SPECIAL ARTICLEEconomic & Political Weekly november 24, 200769Reforms and Regional Inequality in India Sabyasachi Kar, S SakthivelThis article analyses the impact of pro-market reforms on regional inequality in India, both at the aggregate and the sectoral level. It draws on the “new economic geography” literature that evaluates the impact of reforms on regional inequality in a country, in terms of the centripetal and centrifugal forces generated in the economy. The results show that regional inequality in India remained largely unchanged during the 1980s but rose dramatically after the adoption of the reforms. This is mainly due to the fact that the per capita output from the industrial and services sectors showed convergence before the reforms and divergence afterwards. After three decades of sluggish growth of about 3.5 per cent per annum in the post-independence period, the Indian economy attained an impressive growth rate of nearly 6 per cent over the 1980s and the 1990s. The 1980s are now known as the period when India moved from the Hindu rate of growth to a higher growth trajectory. Interestingly, although the average gross domestic product (GDP) growth rate for the two decades was almost identical, the policy paradigm during the 1980s was distinctly different from that of the 1990s. The policy during the 1980s focused on the public sector as the engine of growth, while the government regulated the private sector and its activities with industrial licensing requirements. These licences determined the scale, technology and location of investment projects and any plan to expand, relocate or change the output or input mixes needed prior permission from the government. There were some changes in the policy direction in the second half of the 1980s with partial liberalisation, but it was only in the early 1990s, that a complete paradigm shift took place. These included significant industrial and trade policy reforms with most of the controls in these sectors being dismantled and the state making a conscious effort to reduce its activities. The crux of these reforms was a greater thrust on privatisation and globalisation of the In-dian economy. Although this change in the policy paradigm did not lead significantly to different aggregate growth rates be-tween the 1980s and the 1990s, they did result in very dif-ferent sectoral growth rates during these two periods. The average agricultural growth rate during the 1980s was 4.4 per cent, but fell to a lowly 3 per cent during the 1990s. Similarly, the average industrial growth rate was 6.8 per cent in the 1980s but fell to 5.8 per cent during the next decade. On the other hand, the average services sector growth rate during the 1980s was 6.6 per cent, but rose sharply to a high 7.6 per cent in the 1990s. Clearly the agricultural and the industrial growth rates played signifi-cant roles in the 1980s, but the services sector was the engine of growth during the 1990s. In the context of the growth performance during these two decades, economists and policymakers have become interested in the trends in regional inequality during this period. Rising re-gional inequality can create economic and political problems for any country. For the Indian economy, it has serious ramifications for the continuation of the reforms process. Hence, it is of utmost importance to understand the impact of the reforms on the re-gional inequality of the economy. Moreover, to understand the nature of these impacts at a deeper level, it is also important to move from the aggregate to the sectoral level. More specifically, it is of interest to measure the contributions of each of the major The authors are grateful to Saon Ray and an anonymous referee for useful comments on the paper.Sabyasachi Kar (skar_ieg@yahoo.com) is associate professor and head, Development Planning Centre, at the Institute of Economic Growth, Delhi. S Sakthivel (shakti@phfi.org) is health economist at the Public Health Foundation of India, New Delhi.

SPECIAL ARTICLEnovember 24, 2007 Economic & Political Weekly70sectors to regional inequality in India and study the impact of the reforms on these sectoral contributions. This paper is an attempt to address these issues.The analysis of regional inequality gained prominence with the development of the neoclassical growth literature, based on the Solow model. According to this model, the process of capital accumulation involves a continuous fall in the per capita growth rate due to the falling marginal productivity of capital, until in the long run, it is equal to the rate of labour-augmenting techni-cal progress. As a result, the per capita income of regions continu-ously converge towards their steady state levels, resulting in a reduction in the income inequality among these regions over time. The empirical literature has tested this hypothesis both on multi-country datasets and on regional data within a single coun-try. The studies based on the regional economies within devel-oped countries have provided strong support to this convergence hypothesis. A number of studies have looked at economies such as theUS and Japan [Barro and Sala-i-Martin 1992; Barro and Sala-i-Martin 1995] and find evidence of regional convergence over long sample periods (100 years for US states and 60 years for Japanese prefectures) and also over much shorter sub-periods within the same sample. De la Fuente (2002) records evidence of convergence across Spanish regions in each of the three decades between 1965 and 1995. On the other hand, empirical evidence from developing countries has been much more mixed. Juan-Ramon and Rivera-Batiz (1996) study the states of Mexico over the period 1970 to 1993, and report convergence in incomes between 1970 and 1985 and divergence thereafter. Another study by Jian, Sachs and Warner (1996) finds the provinces of China between 1952 and 1993, and shows evidence of divergence in real per capita incomes, except for the period 1978 to 1990. There is an alternative theoretical literature on regional in-equality that focuses on the increasing returns associated with the process of growth that may lead to regional divergence. This is the literature on agglomeration economies [Duranton and Puga 2004; Rosenthal and Strange 2004]. The central proposi-tion of this literature is that economic activities in general and industrial activity in particular, tend to be attracted to particular geographical locations that are already developed compared to other regions. These are the result of agglomeration economies and lead to urban concentration in these regions whose income and growth rates keep diverging from those in the peripheral regions. Marshall (1920) has suggested three sources of such agglomeration economies. These are: (i) the sharing of inputs whose production involves internal increasing returns to scale; (ii) labour market pooling that allows a better match between an employer’s needs and a worker’s skills; and (iii) knowledge spill-overs between workers. Other sources of agglomeration, that have come up in the literature, include home market effects, economies of consumption, etc. There is some empirical literature SOCIETY FOR CAPITAL MARKET RESEARCH & DVELOPMENTList of Publications1. Indian Share Buyback Practices & Their Regulation: Effect on Share Prices, Dividends & Corporate Finance (2006) by L.C. Gupta, Naveen Jain & Anil Kumar ISBN 81-902410-1-X Price Rs. 200/-3. How Good are Mutual Funds: The Household Investors’ Perceptions (2001) by L.C. Gupta, Utpal Choudhury ISBN 81-900513-9-3 Price Rs. 270/-5. Returns on Indian Equity Shares (2000) by L.C. Gupta assisted by Utpal Chodhury ISBN 81-900513-7-7 Price Rs. 300/- 7. Indian Stock Market P/E Ratios (1998) by L.C. Gupta, P.K. Jain and C.P. Gupta ISBN 81-900513-5-0 Price Rs. 360/- 9. Shareowners’ Geographic Distribution (1994) by L.C. Gupta, Naveen Jain and Yash Kulshreshtha ISBN 81-900513-3-4 Price Rs. 390/-11. Stock Exchange Trading in India: Agenda for Reform (1992)by L.C. Gupta ISBN 81-900513-1-8 Price Rs. 200/-2. * India’s Stock Market & Household Investors 2001-04 by L.C. Gupta, Naveen Jain & Utpal Choudhury ISBN 81-902410-0-1 Price Rs. 300/-4. Indian Households’ Investment Preferences (2001) by L.C. Gupta, C.P. Gupta and Naveen Jain ISBN 81-900513-8-5 Price Rs. 360/- 6. India’s Financial Markets & Institutions (1999) edited by L.C. Gupta ISBN 81-900513-6-9 Price Rs. 390/-8. * The Roots of India’s Stock Market Crisis, 1995-97 by L. C. Gupta (1997) ISBN 81-900513-4-2 Price Rs.180/-10. Mutual Funds and Asset Preference: Household Survey (1993) by L.C. Gupta; ISBN 81-900513-2-6; Price Rs. 250/-12. Indian Shareowners: A Survey (1991) by L.C. Gupta ISBN 81-900513-0-X Price Rs. 180/- * Out of print Teachers, Researchers and Professionals who are Associate Members of the Society and have paid Membership Fee of Rs. 200/-in a year will be allowed 50% Special Discount on the list price without any charge for postage. They will also get the following booklets free of charge: (1) How the P/E Ratio Can Really Help You (22 pages); (2) How to Select Companies for Your Share Portfolio (31 pages); and (3) How Mutual Funds Work (26 pages).Please write to: Dr. L.C. Gupta, Director, Society for Capital Market Research & Development, 10-D BigJo’s Tower, Netaji Subhash Place, Pitampura, New Delhi – 110034. Tel: 27351478; Fax: 27357407; E-mail: scmrd@bol.net.in; Website: hĴp://www.scmrd.org

0.2 0.3 0.4 0.5 0.6 1980-811982-831984-851986-871988-891990-911992-931994-951996-971998-99 Unweighted Weighted

SPECIAL ARTICLEnovember 24, 2007 Economic & Political Weekly72hand, a few studies have found that there is no evidence of regional divergence in the 1990s. According to Singh et al (2003), although there are indications of rise in regional disparities, “but they are neither uniform nor overly dramatic”. Similarly, Dholakia (2003) reports no significant trend for rise or fall of disparity in the 1990s. The theoretical literature on regional inequality is based on the assumptions that may not be equally relevant for all sectors of the economy. In particular, the literature on agglomeration economies is mainly about the industrial sector. Hence, different sectors may contribute very differently to changes in regional in-equality. There are a few studies that have looked at the contribu-tion of different sectors in the context of changing regional in-equality in India. Examining the roles of each sector in aggregate divergence, Das and Barua (1996) find that for the period from the 1970s to early 1990s, agriculture and services are the crucial sectors that contributed significantly to higher regional inequality. Rao et al (1999), suggest that primary sector was largely res-ponsible for the rise in regional disparity from mid-1960s until 1990. However, the standard deviation in secondary sector was stable during the same pe-riod, suggesting the stabilising role of secondary sector in aggregate inequality. Further, they point out that from the early 1990s to mid-1990s, the primary sector had a limited role in the growing inequality in the economy, but the second-ary sector played a significant role. As far as the tertiary sector is concerned, there is no con-sistent trend in accentuating or offsetting regional divergence. Similarly, Dasgupta et al (2000) have studied the period from 1970 to 1995 and concluded that it is primarily the agricultural sector, and to a certain extent themanu-facturing sector, that had important roles to play in the regional divergence over this period, while the tertiary sector had a stabilising influence, with regional inequality decreasing in this sector. It may be noted that the literature on regional inequality in India is motivated completely by the neoclassical growth frame-work and accordingly, tests for absolute or conditional conver-gence of regions over time. There is a dearth of contributions that use the “new economic geography” framework to analyse regional inequality in India, and study the impact of the pro-market re-forms adopted in the early 1990s, on regional inequality during this period. Secondly, although there are a few papers that analyse the contribution of individual sectors to regional divergence, none of them study the effect of the reforms on the contributions from these sectors. This paper is an attempt to fill this lacuna and hence has two objectives. Firstly, it triestoanalysetheeffectof reforms on the trend in regional inequalityinIndia.Thisanalysis is based on a comparison of the trends in regional inequality between the 1980s and the 1990s – two decades that are similar in terms of their growth performance,butdistinctintheirpolicy stancevis-à-visreforms.Secondly, this paper measures the sectoral contributions to regional inequality, and by using these measurements, studies the effect of the reforms at the sectoral level. Section 1 analyses the trends in regional inequality in India during the 1980s and the 1990s, i e, before and after the reforms. Section 2 describes a framework that is used to decompose total regional inequality into the sectoral contributions. Section 3 analyses the impact of the reforms on regional inequality at the sectoral level. Section 4 concludes the paper.1 Reforms and Regional InequalitySince the economic policies were still significantly protectionist during the 1980s and major reforms were initiated only in the early 1990s, we treat these two decades as the pre-reform and post-reform period for the Indian economy. Going by Krugman and Livas Elizondo (1996), we should find an increasing regional disparity throughout the 1980s and a decreasing disparity during the 1990s. Paluzie (2001), on the other hand, would predict a decreas-ing regionaldisparity inthe first decade and increasing region-al disparities in the next one. We test these contradictory hypotheses by studying the regional inequality among 17 major states of India, during the period 1980-81 to 1999-2000. The regional inequality is measured by the coefficient of variation of the per capita income of these states. The figure (p 71) shows the (population) weighted and unweighted coefficient of variation for the given period. From this figure, it becomes clear that regional inequality re-mained largely unchanged during the 1980s but showed a dis-tinctly rising trend during the 1990s. Furthermore, these trends hold whether we consider population-weighted or unweighted measures of inequality, indicating the robustness of these trends. What do the trends tell us about the applicability of the two con-trasting hypotheses discussed above, to the Indian economy? Since the trend remained largely unchanged during the 1980s, the result does not support either of the two hypotheses for the pre-reform phase. Clearly the centripetal and the centrifugal forces balanced each other out during this period. It may be noted that while the agglomeration economies provided the centripetal force, the centrifugal forces were probably gener-ated by the government’s efforts to achieve a spatially balanced growth by channelling development towards backward areas. In the 1990s, however, the rise in regional inequality clearly supports the hypothesis from Paluzie (2001), which predicts increasing regional disparity following reforms. This implies that Table 1: Aggregate Regional Inequality and Its Sectoral Decomposition (1980s)Years 1980-811981-821982-831983-841984-851985-861986-871987-881988-891989-90Contribution of agricultural sector 0.060.080.07 0.06 0.060.05 0.04 0.06 0.060.07Contribution of industrialsector0.09 0.09 0.08 0.09 0.09 0.10 0.10 0.09 0.090.10Contribution of servicessector0.11 0.11 0.11 0.11 0.110.120.120.12 0.110.12Aggregate inequality 0.26 0.27 0.27 0.26 0.26 0.27 0.26 0.27 0.260.29Table 2: Aggregate Regional Inequality and Its Sectoral Decomposition (1990s)Years 1990-911991-921992-931993-941994-95 1995-961996-971997-981998-991999-2000Contribution of agricultural sector 0.05 0.050.070.060.06 0.050.06 0.05 0.050.04Contribution of industrialsector0.110.10 0.12 0.12 0.12 0.13 0.130.14 0.13 0.13Contribution of servicessector0.120.13 0.14 0.16 0.16 0.180.17 0.18 0.18 0.19Aggregate inequality 0.28 0.280.320.340.34 0.36 0.36 0.36 0.360.37

SPECIAL ARTICLEEconomic & Political Weekly november 24, 200773the centripetal forces became stronger than the centrifugal forces during this period. As we shall discuss in some detail in Section3, this may have happened because the reforms weakened the centrifugal forces and strengthened the centripetal forces in a number of ways. 2 A Decomposition FrameworkIn the previous section, we find that the trend in regional ine-quality in India changed significantly following the reforms. Did the sectoral contributions to regional inequality show similar changes in their trends as a result of these reforms? In order to answer this question, we need a framework that will allow us to decompose the aggregate inequality, giving the contribution of each sector. Since we have used the coefficient of variation to measure regional inequality in the previous section, we apply a decomposition framework discussed in De Janvry and Sadoulet (2001) that uses the same statistic to measure inequality. Let there be n regions such that the aggregate output of each region is given by Xi, i = 1….n. Let there be m sectors that contribute to each region’s aggregate output Xi, such that the output of each sector in each region is given by Xij, i = 1…n, j = 1…m. Then,Xi = Σj Xij ...(1) _ _Let X be the arithmetic mean of Xi and Xj be the arithmetic mean of Xij. Next, define Pj as the ratio between the average output of the jth sector and the average output of the economy. XjThus, Pj = …(2) XLet C(Xi) be the coefficient of variation of aggregate output and let C(Xij) be the coefficient of variation of the jth sector’s output, across regions. Also, let rij,i be the correlation coefficient between the jth sector’s output and the aggregate output, across regions. Then, the percentage decomposition of total inequality is given by De Janvry and Sadoulet (2001): C(Xij)Σ(Pj × rij, i× ) = 1 …(3)j C(Xi)Rearranging equation (4) we can write C(Xi) = Σ (C(Xij) × Pj × rij, i ) …(4) jEquation (4) indicates that the aggregate inequality in an economy (measured by the coefficient of variation of aggregate output across regions) can be decomposed to give each sector’s contribution. Furthermore, the contribution of each sector is equal to the product of (i) the inequality within the sector (mea-sured by the coefficient of variation of the particular sector’s out-put across regions), (ii) the relative size of the sector (measured by the average output of the sector as a proportion of the average output of the economy), and (iii) the strength of the linkages be-tween the sector and the economy (measured by the correlation coefficient between the sector’s output across regions and the ag-gregate output across regions). This means that the inequality for the aggregate economy is affected not only by the sectoral inequalities, but also by the relative size of the sectors and their interlinkage with the economy. The size of the sectors adds a scale effect to the sectoral inequality, i e, a larger sector adds more to the economy’s inequality compared to a smaller sector. The interlinkages of a sector with the whole economy – repre-sented by the correlation coefficient between the two – also has an important role. This is due to the fact that a high correlation between any sector and the economy implies that a region that has a relatively high output from that sector also has a relatively high aggregate output, while a region that has a relatively low output from that sector also has a relatively low aggregate out-put. Thus, for a given level of inequality in the sector, an increase in the correlation coefficient increases the economy’s inequality.3 Role of the SectorsWe shall now use the above framework to look at the comparative contributions of the agricultural, industrial and the services sec-tor in changing the regional inequalities in India during the last two decades. Most studies dealing with the regional dimensions of the Indian economy consider the states as the appropriate unit of their analysis and base their study on the major states of India. In consonance with this approach, we use data from 17 major Indian states, i e, Andhra Pradesh, Assam, Bihar, Gujarat, Karnataka, Kerala, Maharashtra, Madhya Pradesh, Orissa, Punjab, Rajasthan, Tamil Nadu, Uttar Pradesh, West Ben-gal, Goa, Haryana and Himachal Pradesh. In order to generate data on aggregate and sectoral output of the states, we use GSDP data classified by industry of origin. The agricultural output is derived by aggregating over ag-riculture, forestry and logging and fishing, while the industrial output is the aggregate of mining and quarrying, registered and unregistered manufacturing, construction and electricity, gas and water supply. The services sector comprises transport, stor-age and communication, trade, hotels and restaurants, banking and insurance, real estate, ownership of dwellings and business services, public administration and other services. The GSDP and the population data series for this study are obtained from the Central Statistical Organisation(CSO). The time period chosen for the study is 1980-81 to 1999-2000. It may be noted that, in or-der to avoid the complexity arising out of reorganisation of states (which resulted in a number of new states) we have used data till 1999-2000, which is the last year for which data from the undi-vided states (such as, Uttar Pradesh, Bihar and Madhya Pradesh) are available. For the rest of the exercise, we have used popula-tion-weighted measures of inequality. Tables 1 and 2 (p 72) ¯¯Table 3: Agricultural Sector’s Contribution to Aggregate Inequality and Its ComponentsYears Intra-SectorRelativeInter-Sectoral InequalitySizelinkageContribution1980-81 0.26 0.430.55 0.061981-82 0.28 0.430.65 0.081982-83 0.29 0.410.61 0.071983-84 0.26 0.420.58 0.061984-85 0.26 0.410.58 0.061985-86 0.27 0.390.49 0.051986-87 0.28 0.370.42 0.041987-88 0.29 0.350.55 0.061988-89 0.27 0.360.60 0.061989-90 0.29 0.350.67 0.071990-91 0.27 0.340.57 0.051991-92 0.30 0.330.52 0.051992-93 0.31 0.330.69 0.071993-94 0.30 0.320.60 0.061994-95 0.29 0.320.61 0.061995-96 0.30 0.300.57 0.051996-97 0.30 0.300.66 0.061997-98 0.31 0.280.55 0.051998-99 0.30 0.270.62 0.051999-2000 0.32 0.26 0.52 0.04

SPECIAL ARTICLEnovember 24, 2007 Economic & Political Weekly74epw RF Ad- Full Page

SPECIAL ARTICLEEconomic & Political Weekly november 24, 200775present the sectoral decomposition of aggregate inequality according to equation (4), for the 1980s and 1990s respectively. From Table 1, we find that it is not only the aggregate inequality that remained almost unchanged (around 0.26) during the 1980s, but also the contribution of each of the three sectors that remained largely unchanged during this period. The only year for which aggregate inequality shows a marked rise is the last year of the decade and this is due to similar rises in the contribution of all the three sectors. Compared to this, Table 2 shows that after the re-forms, while agriculture’s contribution again remained almost un-changed (around 0.06), industrial contribution showed a rising trend (from 0.11 to 0.13) while services sector contribution rose quite sharply (from 0.12 to 0.19), resulting in the consistent rise in aggregate inequality throughout the 1990s (from 0.28 to 0.37). It is clear from Tables 1 and 2 that the reforms had no impact on the contribution to regional inequality from the agricultural sector. As far as the industrial and the services sectors are con-cerned, the reforms completely changed the trends in their contri-butions from the constant trends during the 1980s to the sharply increasing trends during the 1990s. We can throw more light on this issue by looking at the trends in the values of the three factors that determine the sectoral con-tributions, i e, (a) the coefficient of variation of the sector, (b) the relative size of the sector, and (c) the strength of the interlinkage be-tween the sector and the aggregate economy. Tables 3, 4 and 5 present the break-up of the sectoral contribution in terms of these three factors for the two decades. Table 3 (p 73) rep-resents the agricultural sector, Table 4 gives the values for the industrial sector and Table 5 gives those for the services sector.From Table 3, we find that in the agricultur-al sector, there was a moderate upward trend in regional inequality within the sector (from 0.26 to 0.32). Despite this upward trend, a strong decline in its relative size (from 0.43 to 0.26) ensured that the contribution of this sec-tor to regional inequality remained almost constant over the two decades (around 0.06). The rising trend in regional inequality within the sector was partly due to the high growth rate of agriculture in states that had a bigger agricultural sector and partly due to stagna-tion and shrinking of the sector in agricultur-ally backward states like Bihar and Orissa. Since the location of agricultural production is tied to arable land, this divergence cannot be explained by the agglomeration economies discussed in the literature. One plausible ex-planation for this divergence may be that the more prosperous agricultural states have the surplus to make the necessary investment in irrigation, warehouses, cold storages and other infrastructure that sustains the growth rate of agriculture in these states, while the poorer states do not make any surplus from this sector, and hence, stagnate or shrink due to the absence of sufficient investment. The important thing to note is that the reforms seem to have had no effect on the diver-gence of this sector or on its contribution to regional inequality. Tables 4 and 5, on the industrial and the services sector res-pectively, show very similar and interesting trends. It is clear from these tables that there was a distinct fall in regional disparity in the pre-reform period and a distinct rise in regional disparity in the post-reform period, within both these sectors. The tables also show that the relative size of the two sectors and their interlink-ages to the economy grew constantly over the two decades. While the reforms seem to have had no effect on them, they had the ef-fect of dampening the centrifugal forces in the pre-reform period and strengthening the centripetal forces in the post-reform period.Tables 1 and 2 show that aggregate regional inequality remained almost unchanged before the reforms, although it did increase afterwards. Compared to this, Tables 4 and 5 show that the indus-trial and services sectors both show a fall in inequality before the reforms and a rise after the reforms. Clearly, a sectoral analysis pro-vides far stronger support for the Paluzie (2001) hypothesis, than does the analysis at the aggregate level, specifically for the pre-re-form period. The reason for this is that in the pre-reform period, the government’s policies to check regional divergence must have tar-geted the industrial and services sectors spe-cifically, rather than the aggregate economy. Thus the regional disparity within these two sectors went down, while that within the agri-cultural sector was continuously rising, resulting in the aggregate regional inequality remaining unchanged during this period. It may be noted that the policies targeted to check regional di-vergence worked through multiple channels. One channel worked through the public sector, where, a sizeable part of the public invest-ments were made in relatively backward areas. The other channel worked through the private sector, which was encouraged through the use of fiscal incentives and industrial licensing, to invest in these areas. In other words, the state played a crucial role in bringing down inequal-ity in these two sectors during this period.In the post-reform period, the industrial and service sectors show the divergence due to a number of reasons. Firstly, the industrial licensing system, which also determined the location of investments, was dismantled as a part of the reforms. This gave the private sec-tor the freedom to choose its location and minimise the transportation costs to large do-mestic markets, à la Krugman and Livas Elizondo (1996). The result was a shift of their production base to the metropolitan areas of Table 4: Industrial Sector’s Contribution to Aggregate Inequality and Its ComponentsYears Intra-SectorRelativeInter-Sectoral InequalitySizelinkageContribution1980-81 0.44 0.260.82 0.091981-82 0.42 0.250.84 0.091982-83 0.39 0.260.82 0.081983-84 0.41 0.260.86 0.091984-85 0.40 0.260.87 0.091985-86 0.42 0.260.87 0.11986-87 0.43 0.270.87 0.11987-88 0.39 0.270.86 0.091988-89 0.39 0.270.89 0.091989-90 0.41 0.270.91 0.11990-91 0.41 0.280.92 0.111991-92 0.40 0.280.91 0.11992-93 0.44 0.280.95 0.121993-94 0.46 0.280.93 0.121994-95 0.45 0.280.94 0.121995-96 0.49 0.290.93 0.131996-97 0.50 0.280.94 0.131997-98 0.48 0.300.95 0.141998-99 0.48 0.290.94 0.131999-2000 0.49 0.30 0.93 0.13 Table 5: Services Sector’s Contribution to Aggregate Inequality and Its ComponentsYears Intra-SectorRelativeInter-Sectoral InequalitySizelinkageContribution1980-81 0.39 0.31 0.88 0.111981-82 0.38 0.320.87 0.111982-83 0.37 0.330.90 0.111983-84 0.37 0.33 0.88 0.111984-85 0.36 0.34 0.90 0.111985-86 0.37 0.35 0.92 0.121986-87 0.35 0.36 0.91 0.121987-88 0.33 0.380.94 0.121988-89 0.33 0.36 0.92 0.111989-90 0.34 0.38 0.93 0.121990-91 0.34 0.38 0.93 0.121991-92 0.35 0.400.94 0.131992-93 0.37 0.400.93 0.141993-94 0.43 0.400.94 0.161994-95 0.43 0.40 0.93 0.161995-96 0.45 0.41 0.95 0.181996-97 0.43 0.41 0.94 0.171997-98 0.44 0.430.95 0.181998-99 0.43 0.44 0.94 0.181999-2000 0.45 0.45 0.96 0.19

SPECIAL ARTICLEnovember 24, 2007 Economic & Political Weekly76the richer states, which covered most of the large domestic markets in the country.Secondly, the reforms also gave a boost to export-oriented production. From balance of payments data provided by the Reserve Bank of India(RBI), we find that in the commodity- producing sector (i e, agriculture and industry), the average exports toGDP ratio rose from about 8.5 per cent in the 1980s to about 15.5 per cent in the 1990s. Within industry, the manufac-turing exports sector needed to minimise transportation costs to international markets, and hence, preferred to locate itself near the coastal areas that had good infrastructure. Since these facilities were mostly available in the relatively developed states in the western and the southern parts of the country, this preference increased the regional inequality between these states and the poorer ones. There was a modest rise in the average exports ratio in the services sector as well, from about 4 per cent in the 1980s to about 5.5 per cent in the 1990s. The services exports sector, in particular, the information techno-logy exports and the financial services exports, needed highly developed telecommunication infrastructure and high quality human capital. Clearly, both of these were available only in the metropolitan areas of the country, and resulted in the growth of the services sectors in these areas. This again increased the gap between these areas and the relatively underdeveloed regions of the country. The reforms clearly strengthened the centripetal forces through these channels. On the other hand, the reforms also weakened the centrifugal forces in the economy. This is due to the fact that with the reforms, the focus of the public sector shifted to providing utilities and infra-structure. In order to illustrate this change, we have analysed the data onGDP from the public sector and its components, provided by the CSO. We find that within the public sector, the average share of electricity, gas and water (i e, utilities) in total industrialGDP went up from about 24 per cent during the 1980s to about 32 per cent during the 1990s, while the share of all the other sectors within industry declined. Similarly, the average share of banking and insurance (a part of the financial infra-structure of the economy) in total servicesGDP from the public sector, went up from about 12.7 per cent during the 1980s to about 19.5 per cent during the 1990s, while the share of all the other sectors within the services sector declined. Since these utilities and infrastructure are mostly required in developed areas and richer regions, a shift towards them diminished the capacity of the government to check regional divergence through public investments. 4 ConcludingRemarksRegional inequality in India remained unchanged during the 1980s mainly due to a fall in inequality within the industrial and the services sectors during this period. Even though the relative size and interlinkages of these sectors with the economy was rising during this period, the fall in inequality

SPECIAL ARTICLEEconomic & Political Weekly november 24, 200777within the sector was sufficient to keep a check on rising inequalities. The rise in regional inequality during the 1990s is largely due to a sharp rise in inequality in the industrial and services sectors. The rising relative size and interlinkages of these sectors with the economy also contributed to rising inequality in this period. A number of important conclusions can be drawn from these results. Firstly, it is clear that there is no fixed relationship be-tween high growth rates and trends in regional inequality in the Indian context. This is underlined by the fact that while the high rates of growth led to diminishing inequality in the indus-trial and the services sector during the 1980s, the reverse took place during the 1990s. This seems to indicate that it is not the rate of growth but its composition that determines the effect it will have on regional inequality. The second point that emerges is that contrary to the findings of earlier studies that the agri-cultural sector contributed significantly to divergence, for the two decades under study, this sector kept a check on the growth of regional inequality in the Indian economy. Thirdly, and perhaps most importantly, the downward trends in inequality within the non-agricultural sectors in the 1980s and the rever-sal of these trends after the reforms indicate a strengthening of the agglomeration economies in these sectors. We postulate that there are two sets of factors that have generated this outcome. Firstly, in the pre-reform period, the public sector had played a crucial role in maintaining regional equality in the Indian economy by directing resources to backward areas. With a change in the focus of the public sector following the re-forms, this process has become weaker. Secondly, the reforms gave greater freedom and impetus to the private sector and export-oriented production. These sectors, which were attempt-ing to reduce costs and become competitive, were attracted to the areas that were relatively more developed. As a result, investment and activity shifted to these areas, strengthening the forces of divergence. Of course, it is debatable how long these centripetal forces, un-leashed by the reforms process, can continue to generate rising trends in regional inequality. Rising land-rent, rising wages, con-gested urban infrastructure, etc, will increasingly force invest-ment to relocate to less developed areas. Public investment can also help less developed regions to become more attractive in-vestment destinations, by creating better infrastructure within these regions and connecting them to the markets. It would be important to study whether there are already some indications of these trends in the Indian economy. These issues are however, beyond the scope of the present paper. REFERENCESAhluwalia, M S (2000): ‘Economic Performance of States in Post-reforms Period’, Economic and Political Weekly, May.– (2002): ‘State-Level Performance under Economic Reforms in India’ in A O Krueger (ed), Economic Policy Reforms and the Indian Economy, Oxford University Press, New Delhi.Barro, R J and X Sala-i-Martin (1992): ‘Convergence’, Journal of Political Economy, Vol 100, No 2, pp 223-51.– (1995): Economic Growth, McGraw-Hill, New York.Bhattacharya, B B and S Sakthivel (2003): ‘Regional Disparities in Growth: A Comparison of Pre- and Post-Reform Period’, Discussion Paper No 73/2003, Institute of Economic Growth, Delhi. – (2004): ‘Regional Growth and Disparity in India’, Economic and Political Weekly, March.Cashin P and R Sahay (1996): ‘International Migration, Centre-State Grants, and Economic Growth in the States of India’, IMF Staff Papers, Vol 43, No 1.Ciccone, A (2002): ‘Agglomeration Effects in Europe’, European Economic Review, Vol 46, pp 213-27.Ciccone, A and R E Hall (1996): ‘Productivity and the Density of Economic Activity’, American Economic Review, Vol 86, pp 54-70.Das S K and A Barua (1996): ‘Regional Inequalities, Economic Growth and Liberalisation: A Study of the Indian Economy’, The Journal of Development Studies, Vol 32.Dasgupta, D, P Maity, R Mukherjee, S Sarkar and S Chakro-borty (2000): ‘Growth and Inter-State Disparities in India’, Economic and Political Weekly, 35, 27, pp 2413-22.De Janvry A and E Sadoulet (2001): ‘Income Strategies among Rural Households in Mexico: The Role of Off-farm Activities’, World Development, Vol 29, No 3, pp 467-80.De la Fuente, A (2002): ‘Regional Convergence in Spain, 1965-95’, CEPR Discussion Paper, No 3137, Centre for Economic Policy Research, London, UK.Dekle, R and J Eaton (1999): ‘Agglomeration and Land Rents: Evidence from the Prefectures’ Journal of Urban Economics, Vol 46, pp 200-14.Drescher, J (1999): ‘Income Inequality Decomposition by Income Source and by Population Subgroups: A Theoretical Overview and the Empirical Case of Denmark’, Luxembourg Income Study Working Paper No 209, INSTEAD, Luxembourg.Duranton, G and D Puga (2004): ‘Micro-foundations of Urban Agglomeration Economies’ in J V Henderson and J-F Thisse (eds), Handbook of Urban and Regional Economics.Fujita, M and D Hu (2001): ‘Regional Disparity in China 1985-94: The Effects of Globalisation and Economic Liberalisation’, Annals of Regional Science 35, 1, pp 3-37.Glaeser, E L, H D Kallal, J A Scheinkman and A Shleifer (1992): ‘Growth in Cities’, Journal of Political Economy, Vol 100, pp 1126-52.Hanson, G (1992): ‘Industry Agglomeration and Trade in Mexico’, Massachusetts Institute of Technology, Massachusetts, PhD thesis.Henderson, J V, A Kuncoro and M Turner (1995): ‘Industrial Development in Cities’, Journal of Political Economy, Vol 103, pp 1067-85.Jenkins, S P (1995): ‘Accounting for Inequality Trends: Decomposition Analysis for the UK, 1971-86’, Economica, Vol 62. Jian, T, J Sachs and A Warner (1996): ‘Trends in Regional Inequality in China’, China Economic Review, 7, 1, Spring, pp 1-21.Juan-Ramon, V H and L A Rivera-Batiz (1996): ‘Regional Growth in Mexico:1970-93’, IMF Working Paper, No WP/96/92.Krugman, P (1991): ‘Increasing Returns and Economic Geography’, Journal of Political Economy, 99, 3, pp 483-99.Krugman, P and R Livas Elizondo (1996): ‘Trade Policy and the Third World Metropolis’, Journal of Development Economics, 49,1, pp 137-50.Marjit S and S Mitra (1996): ‘Convergence in Regional Growth Rates: Indian Research Agenda’, Economic and Political Weekly, August.Marshall, A (1920): Principles of Economics, MacMillan,London.Mitra, A, A Varoudakis and M Veganzones-Varoudakis (2002): ‘Productivity and Technical Efficiency in Indian States Manufacturing: The Role of Infrastructure’, Economic Development and Cultural Change, Vol 50, No 2.Nagaraj, R, A Varoudakis and M Veganzones (1998): ‘Long-Run Growth Trends and Convergence Across Indian States’, Technical Paper, No 131, OECD Development Centre, Paris, France.Paluzie, E (2001): ‘Trade Policy and Regional Inequalities,’ Papers in Regional Science, 80, pp 67-85.Rao, M Govinda, R T Shand and K P Kalirajan (1999): ‘Convergence of Incomes across Indian States’, Economic and Political Weekly, March. Rosenthal, S S and W C Strange (2004): ‘Evidence on the Nature and Sources of Agglomeration Economics’ in J V Henderson and J-F Thisse (eds), Handbook of Urban and Regional Economics, Vol 4, Elsevier-North Holland, Amsterdam.Shand, R T and S Bhide (2000): ‘Sources of Economic Growth: Regional Dimensions of Reforms’, Economic and Political Weekly, October.Shetty, S L (2003): ‘Growth of SDP and Structural Change in State Economics: Interstate Comparisons’, Economic and Political Weekly, December.Singh, N, L Bhandari, A Chen and A Khare (2003): ‘Regional Inequality in India: A Fresh Look’, Economic and PoliticalWeekly, March.Venables, A J and R Kanbur (2003): ‘Spatial Inequality and Development’, Working Paper No 2003-38, Department of Applied Economics and Management, Cornell University, Ithaca, New York, US.Venables, A J (2003): ‘Spatial Disparities in Developing Countries: Cities, Regions and International Trade’, Discussion Paper No CEPDP 0593, Centre for Economic Performance, London School of Economics, UK. Centre for Research, Planning & Action (CERPA)Requires (Retd) IES, ISS, IAS or equivalent persons. CERPA is UNECOSOC accredited & Centre of excellence for HRD apprd by UNESCSAP. For details visit http://www.cerpaindia.orgPostings are in senior most management cadre.Apply secretary, CERPA, 16 Dakshineshwar, 10, Hailey Road, New Delhi 110001 Tel/23355170 Tele Fax: 23329216 e mail: spahuja@nde.vsnl.net.in