with respect to consumer goods which remained in the “negative” (banned) list.

The reforms of the trade regime in 1991 coincided with an equally signiŽ cant set of reforms in industrial policy. Prior to 1991, there was a system of industrial licensing of private industry in place which governed almost all aspects of Ž rm behaviour in the industrial sector, controlling not only entry into an industry and expansion of capacity, but also technology, output mix, capacity location and import content. In 1991, previous piecemeal efforts towards liberalization of controls were consolidated in a comprehensive wave of domestic deregulation. Industrial licensing was abolished alto- gether, except for a list of environmentally sensitive industries. Along with this came the removal of restrictions on large business groups to merge or expand, and the opening up of several industries, previously reserved for the public sector, to the private sector. Since the 1991 reforms encompassed both the domestic deregulation of private industry and trade reforms that facilitated the purchase of low-cost imported capital goods, this implies that any changes in India’s growth performance in the post-1991 period cannot be directly attributed to one set of reforms. This important issue will need to be dealt with in the empirical sections that follow.

3. Investment Behaviour in India: Trends and Patterns

In this section, we present a brief overview of investment behaviour in India over the period 1955/56–1998/99, highlighting the key features of the data. We begin with a graph of GDP growth (Figure 1). As expected of an economy where climactic factors play an important role in determining total output, growth shows a great deal of variation from year to year. There is, however, a clear rise in the trend rate of growth of output in the 1980s and 1990s as compared with the earlier period–the average annual growth rate in 1981/82–1998/99 was 5.6%, while that for the period 1955/56– 1980/81 was 3.6%. There has also been an increase in gross Ž xed capital formation (GFCF) as a ratio of GDP since the 1980s, with a signiŽ cant acceleration in the 1990s (Figure 2). The increase in the GFCF-to-GDP ratio in this period can clearly be attributed to a rapid increase in the ratio of equipment investment to GDP, particularly since 1991. In contrast, the ratio of structures investment to GDP has been stagnant since the mid-1970s.

The increase in Ž xed investment as a ratio of GDP since the 1980s cannot have resulted from an increase in public Ž xed investment, since this had been falling as a ratio of GDP (both in the aggregate and in each of the components) since the mid-1980s (Figure 3). The rise must be attributed to the sharp increase in private Ž xed investment since the mid-1980s (Figure 4). The ratio of private Ž xed investment to GDP increased from 9.1% in 1981/82–85/86 to 14.9% in 1991/92–98/99. This was primarily due to a sharp rise in the rate of private equipment investment, with the private investment in buildings and structures showing no signs of buoyancy.

One possible explanation of the sharp rise in private equipment investment could be the fall in the relative price of equipment. The latter has shown a negative trend since the late 1970s, in contrast to the relative price of construction investment, which showed a signiŽ cant increase in the 1980s.

This suggests a clear upward trend in the series as a ratio of GDP since the 1980s, driven by a spectacular increase in private equipment investment. At the same time, there was an increase in the average annual growth rate of output and a fall in the relative price of equipment. In the econometric analysis, we shall explore more rigor-

Trade Policy in India 321

Figure 1. Growth of GDP.

ously whether the behaviour of these three series–the relative price of equipment, the rate of equipment investment and the growth rate of output–can be causally linked.

4. The Analytical Framework

The argument that trade policy can positively affect economic growth via an increase in equipment investment is based on three behavioural relationships. These are: (i) the link between equipment investment and economic growth; (ii) the link between the relative price of equipment and the rate of equipment investment; and (iii) the link between trade policy and the relative price of equipment. We discuss each in turn:

4.1 The Link Between Equipment Investment and Economic Growth

There is now a vast number of cross-country studies of the determinants of economic growth using a wide range of explanatory variables. While the factors found to be important in explaining economic growth have differed from study to study, Levine & Renelt (1992) found that the signiŽ cance of the investment rate in explaining economic growth remains robust to different speciŽ cations. There seems to be little doubt that the investment rate is a crucial determinant of economic growth, if not the key determinant. Furthermore, the new growth literature has argued that among the three components of total investment–investment in equipment, investment in construction and investment in inventories–the most important for growth is equipment investment (De Long & Summers, 1991, 1992, 1993; Temple, 1998). This view argues that the

322 K. Sen

Figure 2. Investment in Ž xed capital and its components, as a percentage of GDP.

social return to equipment investment is higher than that for other types of investment. The reasons are not very clear, though as De Long & Summers argue (1991, p. 447), “historical accounts of economic growth invariably assign a central role to mechaniza- tion”. It could be that the role of external economies may be greater for equipment investment than for buildings and construction investment, possibly due to the greater amount of research and development expenditure in the machinery sector.

To assess the impact of equipment investment on the growth rate of output, we use a simple empirical formulation similar to that used by De Long and Summers. We take the growth rate of output to be a linear function of the rates of equipment and buildings construction investment. However, in our context, it would be useful to keep the rate of private equipment investment separate from the rate of public equipment investment in the output growth equation. This is for two reasons. First, the rate of return on private equipment investment may be different from that on public equipment invest- ment. Second, our purpose in the next stage of the analysis is to explain the behaviour of equipment investment, and it can be argued that public investment is, in great part, exogenously determined and in uenced more by political and institutional variables than by economic variables. Thus, the determinants of public investment may be quite different from those of private investment.

The Ž nal speciŽ cation is as follows:2

GY 5 a1PVRE 1 a2PBRE 1 a3RS, (1)

Trade Policy in India 323

Figure 3. Public Ž xed investment and its components, as a percentage of GDP.

where GY is the growth rate of GDP, PVRE is the ratio of private sector investment in equipment to GDP, PBRE is the ratio of public sector investment in equipment to GDP and RS is the ratio of investment (both private and public) in structures to GDP.

4.2 The Link Between the Relative Price of Equipment and the Rate of Equipment Investment

Since we are interested in examining the relationship between the relative price of equipment and the rate of equipment investment, we conŽ ne our empirical analysis for this section to the determinants of private equipment investment (as we have argued earlier, public investment may be taken to be exogenously determined). We model the rate of equipment investment by the private sector as follows:

PVRE 5 b0 1 b1RPE 1 b2RPS 1 b3FINT 1 b4RINT 1 b5PBRI 1 b6DUM91 1 b7PVRE( 2 1), (2)

where RPE is the relative price of equipment (price de ator for equipment investment as a ratio of the GDP de ator), RPS is the relative price of structures investment (price de ator for structures investment as a ratio of the GDP de ator), FINT is Ž nancial deepening, RINT is the real interest rate (the bank lending rate minus the in ation

324 K. Sen

Figure 4. Private Ž xed investment and its components, as a percentage of GDP.

rate), PBRI is total public investment as a ratio of GDP and DUM91 is a dummy variable for the 1991 reforms (value of one from 1991 onwards, zero otherwise).

We would expect from theory that the sign of the coefŽ cient for the relative price of equipment, b1, will be negative–an increase in the latter will decrease the rate of equipment investment in the economy. The sign for the coefŽ cient for the relative price of structures, b2, cannot be determined a priori and will depend on whether construc- tion capital is a complement or substitute for equipment capital. Financial deepening is expected to have a positive impact on equipment investment. Financial deepening can increase both the volume and efŽ ciency of investment. The “debt accumulation” hypothesis of Gurley & Shaw (1955), formalized more recently by Bencivenga & Smith (1991), argues that the spread of organized Ž nance can help overcome indivisibilities in investment through the mobilization of otherwise unproductive resources. Moreover, Ž nancial intermediaries and markets play an important role in selecting the most promising Ž rms and households for lending purposes and thus contributing to the more efŽ cient use of capital (Levine, 1997). Financial intermediaries may also enhance the quality of investment by identifying entrepreneurs with the best chances of successfully initiating new activities (King & Levine, 1993). In the Indian case, Bell & Rousseau (2001) found conclusive evidence of the positive impact of Ž nancial deepening on gross

Trade Policy in India 325

Figure 5. Relative prices of structures and equipment.

domestic Ž xed investment in the post-1950 period. Following Bell and Rousseau, we use the ratio of real bank credit to the private sector as a ratio of GDP as our preferred measure of Ž nancial deepening (FINT).3

An increase in the real interest rate is expected to have a negative effect on the rate of equipment investment via an increase in the cost of capital. Public investment may affect private investment via both supply and demand. On the supply side, the private sector relies on public investment for most of the infrastructure, because this is either a natural or a legal monopoly of the government. Public infrastructural investment can affect private equipment investment by in uencing its rate of return–poor roads, an erratic supply of electricity or inadequate communication facilities can negatively affect the amount of output that it is possible to obtain from a given amount of investment. Thus, public investment in infrastructure and private investment should be comple- mentary (Blejer & Khan, 1984). On the demand side, the relationship is ambiguous. If there is some slack in the economy one would expect a change in public investment to push private investment in the same direction. Otherwise, some private investment may be “crowded out” (Athukorala, 1998).

We also add a post-1991 period dummy (DUM91) to capture the effect of the 1991 deregulation of industrial policy which may have had a positive effect on the private investment rate over and above the effect of trade reforms on the latter working through the relative price of capital goods. Finally, we include the one-period lagged private equipment investment rate to capture the high degree of persistence in the latter.

326 K. Sen

Table 1. Summary data of variables used in econometric analysisa,b

Years GY PVRE PBRE RS PBRI RPE RPS FINT RINT RER

1955/56–60/61 3.98 4.61 2.12 8.86 6.42 0.74 0.90 0.09 0.63 8.38

1961/62–65/66 2.84 5.05 2.78 9.64 8.47 0.81 0.88 0.11 2 0.13 6.64 1966/67–70/71 2.28 4.67 2.46 10.06 7.07 0.89 0.87 0.15 2 1.12 7.51 1971/72–75/76 4.66 4.06 2.38 11.20 7.03 0.89 0.81 0.13 1.73 7.38

1976/77–80/81 3.08 4.68 2.68 9.89 7.23 0.92 0.89 0.16 0.60 8.44

1981/82–85/86 4.69 4.50 3.22 9.99 7.66 0.96 0.91 0.18 3.75 9.39

1986/87–90/91 6.30 7.60 5.03 7.73 9.27 0.91 1.36 0.25 8.02 10.28

1991/92–98/99 5.77 10.53 4.18 7.57 7.47 0.86 1.40 0.24 7.58 14.09

SD: 0.033 0.026 0.011 0.013 0.012 0.089 0.236 0.061 0.055 2.56

a All variables in percentages except RER, RPE and RPS. b Annual averages, except Ž nal row which contains standard deviation of variables over period 1955/56–1998/99.

4.3 The Link Between Trade Policy and the Relative Price of Equipment

Machinery is a key tradable commodity and its relative price will be greatly in uenced by trade policy. In addition, the relative price of machinery would be affected by the real exchange rate as movements in the latter would lead to change in the price of tradables relative to that of non-tradables. We choose a very simple speciŽ cation for the relative price of equipment, modelling it as a function of trade policy and the real exchange rate. As is evident from the discussion in Section 2, the most important shifts in Indian trade policy occurred in the years 1977, 1985 and 1991. We use dummies to capture the changes in the trade regime in these 3 years. We also add a time-trend to the speciŽ cation to capture the upward drift in the relative price series, is evident from Figure 5.

RPE 5 c0 1 c1TIME 1 c2DUM77 1 c3DUM85 1 c4DUM91 1 c5RER, (3)

where TIME is the time-trend, DUM77 is a dummy variable for the 1977 reforms (value of one from 1977 onwards, zero otherwise), DUM85 is a dummy variable for the 1985 reforms (value of one from 1985 onwards, zero otherwise), DUM91 is a dummy variable for the 1991 reforms (value of one from 1991 onwards, zero otherwise) and RER is the real exchange rate (equal to eP*/P, where e is the nominal exchange rate, P* is the foreign price level and P is the domestic price level).4

5. Results

We estimate equations (1), (2) and (3) using ordinary least squares (OLS) regression. Our period of analysis is 1955/56–1998/99. A summary of the data used in the regressions is provided in Table 1.

There is a possibility that equation (1) would be subject to simultaneity bias if we used current equipment and structures investment rates as explanatory variables in the regression. This is because positive and signiŽ cant coefŽ cients on the rates of private equipment investment and structures investment could imply that higher investment rates are the result of economic growth, not the other way around. Therefore, we use one-period lagged investment rates in the Ž nal estimation. We also include dummy variables for the years 1965 and 1979 (denoted DUM65 and DUM79, respectively) to

Trade Policy in India 327

Table 2. Regression results

1. GY 5 0.544*PVRE( 2 1) 1 0.189*RS( 2 1) 2 0.084*DUM65 2 0.10DUM79 (3.29)*** (1.73)* (3.04)*** (3.56)***

R2 5 0.41 SE 5 0.027 SERCOR 2 c2(1) 5 3.10 NORM 2 c2 (2) 5 0.68

2. PVRE 5 0.072 2 0.084*RPE 1 0.174*FINT 1 0.017*DUM91 1 0.490*PVRE( 2 1) (5.34)*** (5.14)*** (5.73)*** (3.60)*** (5.13)***

R2 5 0.92 SE 5 0.007 SERCOR 2 c2(1) 5 3.10 NORM 2 c2(2) 5 1.44

3. RPE 5 0.707 1 0.011*TIME 2 0.049*DUM77 2 0.13*DUM85 2 0.13*DUM91 (26.07)*** (5.77)*** (1.30) (3.97)*** (3.47)***

(3.47)***R2 5 0.62 SE 5 0.058 SERCOR 2 c2 (1) 5 2.65 NORM 2 c2 (2) 5 0.78

Notes: t-Ratios of regression coefŽ cients are given in parentheses. Approximate critical values for the t-ratios are as follows: 10% 5 1.66 (*), 5% 5 2.04 (**) and 1% 5 2.75 (***); R2 5 R-squared; SE 5 standard error of regression; SERCOR 5 Lagrange multiplier test of residual serial correlation; and NORM 5 Jarque-Bera test for the normality of residuals.

capture the large dips in the growth rate of output in these 2 years (due to a negative weather shock in 1965 and an oil price shock in 1979). The rates of public equipment investment (PBRE) for equation (1) and total public investment (PBRI), the relative price of structures (RPS) and the real interest rate (RINT) for equation (2) consistently had statistically insigniŽ cant coefŽ cients in the experimental runs and were dropped in the Ž nal speciŽ cation.5 Similarly, in our initial estimate of equation (3), the coefŽ cient for the real exchange rate failed to attain statistical signiŽ cance at the 10% level and was omitted.6 The Ž nal results are presented in Table 2.

As the estimates indicate, the explanatory power of the regressions is high and they perform well by all diagnostic statistics. While both the coefŽ cients of PVRE and RPS are statistically signiŽ cant (at the 1% and 10% level, respectively), we found that an increase in the rate of private equipment investment has a far stronger positive impact on the growth of output than an increase in the rate of construction investment–the coefŽ cient on the former is more than double that of the latter. This is consistent with what has been observed in the cross-country studies.7In the case of private equipment investment, the relative price of equipment plays a decisive role in its determination–its coefŽ cient is negative and highly signiŽ cant. Financial deepening has also been a key determinant of the private equipment investment rate in India. In addition, the dummy for the post-1991 period indicates a clear increase in equipment investment associated with that period. This provides some support for the argument that the deregulation of industrial policy in 1991 played an important role in the observed increase in equip- ment investment in the post-1991 period (see Section 3) and may provide an alternative explanation to that of trade policy of the post-1991 surge in private investment in India. We shall investigate this explanation in a separate section later.

From equation (3), we see that while the trade liberalization episodes of 1977, 1985 and 1991 led to a fall in the relative price of equipment, the most signiŽ cant were the episodes of 1985 and 1991. This is perhaps not surprising, given the muted nature of the 1977 liberalization episode. However, we found that the 1985 liberalization episode led to a similar fall in the relative price of equipment as the more comprehensive 1991 reforms. While somewhat counter-intuitive, this result is in accord with the Ž ndings from previous studies that the 1985 reforms have had signiŽ cant positive effects on industrial productivity and growth (Ahluwalia, 1991; Srivastava, 1996; Chand & Sen,

328 K. Sen

2002) by beginning the process of relaxing the quotas on capital and intermediate goods. By doing so, the reforms brought about competition in the domestic capital and intermediate goods sectors (albeit in a limited way).

Using the coefŽ cients estimated from equations (1)—(3), it is possible to obtain an estimate of the contribution of the trade reforms of 1985 and 1991 to the average annual growth of GDP following the reforms.8 Our estimates indicate that the com- bined effect of the trade reforms of 1985 and 1991 was to increase the average annual growth rate of GDP during this period by 1.2 percentage points.

6. An Alternative Explanation of the Post-1991 Increase in Investment

As noted in Section 3, the timing of the trade reforms in 1991 coincided with the removal of the industrial licensing system in the same year. This implies that the spectacular rise in private investment, and in particular in equipment investment, can in part be attributed to the domestic deregulation measures. The results in the previous section, and in particular our estimates of equation (2), support this hypothesis. One possible way of disentangling the effects of trade reforms from reforms in industrial policy on investment is to examine the behaviour of imported and domestically produced capital goods in the post-1991 period. Figure 6 provides time-plots of imported and domestic capital goods, and total capital goods as a percentage of GDP over the period 1955/56–1998/99. All three ratios show a signiŽ cant rise in the post-1991 period. This is consistent with arguments that stress the import- ance of trade policy and those that highlight domestic deregulation in explaining investment behaviour. Trade-induced changes in equipment prices could have both income and substitution effects and a fall in the latter may have led to an increase in both imported and domestically produced goods since 1991. On the other hand, the rise in the shares of imported and domestic capital goods in income could also be due to the domestic deregulation measures that allowed Ž rms in India to expand and enter new industries.

To establish the role of trade policy in India’s post-1991 investment performance, we assess whether changes in the relative price of capital have led to a substitution away from domestic capital goods and towards imported capital goods. To do this, we estimate a simple OLS regression, where the share of imported capital goods in total capital goods (MSI) is regressed against an intercept (INT), its past value and the lagged relative price of capital (RPE). We present the estimates (t-ratios in parenthe- ses):

MSI 5 0.160*INT 2 0.140*RPE( 2 1) 1 0.80*MSI( 2 1) (2.16) (1.93) (13.84)

R2 5 0.92.

The coefŽ cient on lagged RPE is negative and statistically signiŽ cant at the 10% level. This indicates that changes in the relative price of capital are induced by changes in the trade regime as such changes lead to a substitution away from domestic towards imported capital goods in the Indian context. Thus, there is indirect evidence that the radical reforms of the trade regime in 1991 pertaining to capital goods was a key contributing factor to the increase in the equipment investment rate.

Trade Policy in India 329

Figure 6. Imported, domestically produced and total capital goods as percentage of GDP.

7. Conclusions

In this paper, we examine an alternative mechanism though which trade reforms may impact on economic growth to those commonly discussed in the empirical growth literature. We argue that trade policy may positively affect output growth via an increase in the rate of equipment investment. There is strong support for this mechan- ism in the Indian case. We Ž nd that the rate of private equipment investment has a strong positive impact on economic growth, and that the former is signiŽ cantly and negatively in uenced by the relative price of equipment, which fell sharply during the trade policy reforms of 1985 and 1991.

The mechanism that we have focused on in the paper provides a strong link between trade liberalization of the capital goods sector and economic growth. Earlier studies on the link between trade policy and economic performance do not make any distinction between reforms in the capital goods and other sectors. Our paper makes the point that trade reforms in the capital goods sector can be a key contributing factor to economic growth.

Notes

1. As Joshi & Little (1997) argue, the concentration of emphasizing an early reduction in tariffs on capital goods in the reform process was probably intended to avoid discouraging investment because of the expectation of a later reduction in tariffs.

330 K. Sen

2. We omit the constant term, as in the standard growth models (both in the neo-classical and endogenous growth tradition), growth of output will be zero if the investment rate is zero.

3. Real bank credit to the private sector is obtained by de ating nominal bank credit to the private sector by the price de ator for Ž xed investment.

4. For e, we use the rupee—US dollar bilateral exchange rate; for P*, we use the wholesale price index in the USA; and for P, we use the consumer price index for India.

5. The possibility of simultaneity bias also exists in the case of OLS estimates of equation (2), as the investment rate is an important demand-side determinant of both the relative price of capital and bank credit to GDP ratio, and so the coefŽ cients on these two explanatory variables in the regression may be biased upwards. The Wu-Hausman test statistic on the endogeneity of the error term, however, does not indicate the presence of simultaneity bias in the regression estimates.

6. The F-statistic associated with the deletion of RER from the Ž nal equation was 2.28 and not signiŽ cant at the 10% level.

7. Given that the growth rate of total output seems to be heavily in uenced by weather-related factors in the Indian context (see the discussion in Section 3), we also experimented with replacing this variable with the growth rate of non-agricultural output. We found, however, that such a model has less explanatory power (as evidenced by a lower R-squared) as compared with the model reported in Table 2. As expected though, the coefŽ cient on PVRE was higher when growth of non-agricultural output was used as the dependent variable instead of total output.

8. It is possible to obtain such an estimate by substituting equation (3) in equation (2), and equation (2) so obtained in equation (1). Using this method, we found that the 1985 and 1991 reforms increased GDP growth by 0.59 percentage points each.

References:

Agrawal, P., Gokarn, S., Mishra, V., Parikh, K.S. & Sen, K. (1995) Economic Restructuring in East Asia and India: Perspectives on Policy Reform (Basingstoke, Macmillan).

Ahluwalia, I.J. (1991) Industrial Growth in India (New Delhi, Oxford University Press). Ahluwalia, M.S (1999) India’s economic reforms: an appraisal, in: J.D. Sachs, A. Varshney & N.

Bajpai (Eds) India in the Era of Economic Reforms (New Delhi, Oxford University Press). Athukorala, P. (1998) Interest rates, saving and investment: evidence from India, Oxford Develop-

ment Studies, 26, pp. 153–169. Bell, C. & Rousseau, P. (2001) Post-independence India: a case of Ž nance-led industrialization,

Journal of Development Economics, 65, pp. 153–175. Bencivenga, V.R. & Smith, B.D. (1991) Financial intermediation and endogenous growth, Review

of Economic Studies, 58, pp. 195–209. Bhagwati, J. & Srinivasan, T.N. (1975) Foreign Trade Regimes and Economic Development: India

(New York, Columbia University Press). Blejer, M. & Khan, M.S. (1984) Government policy and private investment in developing

countries, IMF Staff Papers, 31, pp. 379—403. Chand, S. & Sen, K. (2002) Trade liberalisation and productivity growth: evidence from Indian

manufacturing, Review of Development Economics, 6, pp. 120—132. De Long, J.B. & Summers, L.H. (1991) Equipment investment and economic growth, Quarterly

Journal of Economics, 106, pp. 455–502. De Long, J.B. & Summers, L.H. (1992) Equipment investment and economic growth: How

strong is the nexus?, Brookings Papers on Economic Activity, 2, pp. 157–199. De Long, J.B. & Summers, L.H. (1993) How strongly do developing countries beneŽ t from

equipment investment?, Journal of Monetary Economics, 32, pp. 395–415. Edwards, S. (1998) Openness, productivity and growth: What do we really know?, The Economic

Journal, 108, pp. 383–398. Goldar, B. & Renganathan, V.S. (1990) Liberalisation of Capital Goods Imports in India, Working

Paper No. 8, National Institute of Public Finance and Policy, New Delhi. Gurley, J. & Shaw, E. (1955) Financial aspects of economic development, American Economic

Review, 45, pp. 515–538. Jones, C.I. (1994) Economic growth and the relative price of capital, Journal of Monetary

Economics, 34, pp. 359–382. Joshi, V. & Little, I.M.D. (1997) India’s Economic Reforms 1991–2001 (Oxford, Clarendon Press).

Trade Policy in India 331

King, R. & Levine, R. (1993) Finance and growth: Schumpeter may be right, Quarterly Journal of Economics, 108, pp. 717–737.

Levine, R. (1997) Financial development and economic growth: views and agenda, Journal of Economic Literature, 35, pp. 688–726.

Levine, R. & Renelt, D. (1992) A sensitivity analysis of cross-country growth regressions, American Economic Review, 82, pp. 942–963.

Lucas, R.E., Jr (1988) On the mechanics of economic development, Journal of Monetary Econom- ics, 22, pp. 3–42.

Pursell, G. (1992) Trade policies in India, in: D. Salvatore (Ed.) National Trade Policies (New York, Greenwood Press).

Rodrik, D. (1988) The limits of trade policy reforms in developing countries, Journal of Economic Perspectives, 6, pp. 87–105.

Rodrik, D. (1992) Closing the productivity gap: Does trade liberalization really help?, in: G.K. Helleiner (Ed.) Trade Policy, Industrialization and Development (New York, Oxford University Press).

Romer, P.M. (1986) Increasing returns and long-run growth, Journal of Political Economy, 94, pp. 1002—1037. Srivastava, V. (1996) Liberalization, Productivity and Competition: A Panel Study of Indian Manufacturing (New Delhi, Oxford University Press).

Temple, J. (1998) Equipment investment and the Solow model, Oxford Economic Papers, 50, pp. 39–62.

Tybout, J. (1992) Linking trade and productivity: new research directions, The World Bank Research Observer, 6, pp. 189–211.