Document(s) 1 of 16

The process of economic reform and globalization in India

India embarked on a policy of liberalization and globalization in the latter part of the last century. There has been some discussion in the literature as to when India took steps to move away from the regime of comprehensive state control of the economy and dismantle the restrictive structure. A distinction has been made in this connection between 'reforms' and 'globalization'. Strictly speaking the former is supposed to emphasize the process of easing control of the domestic economy, while the latter refers to the attempts at liberalization on the external account. It is useful to keep the two sets of policy distinct, and will be referred to below.

In practice the former entails the latter. As Nayar (2006a, p. 10) observes:

Economic liberalization within a country creates pressures to integrate the national economy with the world economy . . . Say, for example, a country commences economic reform and removes restrictions on production by the private sector to accelerate growth. Eventually the state would have to allow imports of capital goods and intermediate goods to increase production–and that means integration into the world economy. And if it allows imports of these goods, then it must also promote exports in order to pay for them–further integration into the world economy.

In fact the process is more extensive than suggested in the above quotation. External liberalization also involved removing a good deal of restrictions on the import of consumer goods, not just capital and intermediate goods to aid production. The motivation for this was to promote competition in the domestic economy, and bring the efficiency levels in the Indian economy nearer the levels of the world economy.

It is clear that reforms of the domestic economy started earlier in India. Rodrick and Subramanain (2004) date the beginning of this process to the return of Indira Gandhi to power at the beginning of the 1980s. They ascribe the new direction to an 'attitudinal' shift in the perception of the leaders after the Congress Party had been 'chastened' by its electoral defeat in the earlier election of the late seventies. They also find a 'structural break' in several key indicators including GDP growth in this period. When the party was returned to power in January 1980 it became more inclined to support growth with the help of a more dynamic private sector. Nayar (2006b) maintains that the reform process started earlier in 1975–1976 during the regime of Indira Gandhi herself. The leadership was jolted partly by the turmoil created by the excesses of the 'dirigist' policy followed in the years 1969–1973 (including large-scale nationalization of banking and industry), and partly by external shocks (including war, droughts and oil price inflation). Besides adopting deflationary policy to stabilize the economy, the Gandhi administration undertook deregulation and export-promotion measures on top of the earlier devaluation (Joshi and Little 2000, p. 56).

While the period of 'creeping liberalization' might have been a prolonged one, it was not till the economic crisis of 1991 that there was an open endorsement of 'paradigm shift' embracing a policy of integration with the world economy and recognition of the need to follow the path of the South-East Asian growth strategy. It involved a sharp devaluation of the rupee; removal of quantitative restrictions on imports; reduction of import tariffs; and a unification of the exchange rate as the rupee was made convertible for current-account transactions. On the domestic front of the reform process the system of industrial licensing was removed and the list of items reserved for the small-scale producers was shortened considerably. The program also saw fiscal reforms though the maintenance of important subsidies, particularly on the agricultural front, continued to plague the budget (Ahluwalia 2002; Joshi and Little 2000).

The removal of quantitative restrictions on imports (QRs), an important feature of the controlled economy, came gradually in the decade of the nineties. QRs on intermediate and capital goods were removed in 1991, but they remained significant on a range of consumer goods. Over the next ten years, a series of international negotiations, starting with the 'Uruguay Rounds' of the WTO, saw a gradual whittling down of these barriers. Tariffs do, however, remain as a deterrent to imports on a variety of goods. Table 1.1 is taken from a World Bank Report (Dahlman and Utz 2005) give the extent of the tariff barrier relative to some comparators in the early years of the century.

India's tariff rates remain high by the standards of other developing countries (Table 1.1). But a fair amount of integration with the world economy has been achieved. The following paragraphs briefly discuss the extent of the integration both in terms of the current and capital account of the balance of payments.

External trade

India's merchandise exports had a steep decline during the autarkic regime going down from 2.17 percent of world exports in 1949 to 0.44 in 1980. It hovered around 0.50 percent throughout the decade, and started going up only after 1991. Liberalization of trade has certainly had the impact of starting an upward trend and the share had reached a high of 0.8 in 2004. The share still remains quite low relative to comparator countries in Asia. China increased its share from under 2 percent in 1990 to close to 6 percent in 2003. Even a much smaller country like Korea had a share of 2.8 percent at this date (Table 1.2).

Manufactured exports have been a substantial part of the Indian export growth–reaching 74 percent in 2004 (government of India, Economic Survey–2005–2006). India seems to have performed relatively better in service exports. The gap between India and comparator countries in service exports, particularly vis-à-vis China, is not as large. India's progress in exports in computer and communications services has been much more than China's–which has done better in travel and related services (Figure 1.1).

Figure 1.1 Merchandise and service exports, India and comparators, 2002 (source: World Bank staff analysis using World Bank internal database).

Import volume has generally kept slightly ahead of export volume (Table 1.3). India has been helped in its current account by the terms of trade tilting in its favor in a majority of the years (though in very recent years there is a threat of significant deterioration of the TOT). But in any event the balance of payments position has been helped, increasingly so in recent years, by substantial inflow of foreign funds. This is due to another aspect of India's globalization–the substantial emigration of its nationals and the inflow of remittance from the overseas residents of Indian origin.

Foreign-capital inflow other than remittance has in fact not been significant. In fact India has been a significant laggard in attracting foreign direct investment. Even though the actual value of FDI in India has increased several times from its level before liberalization, it is quite small compared with global trends.

At its height in recent years it has perhaps been no more than 1 percent of GDP–compare to 4.7 percent in China in 2001 and 4.4 percent in Brazil. In value terms India received $4.26 billion in FDI in 2003, compared with $53.5 billion for China (Dahlman and Utz 2005, pp. 30–31). Dahlman, however, reports that according to the Foreign Direct Investment Confidence Index (by A.T. Kearney) India's attractiveness to foreign investors is rapidly rising–although it has been well below China's for sometime.

It is clear from this capsule account that globalization in the sense of integration with the world economy has been significant both on the trade and the capital account. It is equally clear that it has been accompanied by a spurt in the growth rate of GDP. Further, the efficiency of the economy has increased in the aggregate. The slow growth of India's constant price investment ratio (increasing from around 22 percent in the 1970s and the 1980s to no more than 24 percent at the end of the 1990s), while GDP growth was accelerating suggests that the marginal capital–output ratio must have been rising significantly. According to one author this ratio was a meager 0.12 in India's worst decade 1965–1974, but it had doubled to 0.246 in the decade of 1991–2000 (Berry 2006, p. 3).

In spite of this positive effect of globalization, doubts are widespread about the success of the economy in achieving greater equity and acceptable levels of poverty reduction. In the next section we shall review, equally briefly, the major thrust of the literature speculating on the possible links between growth and equity. While this review of the theoretical literature is not to provide exact guidelines to our empirical investigation to follow, some of the ideas explored might help to illuminate or emphasize specific results in the chapters to follow.

Labor markets, poverty and inequality in the growth process

The theoretical discussion on the changes in the incidence of poverty and inequality in the growth process of agrarian economies from low levels of income has been a major topic in development economics. The impact of growth is delineated through the labor market, and any predictions about the impact on poverty and inequality must be based on some implicit or explicit view of the structure of labor markets and their functioning.

Homogeneous labor

The classical view on economic development and its impact on inequality is found in the Lewis model and its elaboration in the early work of Kuznets and of Ranis-Fei. All these models consider the growth process to be driven by a shift of labor from the 'traditional to the developing sector', (variously identified as 'rural and urban', 'subsistence and capitalist'; 'agricultural and industrial'). Ranis and Fei in their work on Taiwan formalized the three different elements in this story which together determine the dynamics of inequality over time. First, there is the 'reallocation effect' of labor moving out of agriculture to the secondary and tertiary sectors. This shift will tend to increase inequality if the distribution is less equal in the latter. Second, we have the 'functional distribution effect' in the 'commercialized' sector (the income accruing to the agricultural sector is best treated as mixed income comprising returns to both labor and capital in family farms). An increase in the share of wages will typically increase equality. Before the 'turning point' in the labor market in the Lewis sense, the unlimited supply of labor at constant wages should induce technological change in labor-using direction and should prevent any increase in technology in a capital-deepening way that shifts the functional distribution towards capital. But after the 'commercialization point' in the labor-surplus economy the trend in the functional distribution of income depends on the nature of technical progress. Thus we have the third element in the dynamics: the 'innovation-intensity effect' that might be sufficiently biased against labor to decrease the share of labor over time and tend to increase inequality. The course of inequality through time depends on the relative strength of all these three effects. In fact the 'innovation-intensity effect' depressing the share of wages in the commercialized sector might not be delayed till after the 'turning point' as suggested by Ranis and Fei, but might already be working in a labor-saving way if entrepreneurs in the commercialized sector choose to adopt imported capital-using technology for a variety of reasons.

The Kuznets hypothesis of the U-shaped pattern of inequality dynamics follows from a theory embodying the above three elements. In the early adages of development the reallocation effect is strong and the share of labor also might, fall inducing a rising inequality. The trend is reversed when the reallocation effect weakens and the rise in wages overwhelms any effect coming from continuing bias towards capital-using technological progress.

Since the reallocation effect shifts labor from the low-income sector, the prediction is that in the early stages of the Kuznets process poverty should decline but inequality has an upward trend. We draw particular attention to this prediction because this is the trend we see in a number of developing countries in recent development history–including China and India.

Labor with different skill-levels

The basic model: skilled and unskilled labor

The economic literature in the last decade has paid a great deal of attention to the reason for rising wage inequality in a number of developed countries (including the USA). A feature of the increase in inequality has been the sharp rise of wages of 'skilled' workers relative to the 'unskilled'. The reversal of the trend towards narrowing wage differentials, which had been going on for much of the twentieth century, coincided with the increase in the share of manufactured exports going from the developing countries to USA and Japan in particular. It was natural for a number of US economists to jump to the conclusion that the critical role in this phenomenon was being played by the increase in imports of labor-intensive goods. A well-know theorem within the framework of the Heckscher–Ohlin model states that when trade is opened up, since each country tends to export the commodity using the more abundant factor, the relative price of the more abundant factor in each country will tend to increase. In developed countries, which shift to more skill-intensive products, the relative price of skilled labor would increase. The corollary of this proposition is that in developing countries the opposite will happen–the wages of unskilled labor would increase relative to those of the skilled, and wage inequality could be expected to fall.

The last corollary is less persuasive than would appear at first sight. The assumption of a homogeneous labor market, which transmits the impulses originating in the tradable manufacturing sector throughout the economy-wide labor market, even if partially acceptable in a developed economy, is wide of the mark in developing countries. The manufactured exports from developing countries originate in the formal sector. The extent to which this sub-sector is linked to the informal manufacturing sector varies from country to country, and in any case it is a small part of the labor market properly considered, which is dominated by the self-employed. We will come back to this point later on. But for the moment let us confine ourselves to some additions to the model as it is applied to the wage-labor market.

Extensions of the model: technological progress

THE DEVELOPED-COUNTRY SCENARIO

The static Heckscher–Ohlin theory is formulated in static terms with trade impacting the economy with an unchanged production function. But the recent decades have seen not only a great deal of technological progress, but of such progress being biased towards skilled labor.

Apart from considerations based on general observations, economists have noted that all explanations of rising wage inequality in the 'North' 'leave unexplained the rising skill intensity in non-traded goods as traded goods sector. In spite of having to pay more for skilled workers, employers in almost all sectors (traded as well as non-traded goods) chose to hire more skilled workers' (Gottschalk and Smeeding 1997, p. 649). 'Only technological change is consistent with rising skill intensity in the face of rising skill prices' (ibid., p. 650).

Daron Acemoglu (2002) has recently surveyed the vast literature on the technical progress and rising wage inequality in the North. The broad facts are clear. While the past 60 years have seen a vast increase in the supply of more educated and skilled workers, the returns to education in the U.S. among other countries fell during the seventies, but have begun a steep rise during the 1980s. This stylized fact is consistent with either a slowing down of the rate of increase of the supply of more skilled workers with a constant pace of technological progress, or alternatively with an accelerated pace of skill-intensive technical progress since the 1980s. It might indeed be a combination of supply and demand side factors.

Acemoglu hypothesizes that technical change is to a very large extent induced by factor market conditions. In the nineteenth century when the North had a plentiful supply of unskilled labor, technical change (e.g., during the industrial revolution) was directed to saving the use of skilled labor. With the rapid growth of education, the supply of skilled labor took a jump in its rate of growth, and the inducement for technical progress shifted towards saving the use of unskilled labor. The skill premium could be held in check as long as the pace of technical progress did not exceed that of the growth of skilled labor. But it is to be expected that with the vast growth of educated labor, its rate of growth in the North had to slow down. In fact, it is possible that the large expansion of international trade might have been a factor in increasing the pace of skill-biased technical progress. As the basic trade model noted, expansion of exports of skill-intensive products from the North tend to increase the relative price of such goods, leading to a search for technology biased towards increasing the productivity in such industries.

THE DEVELOPING-COUNTRY SCENARIO

What have these developments in the North got to do with wage-inequality trends in the South–the developing countries? The availability of a large pool of unskilled labor could be expected to promote technological progress that will economize on the use of skilled labor, as in the nineteenth century North. But there are several important factors which might suggest why this has not happened as a widespread phenomenon.

First, and foremost, it is clear from recent economic history that R&D expenditure is heavily concentrated in the North, and it seems to have the highest payoff in the advanced economies. Thus a more plausible scenario is that, rather than each country and region developing its own technology, new technology is developed in the leading economies of the North and spreads across countries.

Second, we need to emphasize that the techniques of production are not determined only by the relative supplies of different types of labor, but also by the quality of product which the market accepts. It has been noted in the literature that techniques which make use of less capital and less skilled labor often produce a final product which has attributes catering to the demand of low income consumers (Little et al. 1987, chapter 13). When a developing country enters export markets in a big way, the final consumer is located in the affluent North, and the technique of production has to be geared to producing items with superior attributes in the quality spectrum. Often these superior quality ranges would need more mechanized techniques with more skill-intensive labor. The point is reinforced by the need to achieve timeliness and homogeneous quality in the batches exported.

Third, it is useful to think in terms of different stages of production for the market. The stage of physical production might indeed be allocated to dispersed units using techniques which make use of labor of the type that is in plentiful supply (of low or traditional skills), but these have to be integrated with organizational, financial and marketing units to be able to supply the export market effectively. The garment industry, which has played such an important role in the export expansion from the South in recent decades, is a case in point. The tertiary activities needed in this export activity often use labor of high or non-traditional skills, which might be in short supply.

Fourth, the last point brings into focus an important part of the story of export expansion from the South. This is the role of outsourcing. Feenstra and Hanson (1999) among others have emphasized that the change in the degree of inequality or the relative wage of unskilled to skilled labor should be analyzed in terms of a foreign outsourcing model, which emphasizes trade in intermediate products, and not exclusively in terms of trade in final products which the H–O model stresses. The production of a final manufactured good can be broken down into several stages which can be arranged in ascending order of the skill intensity of the activity. Outsourcing from the developed country means that the some of the lower ranges of skill intensity in this chain are shifted out to developing countries. But these activities which are shifted, although they are of relatively low skill intensity in the North, are relatively in the higher rung of skill intensity in the South. The net effect of the outsourcing is then to reduce the relative demand for less skilled workers in the North, but to increase the demand for more skilled workers in the South. Thus while we can expect the skill premium in the North to increase, wage inequality in the South would also tend to increase, contrary to the predictions of the simple H–O model. This kind of outsourcing effect will, of course, be particularly important when the increase in manufactured exports from the South is being driven by direct investment by Northern businesses.

Extensions of the model: several grades of skill

Another dose of realism could be added to the basic H–O model by extending the model to accommodate more than just two types of labor–skilled and unskilled, with the former being complementary to capital in twentieth century technology. Adrian Wood (1994) makes a distinction between at least three types of labor: labor without any modern industrial skill–'raw labor' which is found in agriculture or low services, but not adapted to work in modern factories or businesses; labor with some basic skills for factory work; and labor with higher skills to perform more complicated tasks in the modern sector. Wood believes education is the basis of this classification–he calls the first category NO-ED, the second BAS-ED (those with at least primary or low-secondary education) and SKILD (with higher levels of education) the third category of labor. But the distinction need not be defined by levels by schooling alone. It is known that a significant wage gap exists in favor of labor of low skill in the 'modern' sector even in the face of plentiful supply of labor in the traditional sector (see the section below on 'segmented labor markets'). Wage inequality within the large-scale industrial sector might be squeezed but the over-all wage inequality increases because of the wage of BAS-ED labor increasing relative to that of NO-ED labor.

Even this limited prediction might be thwarted, if technological progress is skill-biased as discussed above, or alternatively, if we introduce factors of production other than labor and capital. Another strand in Wood's set of hypotheses is that factors of production in addition to labor and capital are critical in the comparative advantage of an economy–most notably land and the availability of natural resources. Countries with relatively large endowments of natural resources will tend to export more land-intensive products, while those with a shortage of such resources will tilt towards more manufactured activities. But the land-intensive primary products lead into processing industries which use less skilled labor than other industrial products. Thus expansion of industrial exports in land-abundant countries ceteris paribus would tend to dampen wage inequality, and to increase it in resource-poor economies. This is, of course, only the demand side of the story. The final outcome depends on the relative supply of educated or skilled labor over time–which is to large extent the result of autonomous state policies.

Extensions of the model: shifting boundary of the non-traded sector

In the original discussions of the H–O model there was an implicit assumption that the boundary between the traded and non-traded sectors coincided with that between the manufacturing and the tertiary sector. (For some theorists the implicit assumption was extended further to the distinction coinciding with that between the formal (modern) and informal (traditional) sectors.) Recent developments in the world economy have made this distinction quite unrealistic. For one thing, the services sector has emerged as a major exporter. Second, some products of the non-traded service sector are in close relationship to the traded sector.

Liberalization of the external sector, including devaluation which might accompany it, increases the relative price of traded goods and pushes more resources into the traded sector. But two other effects need to be considered. The first is that some non-traded goods might be complementary to the export sector. Such for instance might be infrastructure, including transport and some supporting services. An increase in the developing countries' exports, even if they are more low-skill intensive than the exports from developed countries, induces complementary expansion of infrastructure which is more skill intensive. Thus the net impact on the demand for labor of different skill levels is uncertain. Second, we should allow for substitution on the consumption side. Consider a developing country with abundant supply of unskilled labor, in which low skill services are close substitutes for the more skill-intensive traded goods (e.g., washing machines). Liberalization reduces the relative price of the latter, leading to a lower demand for low-grade services, and hence a lower demand for low-skill labor, which might offset the increase in demand for such labor induced by the expansion of labor-intensive exports.

The upshot of this discussion is that when the basic trade model is extended by successive doses of realism no definitive prediction about the movements of relative prices of skill, and hence the direction of change in the degree of wage inequality, is possible. This is not to say that empirical analysis would not yield patterns which are uniform over sets of countries or regions. Some work which has been done already has contained the intriguing suggestion that greater openness has decreased wage inequality in East Asian experience in the seventies in the expected way of the simple model, but that in several Latin American countries the opposite has been the case in the eighties. Commenting on this possible generalization, Wood (1997, p. 47) offers a hypothesis apparently based on a suggestion by Jeffrey Sachs:

It might be the case that all manufactures were import substitutes in Latin America, but only skill-intensive manufactures were import substitutes in East Asia. In that case, non-traded sectors (of a given skill intensity) might be more skill-intensive than import-competing sectors in Latin America and less skill-intensive than import competing sectors in East Asia. Hence, if greater openness (through substitution in consumption) caused non non-traded as well as export sectors to expand (and import-competing sectors to contract), the net effect might be to increase the relative demand for skilled labor in Latin America, but to decrease it in East Asia.

Education policies and the supply of skilled labor

While the evolution of the demand for skilled labor is important, countries differ enormously in the way the formal educational system develops over time. Even if skill formation is heavily influenced by on-the-job training, basic formal education is a critical variable. The impact of education on wage inequality has two different effects. The growth of educated population has a 'compositional effect' which yields an inverted U-shaped pattern à La Kuznets. Until a certain proportion of the population belongs to the more educated (and higher wage) groups, an increase in the proportion of the latter will increase inequality, but after the critical point is passed inequality falls as a larger proportion already belongs to the high wage group. The rising inequality in the earlier part of this process will be moderated if the rate of return to education does not increase, and a fall in the returns of sufficient magnitude will in fact reduce inequality. The latter possibility in fact turns on the supply of educated labor running ahead of the inversing demand as the modern sector develops. It has been noted in the literature that the decrease in wage inequality in Korea and Taiwan during their process of export-led industrialization could be traced in large measures to the prior investment in secondary education (see, for example, Gindling and Sun on Taiwan, and Fields and Yoo on Korea). The experience in these East Asian economies contrasts strongly with the development in Thailand, where post-primary education was neglected till the nineties. Thus over the period 1976–1988, Thailand had a strong upward trend in the inequality index as the export-led boom of the latter eighties put strong pressure in the market for skilled labor (World Bank 1996).

Segmented labor markets

The discussion so far has concentrated on the wage-labor market, and for the most part on labor markets differentiated by levels of measurable skills (e.g., education). But in developing countries much of the labor force is self-employed. Even within the wage-labor market discussions in the mainstream literature are generally concentrated on the formal part of the market–which typically excludes the small and micro sectors, if only because the informal sector is poorly served by regular statistical surveys on wages. If labor markets were reasonably homogeneous, trends in the formal wage-labor market would indicate trends in other parts of the labor market as well. But typically labor markets in such economies are segmented. Labor with the same measurable human capital earns significantly different incomes in different segments of the market. The trends in earnings might also diverge as between the different segments of the market.¹

A schematic picture of the labor market in a developing country looks like Table 1.4 (A and B refer to the formal and the informal sectors respectively). There are large gaps in the levels of earnings between the segments of the labor market shown in the table. These gaps persist even after we have controlled for measurable human capital differences between the workers found in the different sectors. The extent of the earnings differences would of course vary from country to country, and one of the tasks of country studies would be to quantify the more important of the wage gaps.

The impact of liberalization or other aspects of globalization on the over-all distribution of income would depend on:

First, the observed levels of employment and earnings in each segment of the stylized classification given above are the product of the intersection of demand and supply curves of labor in that segment. Thus one should be aware that factors affecting the derived demand for labor, as well as the supply conditions of labor, would be instrumental in affecting the outcomes in the segment.

Second, it should be apparent that the movements of these variables over time would be influenced, not only by the way markets for labor of different skills behave over time, but also by the working of the markets of co-operant factors, particularly land and capital. Thus the earnings of the self-employed will be more equally distributed if they are able to accumulate capital more easily over time. A more restrictive capital market would on the other hand both depress their mean earnings relative to those in the formal sector, and also perhaps lead to a more unequal distribution in this sector. For those in the farming sector, the distribution of land and of the basic inputs like fertilizer and water are of crucial significance.

Accounting for the earnings difference between the formal and the informal sectors

The formal–informal sector divide in the labor market cuts through the entire range of non-agricultural industries–both in the tertiary and secondary sectors, and in some economies even in agriculture where there is a large concentration of large Farms or plantations. In view of the importance of this phenomenon it might be useful to review the various hypotheses accounting for the gap in the levels of earnings between the two. In general these hypotheses are not mutually exclusive. They might co-exist in different degrees in any particular economy.

The institutional hypothesis

An important strand in the literature has asserted that labor in the formal sector is 'protected' in the sense that its wage level cannot be undercut by competition from outside labor. This type of 'protection' might be supported by institutions like labor laws or trade unions working independently or working hand-in-hand with the state's labor-regulatory framework. In this case, in so far as wage levels are significantly above alternative earnings outside, entry into the sector is rationed. There is an elastic supply of job seekers but only a fraction can be admitted.

The wage–efficiency hypothesis

The literature has recognized for sometime that the wage–efficiency relationship sets a floor to the wage rate in the formal sector. The most straightforward version is the nutritional one. Efficiency increases with the level of wages because better-fed workers are able to work harder. Thus no employer with a stable body of workers will offer a wage below a level at which efficiency decreases proportionately more than wages. Such a floor to wages is undermined in the informal sector because of a number of factors which include: (i) casual labor without attachment to specific employers; and (ii) self-employment working from households in which earnings of different working members are pooled together. Further, if we do not interpret the wage–efficiency relationship strictly in nutritional terms it will vary with the type of work, quality of machinery used and of goods produced, and the organization of labor. In fact it may become a hazy notion depending very much on the perception of employers. Large formal-sector employers, wary of possible labor unrest or adverse laws protecting job security, might opt for a labor system in which a small elite body of workers produces at a high rate of efficiency in exchange for stable employment at high wages. In any event a significant difference in wage per man is established between the formal and the informal sector although the difference in wage cost per efficiency unit of labor might not be that large. The extent of the differential clearly depends on the quality of labor supplied in the market as well as the institutional setting of the formal labor market.

How does this set of factors affecting the wage differential with respect to the informal sector relate to the discussion above about the skill/education related differentials? The mechanism discussed here in fact establishes a higher wage in the formal sector for all levels of education/skill. This does not, however, mean that formal-sector employers would not use education as a screening or signaling device for the selection of their workforce. If this happens we might find that the educational distribution of the workforce in the formal sector is much more skewed to the higher groups than in the informal. In addition to the wage–efficiency mechanism it would be a supplementary influence enhancing the premium enjoyed by skilled workers in the formal sector.

Constraints in the supply of co-operant factors in the informal sector

As indicated the self-employed constitute a major component of the workforce in the informal sector. Their earnings are in the nature of mixed income, consisting of returns to labor as well as to the co-operant factors used, principally capital (perhaps more of working than fixed capital). It is well known that credit constraints are more severe for the players in the informal sector. Thus the earnings profile in this sector would be critically influenced by the supply function of capital in this sector. It would affect not only the level but the shape of the earnings distribution in the informal sector. The differential with respect to formal-sector earnings is likely to be different for different parts of the distribution–and would also vary from country to country depending on the severity of the credit constraint in this sector.

Regional differences in earnings and employment

In recent work on post-reform developments spatial differences have come to the forefront of discussions. In fact the basic problem of uneven economic growth spatially is in some sense the heart of the subject of development economics. Post-reform developments in the rapidly growing economies like China have drawn renewed attention to the problem. Globalization is heavily directed in the first place to limited areas where producer links to external markets can be most advantageously established and entrepreneurs can exploit important external economies of scale. In fact the concern about unequal development exists equally in a relatively closed economy where major innovations might favor some regions more than others in a cumulative way–as might have happened in the spread of the green revolution in Indian agriculture.

It would be wrong to assume that processes of economic growth necessarily worsen inter-regional inequality. Even if the growth process is concentrated in some regions or enclaves to begin with, rising costs and links to other internal markets can and do produce incentives for producers to diversify to other areas. The net outcome in the dynamic process depends on the relative strength of these centrifugal forces and the process of cumulative causation favoring increased localization.

Plan of the work

The brief review of some of the more important theoretical ideas in the last section provides a background to the empirical investigation of the impact of the reform process and globalization on labor markets, poverty and inequality in India. The major data sources of the data utilized are the 'Thick' rounds of the National Sample Survey (NSS) which are conducted every five years. The latest round available for analysis is the 55th round for 1999–2000. At the time of the completion of this work the subsequent 'thick' round for 2004–2005 is not available for analysis with the unit-level records. But we are able to indicate some broad trends form the published reports issued by the NSS on a limited set of tabulations. This is done in the Epilogue. For the most part the bulk of the work relates to the period ending at the close of the last century.

The book is in four parts. Part I discusses the broad trends for the economy as a whole. Part II focuses on differences between major states and 'broad regions' of the country. Part III carries the analysis to the major sectors–agriculture, manufacturing and the tertiary industries. Part IV discusses issues in labor institutions–both in the formal and the informal sectors.

The analysis for All-India begins with the changes in the incidence of poverty–contrasting in particular the trends in the pre-reform years between the two thick rounds of the NSS before the 55th, and the post-reform period between 1993/1994 and 1999–2000. We discuss, within the framework of a decomposition model, the relative importance of rural–urban shift of labor, growth of the two sectors of the economy and changes in inequality within each. The varying experiences of the urban areas of different population-size groups as well as of the major states in the process are also discussed.

Chapter 3 presents the basic trends in employment and earnings over the period covered, contrasting the post-reform years with the previous periods. It goes on to document the emerging trends in inequality, both for wage earners and all wage- and non-wage-earning households together. We also document the increase in 'rural–urban dualism' in the post-reform years, which had already been suggested by the poverty analysis of Chapter 2.

Part II turns to the analysis of regional differences. This topic is of vital interest in a large country like India. We can do this regional analysis in two ways. First, we can look at differences between major states. This is important because states are political units with a good deal of autonomy in the implementation of economic policies. They are, however, not homogeneous areas in terms of agroecological areas. The latter are of great importance in Indian conditions as they have significant impact on productivity and incomes in the large agricultural sector in particular. From this point of view, working with NSS regions which are more homogeneous in character would seem to be more pertinent. We pursue the analysis at both levels. Chapter 5 is a contribution to the analysis of inter-state differences in labor-market outcomes. In the following chapter we work with NSS regions and look at differences in rural poverty in particular. Since the number of NSS regions is large, an attempt is to group them into seven 'broad regions' defined in terms of agro-economic conditions and geographical contiguity. This attempt at the analysis of rural poverty in terms of grouped NSS regions is, we believe, the first such attempt at regional analysis, and will undoubtedly be improved upon by other researchers.

Part III of the book shifts attention to individual industrial sectors of the economy. The problem of labor absorption at reasonable levels of earnings in agriculture is discussed in Chapter 7. Two important questions related to the performance of the agricultural sector are also addressed. The first is the relationship of agricultural productivity to off-farm activities; and the second is the trends in household welfare of different classes of farmers, especially in the post-reform period. The unit-level data available from the NSS are analyzed to throw light on these major issues. The last three chapters in Part III are on the performance in the non-agriculture sector. Chapter 8 is a detailed analysis of the low elasticity of employment with respect to output in the formal manufacturing sector. The fact that the (relatively) high productivity formal sector has been able to create only a low rate of employment growth, in spite of the fairly high rate of output growth, has meant that labor going into the manufacturing sector has been largely absorbed in the low-productivity informal sector. Chapter 9 in fact shows that the developments in Indian manufacturing have been somewhat more complicated than that. Apart from the labor absorption in the truly informal sector–consisting of household enterprises employing none or only a few hired workers, employment has been disproportionately concentrated at the small end of the formal sector, in units employing 6–9 workers (the so-called DME sector). The bi-modal distribution of employment in Indian manufacturing (with concentration at the smallest and the highest size-groups) has given rise to the problem of dualism in Indian manufacturing. This issue is discussed in detail in Chapter 9–where we analyze the adverse impact of the phenomenon of the 'missing middle' on healthy manufacturing growth, and also the causes leading to its origin and persistence. Chapter 10 takes on the tertiary sector which has been the major source of labor absorption from agriculture.

In the last part of the book–Part IV–labor-market institutions are studied. A critical evaluation of labor laws affecting the formal sector is followed by a review of on-going initiatives to tackle the difficult question of state intervention to improve the conditions of the large numbers of workers eking out a living in India's informal sector.

Document(s) 1 of 16