Organized labour and economic liberalization.
India: Past, present and future

Executive Summary

This paper examines the role of organized labour in India in a structural and historical context. It attempts to trace the economic, political and social effects of the trade union movement and its strategies over time. These effects are felt at enterprise- and/or firm-level, industry-level, regional and national level. First we consider the effect of changing economic conditions on the evolution of trade unions and bargaining institutions in largely urban labour markets in the post-independence period (1947 onwards). Some contemporary issues affecting the organized labour movement in India today are then discussed.

This essay has two main objectives: (a) to present a history of Indian industrial relations, broadly understood as the changing relationships between workers, trade unions, employers, the economy and the state; (b) to posit a political economy of trade unionism in India.

The evolution of trade unionism (and industrial relations in general) is described in terms of the "four phases of unionism". This corresponds with structural changes in the economy and we trace the effect of these changes on labour markets as well as on the (very broadly defined) industrial relations arena. The first phase (1950 to mid-1960s) corresponds to an era of state planning and import substitution, when public-sector employment and public-sector unionism rose phenomenally. Unions and bargaining structures were highly centralized; the two main federations were the nationalist Indian National Trade Union Congress and the communist All India Trade Union Congress. State intervention in the determination of wages and working conditions was the norm and "state-dominated pluralism" was the labour regime during this first phase.

The second phase (mid-1960s to 1979) is associated with a period of economic stagnation and political turmoil. Employment slowed down, there were massive inter-union rivalries, and industrial conflict increased. Centralized bargaining institutions now started feeling the pressure of dissent from below, and both the Hind Mazdoor Sabha and the Centre of Indian Trade Unions made significant progress in the labour movement. The crisis culminated in the May 1974 railway strike that was followed by the 1975-77 Emergency Regime of Mrs. Gandhi. An "involuted" pluralism dominated Indian labour relations during this second phase.

The third phase (1980-1991) corresponds to a period of segmented and uneven economic development. Decentralized bargaining and independent trade unionism enter the stage in a significant way. Two major strikes (the 1980/81 Bangalore public-sector strike and the 1982 Mumbai textile workers strike) marked this phase, and inter-state and inter-regional variations in the nature of labour-management regimes became much wider. In the more profitable economic sectors the unions gained, but in the unorganized and declining sector, workers lost out and unions were left with few strategies.

Finally, the fourth phase of unionism (as yet incomplete, 1991-2000) represents the post-economic reform period. The stabilization and structural adjustment programmes led to demands for increased labour market flexibility, especially employment flexibility. This has led to a recruitment freeze in many public sector sites, and unions in these sectors now have to cope with competition at local level. In non-viable public enterprises, unions are coming to terms with "voluntary" retirement schemes. In the early years of economic reform there were sincere attempts by all parties to engage in tripartite consultations, but there now seem to be several barriers to this form of engagement.

The essay then gives a more detailed presentation of crucial issues which are particularly relevant to the trade union movement in India today. These are: the relationship between unionization and employment; union structure and union density; wages and working conditions; the changing nature of collective bargaining; changing patterns of industrial conflict; the nature of labour-management relations; the (extremely important) issue of inter-state and regional variations in labour regimes; the question of women workers and unionization; and the changing public perception of trade unions. The main findings are summarized below.

• In both the public and private sectors, employment in industry has substantially declined, especially since the economic reforms; however, employment in public sector services has increased since the third phase of unionism. The trend in most industries is to reduce permanent employment and to use more contract, temporary and casual workers.
• Trade unions have performed well in profitable industries, mainly in private (often multinational) enterprises, but also in some public sector firms with decentralized bargaining structures. However, they lack new strategies in the older and declining sectors of production where industry-wide bargaining structures are typically the norm. Whereas workers and unions in non-viable public enterprises are facing closure and an uncertain future, centralized unions representing employees in public services remain strong due to the monopolistic nature of their product markets.
• The number of plant-based independent and unaffiliated trade unions has risen, which may have caused a decline in the power of centralized affiliated unions, especially in the private sector.
• Two critical changes in collective bargaining are the expansion of coverage and scope of long-term agreements. Competitive market pressures, especially since the beginning of economic reform, have forced bargaining outcomes to be decided at local level.
• The procedural environment required for competitive industrial pluralism to work at its best is still absent from the Indian industrial relations system (absence of a secret ballot, absence of a single bargaining agent, absence of third-party arbitration and strong restraints on "legal" strikes). This makes the organized labour movement still very dependent on state recognition and patronage. All parties recognize the urgent need for comprehensive labour law reform.
• Modern human resources management has significantly altered labour-management relations in the advanced sectors of production without necessarily leading to a decline in union bargaining power.
• Inter-state and inter-city variations in labour-management relations have increased since the mid-1980s, and in the absence of a concerted effort on the part of the central government to reform labour laws, these variations are likely to be accentuated with the progress of economic reform.
• Women workers are largely concentrated in the self-employed and unorganized sectors with little representation in trade unions. Non-governmental organizations have made considerable efforts to organize women in these vulnerable occupations.

The paper concludes with a discussion of the future role of trade unions, concentrating on the unionization of informal sector workers. Given the low levels of unionization in India (relative to comparable countries), and given that a large proportion of workers are employed in the informal sector (a trend which is likely to intensify with the economic reforms currently underway), the trade union movement has to make a determined effort to organize informal sector workers. This calls for concerted action on the part of the centralized trade union federations in alliance with other social movements and non-governmental organizations. It is only through this kind of extensive mobilization that "public action" can be triggered to "level-up" the labour market institutions of vulnerable sections of the workforce.

The organized labour movement will have to come to terms with global competition, technology, new industrial organization and structural/demographic changes in the workforce. The earlier, relatively insulated, systems for regulating employment will have to give way to more market-sensitive and flexible systems, ultimately even in the public sphere. Unions, especially in the service sector, will have to become sensitive to consumer needs. In the private sector, trade unions have adapted their structures in response to the decentralization of industrial relations, but these adjustments are painful and difficult in the public sector. There is no other way but to increase membership, which means organizing vulnerable workers in the informal sector. This is the most obvious way for unions to win broader community support. Finally, trade unions can only gain by cooperating and working together. Perhaps the new millennium will witness a series of mergers leading to a united and independent labour movement.

1. Introduction

This paper examines the role of organized labour in India in a structural and historical context, tracing the economic, political, and social effects of the trade union movement over time. We look at union strategies at the level of the enterprise and/or firm, the industry, the region and the nation. The effect of changing economic conditions on the evolution of trade unions and bargaining institutions are described in largely urban labour markets in the post-independence period (1947 onwards). Then we consider some contemporary issues affecting the organized labour movement in India today.

The paper has two main objectives: a) to present a history of Indian industrial relations, broadly understood as the changing relationships between workers, trade unions, employers, the economy and the state; and b) to posit a political economy of trade unionism in India. The contrasting views of Freeman and Medoff (1984) of trade unions as "monopoly" institutions, or as the "collective voice" of workers serves as a theoretical framework here. From a policy perspective in democratic and pluralistic societies, the objective must be to minimize the "monopoly" effects and to strengthen the "collective voice" effects.

The position taken here is "that the evolution of labour institutions is determined by the objective interests of social groups inherent in the logic of a modern industrial society" (Zeitlin, 1987, p. 163), and that these "interests" are inevitably "ambiguous and context dependent". Consequently, it is difficult to establish mechanisms to ensure that labour institutions, such as trade unions, collective bargaining and the state, continue to play the parts assigned to them. The second proposition of this essay is that "no single approach to the study of labour organization is at present adequate - which is to say that the study of these matters is usually informed from several points of view" (Williamson, 1985, p. 241).

The organization of the paper is as follows. Section 2 provides a comparative framework by examining some proposals regarding union behaviour, economic performance and industrial/labour relations strategy. Section 3, on the changing economic environment and its effects on organized labour in India, consists of two parts. The first part expounds the "four phases of unionism". The second part deals with unionization and employment, union structure and density, wages and working conditions, collective bargaining, industrial conflict, labour-management relations, inter-state variations, women workers and unionization, and the changing public perception of trade unions. The concluding section comments on the future role of trade unions in India, with a focus on organizing non-unionized workers.

2. Contrasting views on trade unions as institutions

Freeman and Medoff (1984) examined the impact of unionism on the employment relationship in the US economy, contrasting the "two faces of unionism": the "monopoly" versus the "collective voice" view. These lead to a completely divergent analysis of the union as an economic and political institution; the policy implications of the two views are therefore fundamentally incompatible.

The first view analyses unions as large monopolies in the labour market whose basic objective is to increase their members' wages above the market level by restricting labour supply. This "wage mark-up" leads to a misallocation of human and capital resources, and is therefore not only economically suboptimal, but may also be socially undesirable in that it leads to greater inequalities within the workforce. In addition, strikes of all kinds impose an easily measurable loss on the economy. Finally, "union work rules", embedded in contract provisions often lower the productivity of both labour and capital by creating inflexibilities at the workplace. While these arguments typically emanate from free-market economists, social theorists who subscribe to this view often characterize unions as undemocratic and crime-riddled institutions.

In sharp contrast, the collective voice view asserts that unions have positive economic, political and social effects in pluralist democracies. The union's collective voice, determined by a "median voter", provides management with information on workplace and shopfloor issues, acting as a communication channel. This leads to the development and retention of specific skills, improves worker morale, provides conditions that eliminate quitting, and enables the union to pressure management to act fairly and efficiently in its daily operations. It is believed that unionized work environments are more productive than comparable non-unionized environments. Finally, unions are socially beneficial as they represent the interests of lower income groups and vulnerable sections of society.

Using the above framework, Freeman and Medoff examine union effects on economic efficiency and the distribution of income. They also examine the social nature of the union as an institution of modern industrial society. Questions about economic efficiency usually relate to whether unions enhance productivity through their collective voice or whether they block technological change through "monopoly" routes, such as inflexible contract provisions. These often compel employers to substitute capital for labour and to invest in better "quality" labour. This leads to productivity growth only in the sectors concerned, at the cost of overall employment growth in the economy. Questions relating to income distribution have to do with the effects of unions on non-unionized and unorganized workers. That is, do unions have positive or negative "spillover" effects on unorganized labour markets? Generally speaking, the higher the degree of unionization in an economy, the greater are the positive spillover effects and the smaller the wage inequalities within the workforce. Finally, the "monopoly" view holds that unions fight for their own selfish interests at the expense of others, and blames their political lobbying for bureaucratic and corrupt practices. The "collective voice" view sees unions as the economic and political institutions of the weaker segments of society.

It is clear that the political charter of the "collective voice" is based on pluralistic democratic principles, enlightened management and responsible trade unionism (Turnbull, 1991). The crucial assumptions of the "collective voice" are the following: (a) political and industrial pluralism; (b) interest group politics being played out in a multi-party regime of political liberalism; (c) a strong belief in the efficiency and fairness of "regulated conflict"; (d) an efficient and fair procedural environment, which implies the just implementation of unambiguous labour laws and a degree of symmetry among adversaries' bargaining resources; and, finally (e) trust in neutral third-party arbitration when confronted with a labour relations impasse. According to Williamson (1985, p. 241), "the institution of arbitration lies at the core of industrial pluralism". The basic premise of pluralist bargaining is that interest group conflict approximates the public interest. This paper explores the extent to which the assumptions and beliefs of pluralist industrial relations have evolved in India as structural changes have taken place and the political economy has developed.

A comparative framework

It is now a truism that economic performance is related to the nature of labour market institutions (see Horton et al., 1991; Nelson, 1991; Freeman, 1992; Buchele and Christiansen, 1992). Economic performance here means controlling inflation and generating employment, which implies productivity growth. In the world of unions and collective negotiations, the key variables frequently cited as determining economic performance are: the level at which bargaining takes place (plant, firm, industry/region, nation), and the nature of trade union structures.

Until at least the mid-1980s, the literature suggested that economies with a decentralized bargaining structure (enterprise-based unions negotiating at plant and firm level, as in East Asian countries, Japan and Switzerland) and economies with a centralized bargaining structure (national agreements with centralized trade union federations, as in Austria, Norway and Sweden) "performed" better than economies having industry-wide agreements with industry-wide unions, as in the US and UK. Put another way, "collective voice" effects seem to be maximized in centralized bargaining structures, whereas "monopoly" effects seem to be greatest at medium levels of centralization. This is the famous "Calmfors-Driffil" relationship, based on the premise that the extent of labour-management cooperation (or conflict) is determined by the degree of centralization of the bargaining institutions. Its relevance to developing countries was examined by Nelson (1991), the argument being that as centralized labour market institutions cannot easily be replicated in these countries, decentralized labour institutions may be the most efficient arrangement (see Horton et al., 1991). The whole debate on labour market flexibility and globalization is concerned with the degree of decentralization of labour market institutions (see for example, Katz, 1993).

While most commentators would agree that labour market flexibility is required for facilitating economic restructuring, heated controversy surrounds the question of how this flexibility is to be achieved. As transition will involve considerable social costs, what is to be the role of the state and the trade union movement in managing this transition? The debate acquired particular significance with the release of the World Bank World Development Report: Workers in an integrating world (1995) and an ILO report: World Employment 1996-97 - National policies in a global context. ^{(Endnote 1)} The World Bank Report hypothesizes increasing employment and labour incomes with greater integration of nations into the global market, positing a reduced role of the state in the labour market and expecting trade unions to be responsible and non-political. On the other hand, the central message of the ILO Report is that state intervention with centralized bargaining institutions can return economies to full employment policies, which it sees as the only way to create jobs and better earnings opportunities (D'Souza, 1998).

Where does India fit in this comparative framework? The Indian economy represents a mix of all three bargaining levels and a variety of union structures. In the private corporate sector, plant-level bargaining takes place with enterprise-based unions that may or may not be affiliated to parliamentary political parties. In public-sector enterprises, centralized trade union federations affiliated to political parties bargain with the state (as employer) at industry-and/or national-level. Central and state government employees in the service sector (transport, postal services, banking and insurance, police and firefighters, etc.) are usually represented by politically affiliated unions bargaining at national and/or regional level. Most of these centralized bargaining and union structures were stable during the period of planned industrialization while India pursued a policy of import substitution. However, since the mid-1980s the economy has opened up to greater domestic and international competition, and these structures have come under increasing pressure to decentralize. These pressures have become stronger since the economic liberalization programme was introduced in 1991, especially in the public sector. As in other countries, the state has gradually retreated from its earlier role of creating permanent employment and regulating union-management bargaining.

India has had mixed experience with creating labour institutions that are compatible with a pluralist industrial relations model. In the early years, the government officially promoted industrial pluralism and bilateral collective bargaining, but the institutions responsible for this were largely controlled by the state. This "state-dominated" pluralism, coupled with ambiguous labour laws regarding trade union recognition and "industrial disputes", eventually led to a multiplicity of party-based trade unions. This considerably weakened the political power of the organized labour movement as a whole, although in some strategic sites in the public sector centralized unions had considerable bargaining strength. Although unions could impose severe costs on key sectors of the economy, the organized labour movement as a whole was not strong enough to impose a cooperative solution at national level. This is still true today.

With economic liberalization, competitive forces began to affect the structure of the union movement. In several private enterprises, "independent" rank-and-file led unions came into existence and engaged in informed and militant bargaining, often with multinational employers, securing substantial wage and non-wage gains in the process. As these unions "traded off" increased wages against employment growth, and as employers shifted to "outsourcing" from non-union sites, the traditional party-based unions found their potential recruitment terrain both challenged and curtailed. More recently, since the liberalization process officially began in 1992, many of these centralized party-based unions have united under a common front to resist government attempts at privatization and decentralization in the public sector. However, the organized labour movement as a comprehensive organization continues to face a fractured and segmented constituency, divided by skill, region, industry and ethnicity. In addition, major labour law reform continues as an unfinished (and forever postponed) agenda.

In terms of the "monopoly" versus the "collective voice" framework, the early years after independence witnessed the state acting as the 'collective voice' of workers for the purpose of rapid industrialization with minimum industrial strife. In so doing, the state minimized the potential "monopoly" effects, wages and working conditions were administered rather than decided through collective negotiations in the public sector. An implicit "incomes policy" kept the "'union wage mark-up" in check. Over time, however, as both inter- and intra- industry differentiation developed, especially within the private sector, other, more radical and militant, union "voices" emerged that quite effectively began to challenge the state hold on the organized labour movement. In the private sector, efficient productivity bargaining with informed unions kept "monopoly" effects within the firm in check while "collective voice" effects increased. However, in public sector enterprises and services, the union "voice" led to rigid and inflexible contract provisions. With pay increases unrelated to improvements in productivity, union "monopoly" effects intensified. There seems to be some agreement now among labour commentators and researchers that (since at least the early 1980s) some segments of the organized workforce made substantial wage and non-wage gains in the more profitable sectors of production. The large majority of workers, however, continue to face increasing employment insecurity, if not lower wage growth, both in declining industries in the formal sector and in the growing informal sectors. With the onset of economic liberalization, these "negative union spillover" effects may accentuate inequalities within the workforce. Clearly there is need for a concerted attempt by the state and the organized labour movement to actively reverse these trends by levelling-up the labour market institutions which affect the wages and working conditions of the unorganized.

3. The changing economic environment and its effects on organized labour in India

3.1 The four phases of unionism: An evolutionary approach

3.1.1 The first phase of unionism (1950 to mid-1960s)

The first phase of the (post-independence) ^{(Endnote 2)} Indian trade union movement generally corresponds to the first three Five-Year Plans (1951-56, 1956-61, 1961-66), a period of "national capitalism". A state-led industrialization policy with an import substitution strategy resulted in the formation of large, employment-intensive public sector enterprises, mostly in the capital and intermediate goods sectors. This massive development of the public sector aided the private corporate sector in terms of supplying the necessary intermediate and capital goods (Desai, 1975). Between 1951 and 1965, industrial production increased at an average annual rate of 7.7 per cent, and manufacturing output increased at the rate of 7.6 per cent (Nayyar, 1981). High growth rates were sustained by public investment in capital and intermediate goods, while growth in consumer durables slowed down. Rigid import-substitution policies guaranteed the domestic market and stimulated private sector investment (Patnaik and Rao, 1977).

The expansion of large public enterprises led to employment growth in the formal economy, mainly in such enterprises, but also in the private sector, notably in services, transport and education. Average employment growth increased rapidly from around 0.4 per cent per annum from 1951 to 1956 to around 2 per cent per annum from 1961 to 1966 (Papola, 1994). Public sector employment quite naturally led to public sector unionism. The number of registered trade unions increased from 4,623 in 1951/52 to 11,614 in 1961/62; membership in the registered unions that submitted returns more than tripled during this period (Venkataratnam, 1996). ^{(Endnote 3)}

The communist-led All India Trade Union Congress (AITUC) had dominated the organized labour movement since before independence (for example, in the textile and engineering industries of Mumbai, Calcutta, Kanpur). The growing public sector now provided a new terrain for large- scale unionization. It is here that the Congress Party-controlled Indian National Trade Union Congress (INTUC) made early inroads. Unlike the AITUC, which rose from the rank and file, the INTUC was imposed on the labour movement from the outside. There were no ambiguities in the chain of command from party to union. As (the then) Prime Minister Nehru stated (Chatterjee, 1980, p. 152):

It goes without saying that the INTUC has been sponsored and nursed mostly by Congressmen and derives its strength from the moral and other support of the Congress…….…it is imperative that in all political matters all Congressmen working in the INTUC should treat the Congress as its supreme body and abide by its code of conduct.

Myron Weiner's (1962, p. 78) observation of the INTUC leadership was succinct: "Their loyalties are to the Congress Party, then to the present (Congress) government, to the nation and last of all to the workers who belong to their unions." When confronted with a choice between the patronage of the ruling party and genuine worker support, the INTUC usually opted for the former (Chatterjee, 1980). Since chief ministers, labour commissioners, registrars of trade unions, inspectors, conciliators, tribunal officials, magistrates, police officers and all other officials who dealt with unions (during this first phase) were appointed by the Congress government, management officials were often requested or pressured to assist INTUC unions to establish themselves or defeat rival unions (Kennedy, 1966).

This relationship between the state (i.e. ruling government) and its affiliated trade union federation during the first phase of unionism seems to have tied in neatly with the provisions of the Industrial Disputes Act (IDA) of 1947 which, according to Datta Chaudhuri (1996, p. 12), is "the single most important piece of legislation between the worker and his employer". The Act makes no provision for procedures to determine the representative union in what would normally be a single bargaining unit, and as employers were under no legal obligation to bargain with unions, there were no built-in incentives for either party to engage in collective bargaining. Early writings on Indian industrial relations consistently pointed to this aspect of the Industrial Disputes Act and the way in which it impeded collective bargaining during this period (Kennedy, 1966; Punekar, 1966). Coupled with this was the Indian Trade Union Act of 1926; the Act allowed any seven workers to register their trade union, but made no provision for union recognition (e.g. through a "secret ballot" procedure). While the opposition unions were for a "secret ballot" to determine union strength, the INTUC was against it, favouring instead the "check-off" system of membership receipts, a system that could easily be manipulated. ^{(Endnote 5)}

after the Labour Minister who resigned in protest, and many commentators see this as a major setback to the development of a mature industrial relations system (Ramaswamy, 1984). In more contemporary language, it may have been a strategic mistake: one can only imagine how the industrial relations system would have developed if the "Giri Approach" had been adopted.

State intervention in the determination of wages and working conditions was the norm during the first phase; wages were determined by political and institutional considerations (Myers, 1958; Fonseca, 1964; Jackson, 1972). The structure of bargaining was very centralized, usually at national level, but at industry level in some regions (e.g. Mumbai textiles). In a few cases private- sector bargaining was at enterprise level. Wage determination during this first phase was known at the time as "tripartism" and "political bargaining".

In terms of the movement of real wages of industrial workers during this first phase, India was held out as an example of the Lewis model of growth at work, with both product and consumption wage growing slower than labour productivity (Jackson, 1972). Low unionization, inter-union rivalries sharpened by political affiliation, excess supply of labour and state intervention of a complex and peculiar type contributed to a wage lag (Deshpande, 1992). As the famous Report of the National Commission on Labour (1969), examining the real wages of industrial workers for the period 1951-64, stated: "increases in money wages of industrial workers since Independence have not been associated with a rise in real wages nor have real wage increases been commensurate with improvements in productivity". The labour relations regime promoted "responsible unionism" subject to maintaining industrial peace (Johri, 1967). Both the number of strikes and the number of workers involved in strikes during the first phase were significantly lower than during the second phase and beyond (Johri, 1967; Sengupta, 1992). By the end of the first phase of unionism, further splits had occurred in the labour movement: the Socialists broke away from the Congress and formed their own trade union federation, the Hind Mazdoor Sabha (HMS), and during the Indo-Chinese conflict, the radicals broke away from the Communist Party of India and formed the Communist Party of India (Marxist) which generated its own trade union, the Centre of Indian Trade Unions (CITU).

Commentary

The first phase of unionism represented a period of state-driven industrialization that possibly required the government to guide or control the labour movement. At enterprise level, the capital-labour relationship was dominated by a paternalistic labour relations system based on the belief that the state knew more about workers' needs than they did themselves. In this way, the state appropriated the various "union voices" for the "collective" purpose of rapid industrialization with minimum industrial strife. In public sector enterprises and services, internal labour markets ^{(Endnote 6)} generated social harmony and a feeling of belonging.

Given the large number of statutes governing the terms and general conditions of employment^{(Endnote 7)}

it is clear that the state attempted to execute the "idea of a national minimum". However, freedom to contract between capital and labour was restricted and mediated through the state. In addition, the idea of planned and rapid industrialization possibly excluded the basic premise of mature industrial pluralism, i.e. the belief that "regulated conflict" can be both productive to the economy and fair to the involved parties.

During this first phase of unionism, few strategic choices were open to unions or to employers as labour relations outcomes were over-determined by the state. The institutional structures that facilitate efficient collective bargaining remained underdeveloped. Government attempts at developing participatory choices at the small group level, through its legislated Work Committees (1947) and later through Joint Management Councils (1958), received at best a lukewarm response from management and unions (Venkataratnam, 1996).

During this first phase, union "voice" effects were probably larger than "monopoly" effects, as both unionization and employment in the organized sector increased significantly (see tables 1 and 2). In addition, given a phenomenal rise in the number of registered trade unions during this phase, one could argue that there were positive union spillover effects to previously unorganized sectors.

3.1.2 The second phase of unionism (mid-1960s to 1979)

The second phase of unionism more or less corresponds with the Annual Plans for 1967-69, the Fourth (1969-74) and Fifth (1974-79) Five-Year Plans. "The rate of inflation rose above the politically sensitive danger-mark of 10 per cent in 1966/67 and 1967/68, and food price inflation was even higher (around 20 per cent)" (Joshi and Little, 1994, p. 48). Inflation worsened in 1973/74 and there were food riots in various states. The period is associated with industrial stagnation (Nayyar, 1981). Thus, between 1965 and 1975, the average annual rate of growth in total industrial production and in manufacturing output increased at only 3.6 per cent and 3.1 per cent respectively (Nayyar, 1981). Political economists have posited various explanations for this slowdown: the deceleration in public investment (Raj, 1976; Vaidyanathan, 1977; Patnaik and Rao, 1977), the unequal terms of trade between agriculture and industry (Mitra, 1977), the inefficiencies of state regulation in the public sector (Bhagwati and Srinivasan, 1975), and changes in the structure of demand resulting from growing income inequalities (Nayyar, 1981).^{(Endnote 8)}

In addition, the economy suffered oil price shocks in 1973 and in 1978. During this phase, especially during the Fourth and Fifth Five-Year Plans, actual growth rates for industrial production were far below plan targets (Ahluwalia, 1991, p. 11). It is quite clear that the deceleration adversely affected the level of employment in the economy as a whole.

Structural changes in the economy obviously had an effect on union activity, collective bargaining practices, industrial labour markets, and labour relations in general. There was a dramatic increase in the number of disputes (strikes and lockouts), the number of workers involved and the number of workdays lost between 1966 and 1974 (Sengupta, 1992). New forms of protest, such as the "hartal" (go-slow), often resulting in considerable violence, emerged during this second phase. In certain regions, such as West Bengal, these were used quite frequently and effectively, leading to a significant outflow of capital to other parts of the country.

Disillusionment with the INTUC's internal practices and its ineffectiveness at enterprise level led to a proliferation of unions affiliated to other, more radical, political organizations during the first part of this second phase. Workers sought more skilled politicians and negotiators to lead their struggles. The relationship between the rank and file and outside leadership in the non-INTUC unions could be characterized, to use Holmstrom's (1984, p. 289) phrase, as "a transaction between the membership and a politician known as a good negotiator". Two demographic factors may be associated with this change in worker preferences: a) a growing proportion of workers were young people who had not participated in the pre-independence labour struggles; and b) (probably more important), the leaders of the radical unions at this time were mostly not party politicians but committed lawyers and student activists well-versed in the bureaucratic rites of the Indian industrial relations system (Chatterjee, 1980). An analysis of industrial disputes by "union-type" revealed that the number of disputes involving two or more central trade union federations increased between 1966 and 1973 compared to the first phase (Bhattacherjee, 1987a, p. 57), reflecting a period of intense inter-union rivalry. Uncertainties within the organized labour movement finally culminated in the all-India May 1974 railway workers' strike that shook the economy and the country.

During Mrs. Gandhi's Emergency Regime of 1975 to 1977 the right to strike was suspended, and this is clearly reflected in time-series strike data (Sengupta, 1992). In addition, the emergency regime pre-empted bargaining on key issues, froze wage increases, reduced the minimum annual bonus from 8.3 per cent to 4 per cent, and transferred increments in the cost-of-living allowance to a compulsory savings scheme (Rudolph and Rudolph, 1987). Two important interventions took place during this regime. The first was an attempt by the government to establish the National Apex Body, composed of 12 trade union federations and 11 employer representatives, in order to encourage a bipartite approach to industrial relations (as opposed to the earlier tripartite approach). On the surface this seemed more responsive to trade union preferences for voluntary collective bargaining, but this was not the case.

The second intervention was the 1976 amendment to the Industrial Disputes Act, that arose either from trade union pressure (unlikely during the Emergency!) or as a populist measure, which led to employment inflexibility: firms employing more than 300 workers had to get (state) government permission before they could lay off workers. Needless to say, government permission was seldom forthcoming. However, this apparently pro-union measure had an unexpected (at the time) effect: for the first time since independence, the number of workdays lost to lockouts exceeded the number of days lost to strikes (Sengupta, 1992).

During the post-Emergency regime of the Janata government (1977-80),there was a qualitative change in the political climate but not much rethinking in the industrial relations structure (Sengupta, 1992). A government attempt to put forward an industrial relations bill banning strikes and lockouts in essential industries and services met stiff opposition from most trade unions. The bill was not passed (Sengupta, 1992). In 1977, the major central trade union organizations in decreasing order of membership were the Indian National Trade Union Congress, the All India Trade Union Congress, the Hind Mazdoor Sabha, and the Centre of Indian Trade Unions (Bhattacherjee, 1987a).

The second phase of unionism saw significant changes in collective bargaining practices,. The Industrial Disputes Act of 1947 did not provide for the compulsory recognition of a representative union as the sole bargaining agent, nor did it encourage or compel parties to bargain in good faith; more importantly, it gave no legal status to collective bargaining agreements. However, the 1965 amendment to the IDA gave a "higher legal footing" to agreements reached through conciliation and adjudication. Patil (1982) describes how employers and unions have used the 1965 amendment to transform agreements into legal documents. First, there is a form of coalition bargaining between multiple unions and employer in an enterprise so as to arrive at a satisfactory settlement. Then conciliation is sought (not after the failure of direct negotiations as in the first phase) in order to convert the agreement into a legally binding document. The terms of the agreement are signed in the presence of the conciliation officer, making the contract legally binding on all parties.

The states can add their own labour legislation to the central labour statutes, and in the early 1970s, Maharashtra, Gujerat, Rajasthan and Madhya Pradesh enacted laws on trade union recognition. The Maharashtra Recognition of Trade Unions and Prevention of Unfair Labour Practices Act became effective from 1975. Failure to bargain with the (now-defined) representative union became an unfair labour practice under this Act. Its actual capacity to penalize is, however, questionable (Deshpande, 1992).

In terms of wage dynamics in the urban labour markets, the second phase of unionism was marked by significant changes as well. While the evidence suggested declining, or at best stationary, real wages during the first phase of unionism, Madan (1977) pointed out that the data used in earlier studies suffered from a serious downward-bias as they referred to a restricted category of low-paid workers. Using wage data generated by the Annual Survey of Industries (all workers) he found that the real wages of manufacturing workers had in fact increased since the early 1970s; he also showed that the proportion of low-paid workers to all workers had declined during the second phase. This hypothesis finds further support in the papers by Tulpule and Datta (1988; 1989) who, using Annual Survey data for 1967-84, found evidence of substantial real wage gains since the late 1970s, with substantial variations across industries.

It could be hypothesized that since the mid-1970s, segments within the union movement shifted their goals from those of right to those of interest. This distinction roughly corresponds with the value placed by unions on centralized lobbying (rights) vis-à-vis decentralized collective bargaining (interests). Various factors conditioned such a shift and these became apparent during the third phase of unionism; see below. Some of the key factors were: (a) uneven development of firms within an industry, as well as increasing inter-industry differentiation, meant that some sites were considerably more profitable than others. Unions in these sectors exploited the increased "capacity to pay" during collective bargaining, while unions in the declining sectors had no such opportunity; (b) workers and unions in the profitable sites were more aware of their firm's financial performance through their informed bargaining practices and/or through management willingness to share this information with unions, and (c) workers in these units realized that the leaders of many traditional party-based unions were averse to intense decentralized bargaining, because of their party commitments and their more national concerns.

These shifts and fractures within the organized labour movement had serious implications for union strategies, especially in maintaining solidarity across the entire organized labour movement. Specifically, it become increasingly difficult for the trade union movement in India to act as an "all encompassing" organization, especially during the third phase of unionism. A rather pessimistic variant of neo-institutional analysis of labour unions and collective action (Olson, 1971) can partly explain the above difficulties. According to this "logic of collective action", the assumption of rationality and self-interested behaviour on the part of individuals does not always lead to groups acting in the (collective) interest. This is because of the "public goods" nature of this collective interest: "though all of the members of the group therefore have a common interest in obtaining this collective benefit, they have no common interest in paying the cost of providing that collective good" (Olson, 1971, p. 21). In our context, the "cost" of obtaining collective benefits is different for the various segments of the organized labour movement. To the extent that these "costs" become similar across certain sectors with the onset of liberalization (especially in sectors facing closure, privatization and restructuring), there will be fewer barriers to the trade union movement acting as an all encompassing organization.

Commentary

The second phase reflected an objective crisis of accumulation in industry under the state-led industrialization regime, and a subjective crisis of legitimacy of the "state pluralism" model in the existing industrial relations system. This mode of regulation caused the organized labour movement to split up. The unions of the left had alternate voices, and employers often found them easier to deal with in spite of their greater militancy. The Emergency Regime represented a failed attempt by the state to impose a Latin American-type of corporatism in labour-management relations. Dissension occurred at the end of this second phase both within capital and labour. In the growth sectors the labour-management relationship increasingly turned to an informed decentralized bargaining mode, while in the declining sectors, the government and the large centralized trade unions found themselves enmeshed in the "state pluralism" mode.

The analysis of Rudolph and Rudolph (1987, Ch. 10) illustrates the fact that "organized labour has not been able to challenge India's centrist ideology and politics, that is, to mount or support a left class party" (p. 259). These authors characterize the Indian industrial relations regime as one of state domination and "involuted pluralism", using the term involution "as a metaphor for a decline or loss of vigour that results from a replication of units whose increase in number is accompanied by a decline in effectiveness" (p. 269). They refer to the multiplicity of trade unions: in 1950/51 there were 2,002 registered trade unions, and in 1979 there were 10,021, while the average size of unions increased only marginally (see table 2). According to Rudolph and Rudolph, this "debilitating process" tends towards entropy, so that even though the number of unions increases manifold, union density keeps falling.

One might argue that the silent changes that took place in collective bargaining practices in the growth sectors of the economy represented attempts by both unions and capital to arrive at new governance and institutional structures so as to minimize transaction costs. The fractionalization of unions into smaller groups could then be interpreted as an efficient solution to organizing workers in a changed economic environment. In the public sector sites, economic crises created uncertainties in previously stable internal labour markets. In terms of industrial relations strategy, employers and unions in a few sectors had some strategic choices at their disposal.

Towards the end of this phase it became apparent that there were troubling union "monopoly" effects, especially in the public sector. Public sector and government employees received considerable fringe benefits, whereas in the private sector, unions had to bargain hard for such benefits in a competitive economic environment. Inflexible contract provisions, especially those relating to deployment and retraining and to seniority rules, increased costs substantially in the public sector. Overall, fractured union "voices" searched for a collective mode of expression as this second phase of unionism witnessed more industrial strife than the other three phases both in terms of the number of disputes and the number of workers involved (see table 3).

3.1.3 The third phase of unionism (1980-1991)

This phase corresponds to the Sixth (1980-85) and the Seventh (1985-90) Five-Year Plans, as well as the two Annual Plans (1990-92). Average annual growth rates during this decade were about 5.7 per cent but employment grew at only around 1.8 per cent (Papola, 1994). Employment elasticities in major sectors, especially in services, fell drastically during this time (Papola, 1994). In terms of Joshi and Little's (1994) analysis, the third phase of unionism corresponds to two distinct sub-periods: 1979/80 to 1984/85 and 1985/86 to 1990/91.^{(Endnote 9)} During the first part, the economy suffered from severe internal and external shocks: one of the worst droughts since independence occurred in 1979, there was trouble in the northeast, an industrial recession in 1980/81, rising inflation and increasing oil bills. All this led to a balance of payments crisis and then to a massive IMF loan. In May 1984, India terminated the programme after drawing SDR 3.9 billion (Joshi and Little, 1994, p. 60). This period was also turbulent on the political front with Mrs. Gandhi's return to power in 1980, the assault on the Golden Temple in Amritsar, Mrs. Ghandi's assassination in October 1984 and the landslide victory of the Congress Party with her son Rajiv as Prime Minister.

The second part of this phase is associated with Rajiv Gandhi's economic liberalization measures. The economy moved away from import substitution towards strategies that encouraged export promotion and domestic competition. This was bought about by partial deregulation, financial liberalization, exchange rate policy, taxation, and export incentives (Joshi and Little, 1994, p. 62).^{(Endnote 10)} After 1988 the country experienced severe unrest. The 1990 Gulf crisis was followed by economic recession and political turmoil.

India was now facing a full-scale macroeconomic crisis. "Agreement was reached with the IMF in January 1991 on a loan of $1.8 billion, partly out of the Compensatory Financing Facility (to offset increased oil imports) and partly as a first credit tranche standby" (Joshi and Little, 1994, p. 66).

The macroeconomic changes during this phase had a profound effect on the political economy of trade unionism and labour markets, as well as on the structure of industrial relations. On the union front, this phase started with a massive public sector strike in Bangalore during 1980/81 which involved giant public enterprises such as Hindustan Machine Tools, Hindustan Aeronautics Limited, Electronics Corporation of India, and Indian Telephone Industries.

The event that characterized the first part of this third phase was the famous, and much studied, Mumbai textile strike of 1982 (see Patankar, 1981; Pendse, 1981; Bhattacherjee, 1988; 1989; Van Wersch, 1992). This started as a wage and bonus issue in a few mills in late-1981, and soon developed into an industry-wide stoppage that ultimately became the longest strike in post-independence labour history. The basic cause was disquiet among the rank and file about the 1947 Bombay Industrial Relations Act, which had imposed an industry-wide bargaining structure with an unrepresentative union (affiliated to the Indian National Trade Union Congress) as the sole bargaining agent of workers. As a result of the internal differentiation within the mills which took place during the 1970s, workers wanted more control over their labour market- and industrial relations outcomes at the level of the individual mill. They approached Dr. Datta Samant to lead their struggle: his main project was to form and lead an "independent" trade union movement in Western India.

The textile strike ended in a whimper late in 1983: many workers returned to their villages, their lives ruined, employers restructured their mills in the advanced textile sector, and the credibility of the government-installed union in the industry declined to levels from which it could never recover (Bhattacherjee, 1988). Evidence from the immediate post-strike period seems to suggest that workers and unions, at least in the profitable mills, negotiated their own decentralized bargaining agreements (Bhattacherjee, 1989; Van Wersch, 1992). After Datta Samant formed the Kamgar Aghadi Party and won a few seats in Parliament in 1984, many commentators felt that this was a new and encouraging beginning for the organized labour movement in India.^{(Endnote 11)}

Another important feature of this third phase was the proliferation of "independent" unions operating in the major industrial centres and competing with the traditional party-affiliated trade unions (both of the Congress and of the Left). In Mumbai, for example, the decline of the Left unions is partly attributable to their general opposition to intense decentralized bargaining (Pendse, 1981). Segmented and uneven developments in the industrial sector tied workers' earnings to the fortunes of the plant in which they were employed. An analysis of plant-level contracts from the Greater Mumbai-Thane industrial corridor revealed that, ceteris paribus, the "independent" unions delivered a higher wage and fringe package than did the affiliated unions at the beginning of this third phase (Bhattacherjee, 1987b). In a number of multinationals, it was found that workers with their "independent" unions exerted considerable control over the labour relations process, often more than their counterparts in the host country (see the example of Phillips in Banaji and Hensman, 1990).

In the late 1970s there was a phenomenal rise in the number of disputes led by unaffiliated unions and the importance of politically affiliated unions declined (Bhattacherjee, 1987a, p. 57). The ratio of registered unions that submit returns about their membership size to the total number of registered unions fell sharply from 60 per cent in 1962 to 21 per cent in 1974 to 13 per cent in 1982 (Bhattacherjee and Datta Chaudhuri, 1994a, p. 70). This would support the hypothesis of the rise of "independent" unionism, if the total number of registered unions that submit returns proxies the traditional party-affiliated trade unions.^{(Endnote 12)} By 1989 the Labour Ministry listed the following union organizations in their registry: Indian National Trade Union Congress, All India Trade Union Congress, Centre of Indian Trade Unions, Hind Mazdoor Singh, the Bharatiya Mazdoor Sabha (BMS, affiliated to the BJP), the Hind Mazdoor Kamgar Party, United Trade Union Congress (UTUC), United Trade Union Congress Lenin Sarani, National Labour Organization, and a number of small independent unions.

Finally, this phase was marked by increasing inter-regional, inter-state, and inter-city variations in the nature of labour-management relations.^{(Endnote 13)}

In a study of Bombay, Calcutta, Madras and Bangalore, Ramaswamy (1988) points to significant inter-city differences in the texture of labour-management relationships. He writes (1988, p.17): "The driving force of the Bombay labour movement are union leaders who disclaim allegiance to political parties and their trade union federations. What we find here is the most evolved Indian version of business trade unionism", and "the city has witnessed the steady decline, if not eclipse into oblivion, of ideological trade unionism". This clearly has something to do with the fact that private and multinational firms dominate Mumbai's urban economy. As Ramaswamy puts it: "there is an organic linkage between the buccaneering spirit of Mumbai unions and the eclecticism and adaptability of private enterprise".

In sharp contrast is the case of Calcutta, where a highly politicized industrial relations regime prevails with the dominant trade union federation (the CITU) under the close watch of the dominant political party (the CPI-M). This has created considerable inflexibilities for management, and has partly prohibited the growth of independent trade unionism. Bangalore, a city where both private and public sector enterprises thrive, especially those in the information technology industry, has witnessed the rise of plant and firm-based unions (as in Mumbai). Inter-city differences, attributable no doubt to different political, social, and urban histories, emphatically suggest the inherent difficulties in trying to generalize about an "Indian" labour relations system. With the passing of time, especially with the onset of the fourth phase (see below), attempts at this kind of generalization will become more and more difficult.

Changes in union structure, together with macroeconomic developments, considerably affected both employment and the wage structure. Between 1980/81 and 1988/89, while employment growth declined, the capital-labour ratio and labour productivity increased by 8 per cent and 7.5 per cent per year respectively (Ahluwalia, 1992). According to Ghose (1992, p. 95), "The most striking fact is that the 80s have been the best decade in terms of economic growth but the worst decade in terms of employment generation". Moreover, employment growth decelerated in all sectors of the economy and open unemployment increased in the 1980s (Ghose, 1992). The search for labour market flexibility in Indian manufacturing led labour-intensive firms and those engaged in the production of consumer non-durables to subcontract and outsource their production to the unorganized sectors (Ramaswamy, 1999).

Unions in the organized and more profitable sectors (often "independent" unions), managed to secure part of these productivity increases through militant bargaining and/or through productivity bargains that contained effective incentive structures. This resulted in slower employment growth. In the relatively unorganized and less profitable sites, workers and unions lost out. While formal sector employment as a percentage of total employment in manufacturing fell from 24.5 per cent in 1972/73 to 17.4 per cent in 1987/88, real wages of workers and "other employees" in organized manufacturing increased at a rate of 5.8 per cent and 4.1 per cent between 1983 and 1986 respectively, whereas low-paid workers suffered declining real wages (Ghose, 1992, p. 97). As the union wage effect increased significantly in the profitable sites (especially in consumer non-durables) during this phase, employers cut back on hiring and started retrenchment, increasing the capital-labour ratio which in turn increased labour productivity. Thus, according to this scenario, the faster growth of real wages in the 1980s played a role in slowing employment creation (Ahluwalia, 1992).

Bhattacherjee and Datta Chaudhuri (1994b) found that: (a) in the high-paid sector, real wages increased after the late 1970s and there were wage returns from striking, (b) in the low-paid sector, real wages declined after the early 1980s, and employers could lower wages by imposing lockouts; and (c) in terms of union structure, low-paid workers gain as unions submitting returns (proxy for traditional unions) increase their dominance within the union movement, whereas, high paid workers gain as registered unions not submitting returns (proxy for plant-specific unions) increase their dominance. The traditional unions predominant in the older industries provided overall protection to their members as long as these industries grew. As they declined, and as plant-specific unions reaped returns in the high-paid sector, workers in the low-paid sector "became more vulnerable to competitive forces and could no longer count on the traditional 'wage-welfare' functions provided by the party-based unions. Workers in this segment will find it difficult to form strong plant-specific unions due to increasing instability in their product markets" (Bhattacherjee and Datta Chaudhuri, 1994b, p. 459).^{(Endnote 14)}

Jose (1992) examined earnings, employment, and productivity trends for 19 industry groups for the period 1970/71 to 1987/88. His findings were: (a) whereas the 1970s are associated with employment growth in both high and low wage sectors and with stagnant and even declining productivity levels, the 1980s are associated with slower employment growth with rising productivity levels, especially in the high wage sectors; (b) higher productivity brought about the (modest 2-3 per cent per year) increase in real earnings in the high-wage sectors. These findings seem to fit with the characterization of different types of union dominating these two labour markets since the early 1980s. Jose (1992) maintains that technological change led to the rise in wages and productivity, whereas subsequent analysts inverted the hypothesis to argue that union militancy and higher wages resulted in technological change that subsequently led to a decline in employment.

Labour researchers towards the end of this third phase focused on employment inflexibilities embedded in the Industrial Disputes Act (see Mathur, 1992). The 1982 amendment of the Industrial Disputes Act provided that a firm employing more than 100 workers (reduced from >300) needed permission from the state government to lay-off or retrench workers. Fallon and Lucas (1991) showed how employment would have been higher in several sectors without the 1976 and 1982 amendments. Mathur (1992) recommended that the sections pertaining to permission for lay-off, retrenchment or closure be deleted from the Act.

To remedy the limitations of the Industrial Disputes Act and the Trade Union Act, from both the union and the employer point of view, a number of changes were proposed in the Trade Unions and Industrial Disputes (Amendment) Bill, 1988 (Mathur, 1992). The proposed changes would reduce the fragmentation and multiplicity of unions, clearly define the bargaining agent by providing for a secret ballot, promote internal leadership, create state-level industrial tribunals, force employers to set up comprehensive bargaining councils to facilitate internal grievance settlement, and so on. After considerable debate however the bill was rejected because of controversy over the definition of "industry" (Mathur, 1992, p. 50).

Commentary

The "independent" unions arose in Western India from within the rank and file, throwing up their own shopfloor leaders, and initially encountering considerable hostility from management due to their militant bargaining strategies.These unions arose primarily as a result of dissatisfaction with the bargaining weakness of the traditional unions at enterprise level.

One way to distinguish between unions is to look at the arguments which they emphasize at micro level (Bhattacherjee and Datta Chaudhuri, 1994b). Given the level of demand for organized labour, unions, like employers, face a wage-employment trade-off during contract negotiations. The situation that prevailed during the third phase, and continues to prevail, can be described in the following way. There are two types of union: those that are "altruistic" and those that are "selfish". Given a choice between wage increases and an increase in employment, the "altruistic" unions choose the latter, as membership is the crucial argument to be maximized; consequently, their bargaining strength is positively related to their size. The large and established national trade union federations that are typically affiliated to political parties, display this type of union behaviour. On the other hand, "selfish" unions are interested in maximizing the wages of those already employed in a particular enterprise, and confronted with the trade-off, they opt for wage increases. In this case, small is powerful. The so-called "independent" unions seem to approximate the latter. It is clear why the "selfish" unions would be more likely to sign productivity bargains than the "altruistic" unions; for the latter, productivity bargains cut at the source of their strength, i.e., growing union membership. While at micro level, in profitable sites, bilateral collective bargaining with "selfish" unions seems to have generated efficient and cooperative outcomes, at macro level, the trade union "movement" in a democracy has to be an all-encompassing organization. Here lies the difficulty in formulating a fair but efficient industrial relations policy: what is efficient at the micro level is not fair at the macro level (Bhattacherjee, 1996).

In terms of the "monopoly versus collective voice" framework, labour market evidence from the third phase of unionism seems to suggest that some monopoly effects dominated the collective voice, and that the proliferation of "selfish" unions and their microeconomic success may have had a negative effect on unorganized labour markets. In the private corporate sector, firms attempted "efficiency wage" strategies, usually with the cooperation of unions, but often changing the nature of plant or firm-level unions in the process. In public-sector internal labour markets, unions often became involved in unproductive activities while managing these enterprises. The importance of concepts such as internal governance structures of trade unions and their effects on efficiency and fairness became increasingly relevant during this phase of unionism.

In terms of industrial relations strategy, the actors had more options than before and one could even suggest that some segments operated their own strategy, insulated from outside forces. Thus, effective "gainsharing" resulted from efficient bargaining in some sectors, whereas in the unorganized sectors, a free market regime prevailed. Once more, attempts at serious reform of industrial relations law came to a predictable grinding halt.^{(Endnote 15)}

3.1.4 The fourth phase of unionism (1992-2000)

In June 1991, the ruling minority government decided to adopt the World Bank-IMF stabilization and structural adjustment programme.^{(Endnote 16)}

"In July 1991 the rupee was devalued twice, quotas on the import of intermediate and capital goods were reduced, tariffs were brought down, the state monopoly on exports and imports was ended and a statement on industrial policy was presented along with the Union Budget, which was aimed at reducing the fiscal deficit by two and a half percentage points" (Mathur, 1993, p. 333). The fourth phase of unionism more or less corresponds to the Eighth (1992-97) and the Ninth (1997-2002) Five-Year Plans.

According to Nagaraj (1997, p. 2870) "On average, the Indian economy grew at 5.3 per cent during the first five years of the reforms (1992-96), compared to 5.9 per cent during 1986-91". The tertiary sector grew fastest in the 1990s, at about 6.8 per cent per year. The economy has become considerably more "open" than ever before. There was some apprehension that government expenditure in the social sector would decline significantly, but Nagaraj (1997) found that social spending, averaged over four years since the reforms, did not suffer; most of the cuts made in defense and economic services. Again, contrary to expectations, investment performance in India actually improved after the reforms, with private corporate business emerging as the economy's "leading sector". However, in terms of industrial growth performance, "the manufacturing growth rate since the reforms is lower and its composition is uneven" (1997, p. 2875). Even though public investment had witnessed deep cuts since the reform, "public sector output growth and profitability improved" suggesting better resource utilization. Nagaraj concludes: "In sum, the good news (so far) is that there is no major, unqualified, bad news".

Transnational corporations, as expected, have reacted very favourably to the new economic policy. Chaudhuri (1995) examines the mechanisms through which transnational corporations have achieved entry and growth: through the ousting of Indian partners, through extensive mergers and acquisitions, and through expansion and fresh entry. The gross inflow of foreign direct investment rose from Rs.5.3 billion in 1991 to Rs.141.9 billion in 1994 (US$1 is approximately Rs.0.44), and although Chaudhuri is sceptical about the prospects of this investment leading to export-oriented growth in India, it is expected to generate some employment.

If the 1995-96 period was a slowdown, and 1996-97 was a year of "industrial recession and political uncertainty", the year 1997-98 can be described as an "elusive economic recovery" with a "hesitant government".^{(Endnote 17)} A whole range of issues on the economic reform agenda remain incomplete: infrastructure development, greater transparency in investment procedures, restoration of business confidence, review of import duties, further banking sector reforms, and most important, public sector and labour market reforms.

The available data seem to suggest that economic reform has led to an increase in rural poverty and a decline in urban poverty; in fact, urban poverty was lower in 1993-94 than in any pre-reform year (Sen, 1996). Deshpande and Deshpande (1996) found that although the initial stabilization years "took some toll of organized manufacturing employment", subsequent structural adjustment led to employment growth at around 2.3 per cent between 1992/93 and 1994/95. If this rate continues for the next few years, "employment in the factory sector would be about 12 per cent higher at the turn of the century than in 1990/91" (1996, p. 18). In terms of the structure of employment, these authors found that in some sectors the employment share increased after the reforms (textiles, transport equipment, chemicals, beverages, metal products) while in others the employment share decreased (machinery).

Mundle (1992) presented employment and unemployment projections under "high" and "low" growth scenarios taking into account employment elasticities and labour force participation rates. Even under the most optimistic scenario in the private sector, given declining employment elasticities in organized manufacturing and the unlikely expansion of employment in the public sector, it is quite evident that a "large majority of the nearly 80 million persons who will join the labour force during 1999-2000 will have to find work as self-employed or casual workers" (Visaria and Minhas, 1991, p. 978). In terms of the impact of economic reforms on women workers, Deshpande and Deshpande (1992) believe that female unemployment may go down but the openings available will be low-wage jobs in secondary labour markets. In the light of the above, it seems imperative to expand the existing public employment schemes. However, the most appropriate strategy for a predominantly agrarian economy such as India must be to raise labour force participation rates and shift the workforce away from agriculture into more productive sectors so as to generate large economy-wide productivity gains (Bhaduri, 1993).

One of the main objectives of the economic reform package is the restructuring of unprofitable public sector enterprises.^{(Endnote 18)} These enterprises are free to reduce their workforce through voluntary retirement schemes (VRS) assisted by the national renewal fund (NRF) instituted by the government, and by amendments to the Sick Industrial Companies Act 1985. The strengthening of the Board for Industrial and Financial Reconstruction considerably facilitated this process (Mathur, 1993). The objectives of the NRF were to provide assistance to cover the costs of retraining and redeployment made necessary by modernization, technological upgrading, industrial restructuring, and possible closure. In 1993-94, Rupees 7 billion (US$1 = approx. Rs0.44) were allocated to the VRS in the central public sector enterprises; nearly Rupees 4.9 billion was allocated to the textile sector alone (Muralidhar, 1993).

While recruitment was all but frozen (especially at lower levels), the government also froze the centralized wage bargaining process for a few years after 1992. It later opened the negotiation process and attempted to decentralize bargaining by announcing that any wage increases would have to be absorbed by the specific enterprise as these could no longer be passed on to the final price. In other words, the new policy clearly stated that any additional wage burden would not receive budgetary support (Venkataratnam, 1996).

The need for tripartite consultation was clearly felt during the early years and many meetings took place. Mathur (1993) documents the experience of consultation during the early phase of structural adjustment in India (1990-92), and suggests that although the government partly diffused possible tension through its consultative approach, unions had "serious misgivings about the adequacy of consultation at (the) industrial or enterprise level" (p. 344).

Labour commentators have found that the centralized and traditional union structures often fall short of worker expectations, and in many instances, they are giving way to independent and decentralized union structures (Davala, 1992; Muralidhar, 1994). However, this (current) fourth phase has witnessed the growth of the Bharatiya Mazdoor Sangh, affiliated to the Bharatiya Janata Party. In the state of Maharashtra, the trade union movement has become quite volatile since the locally based Shiv Sena party and its affiliated union, the Bharatiya Mazdoor Sena, made deep inroads.

During this phase the public has become acutely aware (largely through the print and visual media) that trade unions represent a declining "sectional interest group". Bhaduri and Nayyar (1996, p. 139) point this out in no uncertain terms: "The government also needs to protect consumers against sectional interests of many unrepresentative trade unions. While the trade union rights of workers must also be respected in any democracy, the government must also ensure, perhaps through secret ballot, that no unrepresentative union harasses ordinary consumers". They go on to add: "…recognition of workers' rights must go with appropriate regulations for recognizing these rights. All such rules of the game need to be set transparently, and without partisanship".

On 10 January 1999 the government announced the second National Labour Commission (the first NLC was set up 30 years ago). The terms of reference lay down that the commission should suggest rationalization of existing labour laws in the organized sector and recommend umbrella legislation to ensure minimum protection for unorganized workers. The commission has a two-year term and is made up of representatives from government, trade unions and industry. Trade unions feel that workers have little protection from the whims of errant management, and that any alteration in the law would only add to managerial power. For example, the proposal to relax the law on contract labour in order to generate more jobs on contract for the unorganized sector is interpreted by the unions as a move to undercut permanent unionized jobs. More recently, proposed changes in the Industrial Disputes Act will make it difficult for trade unions to call wildcat strikes and the amendments will dilute the need for employers to have government approval for a lockout. On the other hand, they give the tribunals more power to penalize errant employers.

The present government dissolved the National Renewal Fund and entrusted the corpus to the Ministry of Industry. The money in the fund will now be given to public sector enterprises directly by the Ministry. The government hopes to monitor the use of funds more effectively by not involving the administrative departments that control these enterprises. More recently, the government has announced that it will develop a comprehensive strategy to deal with unprofitable public enterprises and "hammer out a mechanism which makes it easier to close chronically sick units". Officials in the Ministry of Finance announced that more funds have to be pumped into the voluntary retirement schemes so that workers can be retrained and redeployed in viable public sector units.^{(Endnote 19)}

Commentary

In terms of labour market and industrial relations reforms, the continuation of economic liberalization would undoubtedly lead to more employment flexibility, greater decentralization in bargaining structures (especially in public enterprises) and hence less government intervention in the bargaining process, fewer strikes, and a possible halt to the fragmentation of the union movement. On the positive side, this could mean more employment and a more effective union voice at both micro- and macro-level, with industrial pluralism being strengthened. On the negative side, the proposed reforms could increase managerial power and accelerate the growth of the non-union sector, leading to a decline in the power of organized labour. At microeconomic level, unions of all political hues are cooperating with management in the restructuring process (often because they have no choice). However, at the macro-level, the organized labour movement is generally extremely critical of globalization and reform. From a strategic viewpoint, trade union federations of different political affiliations will have to forge a united front at the macro-level to ensure their continuing growth as "all-encompassing" organizations. This means organizing non-union workers.

Differences within the centralized trade unions, political parties, and state governments have indefinitely delayed the passage of industrial relations reforms. The lack of consensus and political instability at the centre since 1992 (several coalition governments have failed to complete their full term), mean a lack of political will to carry out these reforms. A worrying consequence of the inability to reach a national consensus on labour market reform is the future effect on regional labour markets and labour relations in general of heightened inter-state competition to attract foreign and local capital. In the absence of national reforms, various states may attempt to level down their labour market institutions by offering incentives to investment. The latter will have a profound effect on regional variations in labour standards. As these divergent trends will make it increasingly difficult for the centralized trade union federations to act at national level, it is in their immediate interest to press for industrial relations reform.

3.2 The issues

The above discussion on the four phases of Indian trade unionism sketched out the interaction between changes in the overall economy and their effects on the labour market and on industrial relations. In this section we examine several issues that arise from the discussion and speculate on trends in the near future. These issues are: unionization and employment; union structure and union density; wages and working conditions; collective bargaining; industrial conflict; labour-management relations; inter-state variations; unionization and women workers; and the changing public perception of trade unions.

3.2.1 Unionization and employment

Declining employment elasticities imply that more output is attained with less employment. This could be due to the fact that employers are investing in more capital-intensive technologies, and that there has been a considerable amount of labour shedding in the private and public-sector enterprises since the mid-1980s. Unions can affect these employment elasticities by resisting technological change that increases the possibility of substituting between capital and labour and by limiting the availability of goods and services that compete with the output of unionized firms. In addition, union bargaining power varies indirectly with labour's share in total costs: unions are more powerful in relatively more capital-intensive firms and industries, as the demand for labour is relatively inelastic compared to labour-intensive sectors. Employers in capital-intensive firms find it much easier to meet union wage demands compared to employers in labour-intensive firms. Finally, it is in the interest of unions to raise the price of other inputs, particularly non-union labour, as this increases the cost of switching from union to non-union labour (see Borjas, 1996, pp. 126-127).

To illustrate the usefulness of the above, consider the following. In the current scenario, labour and unions in the more labour-intensive sectors in India face considerably more uncertain and insecure times than their counterparts in more capital-intensive sectors. Several firms in the textile and jute industries are near closure. On the other hand, the majority of plant-based "independent" unions are located in more capital-intensive industries, and union bargaining power in several public sector sites (such as transportation and banking) is high due to its specific market characteristics defined by state monopoly. In some states, such as in Kerala and West Bengal, progressive unions in conjunction with their state governments have improved the wages and working conditions of agricultural and unorganized workers, thereby curtailing the growth of income inequalities within the workforce as a whole.

Over time, average annual growth rates in GDP have outstripped average annual growth rates in employment (table 4) and employment elasticities in major sectors have fallen (table 5). Table 1 shows the level of employment in the private and public sectors over the "four phases of unionism". This table clearly shows the following: in the first phase, public sector employment increased sharply and private sector employment increased marginally; during the second phase, public sector employment increased rapidly, whereas private sector employment remained sluggish; during the third phase, employment growth in both sectors sharply tapered off; and during the first few years of the fourth phase, growth in both sectors remained nearly static.

Table 6 presents a sectoral distribution of employment in the formal economy. We note two trends: since the third phase, growth in employment in manufacturing, both in the public and private sectors, has remained virtually stagnant; however, during the third phase, employment increased in public services. We also observe the phenomenal growth in employment in services (especially in the public sector) relative to manufacturing. Finally, except in manufacturing, the public sector continued to be the dominant employer in the organized economy. Table 7 shows a sector-wise ranking of the number and membership of trade unions that register and submit returns over the four phases of unionism. We observe that: (a) manufacturing leads in the number of unions and membership size; (b) construction lags behind most other sectors.

Given that the macro data on employment and unionization in India are riddled with problems and contain errors of omission, detailed case studies of specific industries and regions have revealed significantly different trends. The Friedrich Ebert Stiftung study (Davala, 1992) is an example. It covers tea plantations and the jute industry in West Bengal, the coal sector, ports and docks, the engineering industries of Andhra Pradesh and West Bengal, the power sector in Andhra Pradesh, and the chemical and pharmaceutical industry in Maharashtra. Although the study shows a fair number of inter-industry and regional variations, there were some striking similarities. The trend everywhere was a downsizing of permanent employment and the proliferation of contract, temporary, and casual jobs. There were very low rates of unionization of contract and casual labour, with the unions being more sensitive to the plight of such workers in the newer industries. ^{(Endnote 20)} In the state of Maharashtra, permanent workers and their unions in the chemical and pharmaceutical industries have realized that their well-being is ultimately tied to their fellow workers in the "reserve army of labour". The study found that unions organize on an industry- and/or region-wide basis in the older industrial sectors, but the enterprise becomes the unit of organization in the newer industries. This pattern corresponds with the prevalent bargaining structure (i.e., industry- and region-wide in the older industries, and plant-level in the newer industries). As Ramaswamy concludes in this study (Davala, 1992, p. 231):

The common thread running through most of these variations is the desire of workers to gain greater control over their unions. This might well be the case in the older industries as well, but the structure of trade unionism does not permit much scope for the expression of these aspirations. In the newer industries, on the other hand, trade union structure itself appears to have been influenced by this fundamental force.

The "market friendly" views of the relationship between unionization and employment in India, especially during the third and fourth phase, was elaborated earlier, but needs to be emphasized in this section as well. The "monopoly effects" of trade unions, together with inflexible labour laws, have enabled employers to move up their demand curves and have practically frozen employment growth in permanent unionized jobs. Due to union wage mark-ups, employers have increased the capital intensity of production, thereby raising productivity. But this route can only lead to jobless growth, clearly a sub-optimal outcome in a labour surplus economy. Those who oppose this view point out that wage increases took place not because of union power but largely because of an intensification of the labour process resulting from a decline in union power. A disaggregated analysis probably comes closest to reality: in the more prosperous sectors, with low elasticities of labour demand, "selfish" plant-specific unions managed to extract generous wage increases; in the less prosperous sectors, largely in the public sector, "altruistic" unions affiliated to political parties have had little success with centralized bargaining procedures in unstable product markets.

In the buoyant sectors of production, even before 1991, employers have managed to execute viable exit policies through generous Voluntary Retirement Schemes, with the cooperation of enterprise-based unions. In these largely private sector sites, "bringing the union in" has clearly paid dividends in terms of generating "strongly efficient contracts" (that is, when the labour contract leads a unionized firm to hire the competitive level of employment). In the public sector, however, exit schemes (available through the National Renewal Fund) have few individual takers or else they face union resistance at national level. The general secretary of the All India Bank Employees' Association says, "We will resist any attempt to introduce a VRS. As far as we are concerned, job security is more important than wage revision".^{(Endnote 21)}

The practice of "featherbedding" (employing more people than required) in a whole range of public enterprises is no longer financially viable. The union movement as a whole, especially the large centralized and industry-wide public sector unions, may benefit more from unionizing the unorganized than from attempting to preserve unproductive jobs.

3.2.2 Union structure and union density

The number of registered unions increased from 3,766 in 1950/51 to 55,784 in 1993, whereas the average size of those that submitted returns decreased from 577 to 460 during the same period (see table 2). As stated earlier, not all registered trade unions submit returns on their membership. If we assume that the registered unions which submit returns are the large, established politically-affiliated unions with the organizational resources to engage in this process, whereas the registered unions that do not submit returns are the plant-based independent unions, then movements in the fifth column of table 2 (percentage of unions submitting returns to total registered unions) makes for an interesting interpretation. It appears that the centralized unions dominated during the first phase, lost ground during the second phase, and then sharply declined during the third phase, dropping to an all-time low of only 12 per cent in 1993. Although this is only one possible reading of the data, several detailed case studies have also pointed to these movements in union structure (Ramswamy, 1988; Davala, 1992).

To illustrate the above with some region-specific examples, consider the following: (a) in the jute industry in and around Calcutta, union membership for each of the major unions declined with a fall in the proportion of permanent workers, even though "a large number of small, ultra-left and even unaffiliated unions have emerged during the last decade" (Sen, 1992, p. 48); (b) in the coal mines in Eastern India, the independent unions represent more than a quarter of the total workforce (Sen, 1992, p. 71); (c) in the state of Andhra Pradesh, "the emergence and growth of non-affiliated unions is perhaps one of the most striking features of the trade union scenario in the engineering industry during the last decade" (Guptan, 1992, p. 124); and (d) more than 50 per cent of the unions in the chemical and pharmaceutical industries in the state of Maharashtra are unaffiliated independent unions (Davala, 1992, p. 210). The fourth phase is also associated with the sudden rise of the Bharatiya Mazdoor Sabham, especially in those states governed by the Bharatiya Janata Party.

The decentralized independent unions have in some instances attempted to generate firm-wide agreements but they have been unsuccessful because of considerable employer resistance at firm level. Some recent commentators have suggested that these unions have been on the defensive ever since the reforms. This is partly because they lost staff through voluntary retirement schemes; it was also due to the strategies of multinationals which severely cut back on permanent unionized employment. Other factors were the setting up of non-union facilities at new sites and the practice of subcontracting work to the informal sector (Banaji and Hensman, 1995; Noronha, 1996).

3.2.3 Wages and working conditions

Table 8 provides data on the movement of the consumer price index for industrial workers and urban non-manual employees across the four phases of unionism. Tables 9 and 10 provide some data on the movement of nominal (a general index for 12 industries) and real wage rates (in organized manufacturing).

Standard neoclassical economic theory would argue that increased global trade raises the earnings of unskilled workers relative to those of skilled workers in a country such as India, which has unlimited supplies of the former. This implies that India's exports are largely unskilled labour-intensive products. Nambiar et al. (1999) found that wage disparity for the period 1980/81 to 1992/93 increased but only marginally. In addition, they found that the disparity rises from less skill-intensive to more skill-intensive sectors. To the extent that the union voice reduces earnings inequality within the unionized workforce, employers, especially in the private sector, have "manufactured consent" with the unions in setting up ingenious pay incentive systems.

Wage determination in the organized economy varies significantly between the private and the public sector (Datta Chaudhuri, 1996; Anant and Sundaram, 1998). In the private corporate sector, where collective bargaining largely takes place at enterprise level, unions that are willing to accept some risk have benefited from a form of gainsharing by agreeing to tie a significant part of the monthly pay to incentives. The incentive structures are designed to generate cooperative behaviour at the departmental, plant and firm level. Risk-averse unions, usually more concerned with employment growth than with members' wages have resisted management attempts to impose such systems.

In many of the older sectors, such as tea plantations and jute in Eastern India and textiles in Western India, industry-wide bargaining is the institutional norm. With the advent of economic liberalization, this bargaining structure will be under pressure to decentralize some of the outcomes as inter-plant and inter-firm differences become wider (see 3.2.4 below). Similar decentralizing pressures will be felt increasingly in public sector industries, even though the Bureau of Public Enterprises "sends guidelines for wage settlement to all administrative ministries, setting down norms to be followed in determining basic salaries and the various categories of benefits for different classes of employees" (Datta Chaudhuri, 1996, p. 18). In Coal India, for example, employees in the better-off units feel that their earnings could increase substantially if they were linked more closely to productivity at the unit level. Centralized public sector unions will have to come to terms with these decentralized union "voices".

Salaries and benefits for central government employees in public administration, academic institutions, posts and telegraph, etc., are determined in detail by the Pay Commissions which are periodically set up by the government.^{(Endnote 22)} As Datta Chaudhuri (1996, p. 18) puts it: "The award of the Pay Commission for the Central Government employees becomes the reference point for wage determination in the rest of the public sector." The Pay Commission is the object of considerable lobbying by various unions and employees' associations prior to and during the deliberations.

The recommendations of the Fifth Central Pay Commission have been implemented and employees in the central and state governments have seen their incomes rise substantially. If it is true that the disparity between the average salary of government employees and per capita income is far higher in India than in most other countries,^{(Endnote 23)} the public clearly need to see vast improvements in productivity in this sector given the fairly high additional cost involved in delivering the pay recommendations (Joseph, 1997). The All India Federation of College Teachers' Organizations could, for example, take a lead in self-monitoring their constituency so as to impose work norms and discipline. There are already many rules to deal with employees who do not work and/or who are late; applying these rules occasionally will send strong signals to both employees and to the public (Joseph, 1997).

Industrial accident rates in India, both fatal and non-fatal, are extremely high compared to most countries. While it is true that the occupational health and safety monitoring agencies are weak, it is also the case that unions can intervene significantly more in this area than they are presently doing. Often, workers demand higher wages for increased safety measures during contract deliberations.^{(Endnote 24)} But there are other examples where unions have been closely involved in occupational health and safety matters. One such case is the Occupational Health and Safety Centre that operates out of two union offices in Central Mumbai and was set up largely due to the inefficiencies of the Employees' State Insurance (ESI) Scheme. Among its many achievements, the Centre was able to get the ESI medical board to recognize and compensate mill workers suffering from occupational bysinosis in 1994.

The linking of minimum labour standards and trade agreements, i.e. the "social clause", has generated considerable debate and discussion among trade unions and labour commentators (FES-IIRA, 1996; Hensman, 1996; Bhattacherjee, 1997). There are various economic arguments in favour of the imposition of international labour standards (ILS). To the extent that labour markets in developing countries are beset with imperfect and asymmetric information (for example relating to industrial safety), ILS may level up labour market institutions to correct for market failures of this type. ILS can be used as a redistributive mechanism if the government feels that the market- determined income distribution profile is skewed towards the more skilled workers in relatively protected environments. Trade unions in India interpret this imposed link as a disguised form of protectionism (for the various trade union views, see FES-IIRA, 1996, and Hensman, 1996). This argument leads to the same outcome as those put forward by the 'comparative advantage' trade theorists that the imposition of ILS will lead to a reduction in the net gains from trade, and therefore individual countries should decide on their own labour standards. But surely one has to question this simple view. Why is it that countries with relatively abundant and cheap labour find it difficult to compete in international markets, except in those sectors that have (relatively) lower labour standards? Clearly, this kind of participation in global trade, where inferior labour standards are the "comparative advantage" is unlikely to lead to social progress. Since it is not enough to wait for sustained economic growth to upgrade domestic labour standards, the unions need to forcefully generate demands, both from above and below, for improvements in working conditions. Hensman (1996) spells out an agenda for trade unionists and NGOs that strongly believe that labour standards in India have to be substantially improved, perhaps even through trade links and other forms of international pressure.

3.2.4 Collective bargaining

Two critical aspects of the collective bargaining system are an expansion of the coverage and scope of long-term agreements and the increasing pressure for decentralized bargaining. These tendencies originated during the third phase of unionism and became acute during the post-reform fourth phase. The restructuring agreements cover (among other things): ban on recruitment, job transfers to non-bargainable category, introduction of parallel production, automation and flexibility, transfer of production to subcontracted units, introduction of voluntary retirement schemes, transfer of permanent jobs to contract/temporary workers, merger of units, and a host of other shopfloor restructuring provisions (Venkataratnam, 1996). Concession bargaining in several units has led to job and wage cuts, a freeze on cost-of-living allowances, and suspension of industrial action for a period of five years. While the above suggests that management has had the upper hand in recent years, Venkataratnam (1996) also lists some "unusual" clauses that suggest a positive union "collective voice" effect: linking allowances such as house rent and children's education to attendance, permitting pregnant women to refuse to work on computer terminals, voluntary retirement schemes for contract labour, and so on.

We have already mentioned government attempts during the post-reform period to decentralize bargaining in public sector units by tying unit-level wage increases to productivity increases (rather than passing them on as price increases), and by consistently announcing its refusal to provide budgetary support to these wage increases. The central trade unions are uniting to pre-empt a government attempt to switch to a 10-year wage settlement in public sector units instead of the present 5-year duration. Union leaders say the move will affect the pay revision prospects of about 2 million workers in the public sector. Private employers have been lobbying for some time to increase the duration of contracts in public sector units so as to prevent disruptions arising from frequent industrial disputes, the costs of which they have to often bear. Unions, especially the All India Trade Union Congress and the Indian National Trade Union Congress, are strongly against such a move.

Even outside the public sector, pressures for decentralized bargaining are emanating from both workers and employers. What happened in the textile industry is now being repeated in the jute industry in and around Calcutta. Due to the uneven development of textile firms in Mumbai in the early 1980s, and due to the undemocratic nature of union representation in the industry, both employers and unions in the more profitable mills wanted to break away from the industry-wide agreement in force and set up their own mill-level agreements (Bhattacherjee, 1988; 1989). Recently, employers in two jute mills defied the industry-wide agreement between the Indian Jute Mills Association and the many unions by offering a higher cost-of-living allowance to avert mill-level strikes. The central trade unions are in a dilemma for the first time in the jute industry: on the face of it, they oppose any kind of bipartite settlement by insisting on an industry-wide agreement at tripartite level. But by encouraging strikes at individual mills, they are indirectly opening up routes for direct negotiation between management and mill-level unions. The trend towards decentralization is reflected in the rising number of unions which have signed the industry-wide agreement: in 1972, there were only four signatories, in 1979 there were 11, and in 1992 there were 16 (Sen, 1997, p. 104).

By decentralizing bargaining structures and expanding the scope and duration of labour contracts, employers and the government are trying to minimize the "monopoly effects" of union work rules embedded in contract provisions that lead to considerable labour market inflexibilities, especially at micro-level. The determinant of public sector negotiations has been pay parity, and as a result there are few incentive structures in this system. Centralized public sector unions will have to come to terms with the microeconomic requirements of productivity growth and increased competition. They can learn from union responses to restructuring programmes in Europe. ^{(Endnote 25)}

3.2.5 Industrial conflict

Even though union density is very low by international standards, India loses more days every year as a result of strikes and lockouts than almost any other country (ILO, 1997/98, pp. 253-254). However, the number of workers involved in strikes and lockouts is considerably lower in India than in Brazil, Italy or Spain (pp. 251-252). This raises the classic question of whether conflict reflects union power or union weakness? It certainly indicates that the basic premise of industrial pluralism, the regulation of conflict, has not been achieved.

The answer to this question in India depends on whether workers resist strikes or whether employers are on the offensive during lockouts. Table 3 presents some time series conflict data. We note that during the third and fourth phase, the number of lockouts rose whereas the number of strikes decreased. The same applies to the number of workdays lost and the number of workers involved. Until 1989 industrial conflict occurred mainly in the private sector in terms of number of disputes, workdays and wages lost, and lost production, although the actual number of workers involved in disputes was higher in the public sector in the early 1980s. Since 1990, however, industrial disputes in both sectors have increased significantly (Venkataratnam, 1996).

Since the economic reforms of 1991, successive national governments have had to deal with considerable industrial conflict in the public sector, especially in banking, insurance, and transport. At regional level, some state governments have had to contend with continuing inter-union and inter-party rivalries, leading to the disruption of public life. Of late, the government has followed a tough line on striking public sector unions, as demonstrated in the dismissals and criminal proceedings against air traffic controllers. There is an imperative need for industrial relations reform in dispute settlement as the average consumer and voter has increasingly come to be a key actor and end-user of the industrial relations system. There are significant differences between the main trade union federations on the issues of secret strike ballot, prior notice to striking and the period of notification, lay-off provisions, the role of voluntary arbitration, multiple union situations, etc. The country as a whole would gain if the union movement could arrive at a consensus and if the recommendations of the Ramanujam Committee (statutory recognition of "representative" trade unions and the creation of an independent authority to arbitrate, mediate and enforce) were implemented. The institution of arbitration, if effective, fair and credible, will significantly improve the quality of industrial pluralism.

3.2.6 Labour management relations

Since the mid-1980s the practice of human resources management (HRM) has significantly altered traditional union-management relations in the advanced sectors of production, notably in multinationals and other private firms. Since the economic reforms of 1991, some public sector firms have also incorporated modern HRM practices into their otherwise traditional labour-management relationship. Some of the essential characteristics of these HRM practices are: attempts at direct communication between managers and employees; individualized and/or contingency pay systems; modular organization of production through work teams with team leaders who often form part of the management structure; carefully designed and fairly implemented performance appraisal systems; and so on. While many would argue that modern HRM practices undercut union effectiveness at enterprise level, there is no clear evidence of this in India. Unions have a strong presence in the firms where modern HRM practices are implemented successfully, and it is only with cooperative union-management behaviour that this has been possible. But this applies mainly to the manufacturing sector.

In the skill-intensive service sectors such as information technology, HRM practices continue to pose a challenge and possibly create permanent barriers to union entry and organization. To the extent that information technology redefines the nature of work in banking and insurance, unions will have little success if they resist modernization, given the entry of new, mainly foreign, players in these sites. It is no coincidence that the most publicized post-reform industrial conflict at national level has been in banking and insurance.

If labour law reforms facilitate competitive pluralism and lead to efficient collective bargaining in the private sector, and to tripartism with responsibility and accountability in the public sector, we can postulate that the effectiveness of unions would not necessarily diminish if modern HRM practices were introduced in enterprises.

3.2.7 Inter-state variations

One critical reason why labour law reform has been continually debated but ultimately shelved, is the lack of a consensus within the labour movement and across different state governments. Since economic liberalization, every state government has attempted to attract national and international capital. This has involved significant media costs, and many states have even hired international consultants to draw up economic renewal agendas. A vital consideration in this post-liberalization discourse has been the question of work ethic and comparative union militancy.

Table 11 ranks the states in terms of the number and membership of registered unions that submit returns according to the four phases of unionism. Although the data are sketchy, some trends can be discerned: (a) the more industrial states of Maharashtra and Tamil Nadu have maintained their high rankings through the four phases; (b) West Bengal ranks high during the first two phases, but we could not find comparable data for the following two phases; and (c) Gujarat is doing well during the fourth phase, no doubt reflecting the rise of unions affiliated to the Bharatiya Janata Party.

One of the most important concerns of social scientists in India today is the effect that economic liberalization will have on inter-state variations in human development, social productivity and civil society at large. States with a less organized labour movement, if controlled by pro-capital state governments, may attempt to attract capital with implicit promises of a union-free environment. This has clearly happened to some extent and has often taken violent forms. Ruling governments in other states with a long history of proletarian politics are desperately attempting to change their signals. These attempts are causing confusion within the union movement, both among the leaders and the rank and file, and also leading to chasms between political parties and their affiliated unions. In several instances, temporary or issue-based alliances have been formed between unions affiliated to opposition parties, especially with regard to privatization of public sector services and utilities.

One example of a state government changing its signals after being in power for over 20 years is the case of the CPI(M) in West Bengal. For the last 20 years, the CPI(M) and its trade union wing, the CITU, has been at the forefront of proletarian resistance to capitalist production processes and management hierarchies that control the labour process. The large-scale flight of industrial and commercial capital from West Bengal, which has been occurring since at least the mid-1960s, is attributed by many commentators to heightened industrial conflict supported by a sympathetic state government. In the meantime, the state's reputation declined, fuelled by a hostile local press. But the fact that the industrial health of the state withered during this time was there for everyone to see and experience.^{(Endnote 26)}

Since the mid-1990s, the CPI (M) has attempted to attract national and international capital at several well-publicized gatherings in Calcutta, through government funded ministerial visits abroad, and through promotional advertising. The state government has been sending strong signals to the trade unions in the state to accept the requirements of industrial regeneration. Unions and employees in hospitals and nursing homes have been warned. As an important CPI(M) minister put it recently: "In the name of trade unionism, we have allowed many things. But enough is enough. We will take immediate action if there's the slightest hint of agitation at a hospital".^{(Endnote 27)} After nurturing (some would say pampering) the trade unions in the state for so long, it is now very difficult to change attitudes, behaviour, customs and institutionalized practices to facilitate capitalist expansion, a process that was anathema to the CPI(M) for so many years. With hindsight, one could argue that if the party had thought out a strategy only a decade ago, the state could have reaped considerable returns from an effective model of social democracy. In the meantime, trade unions are now realizing that there is nothing voluntary about the 'voluntary retirement schemes' being imposed on workers in some well-known large firms.^{(Endnote 28)}

Recently, tensions between the party and its affiliated union (the CITU) came to the fore on two counts. At the Haldia Petrochemicals Limited township the union is demanding that "sons of the soil" be given permanent jobs in the Rs.51.7 billion project. According to management such a recruitment policy would spell financial disaster, and consequently, the state government has gone against the wishes of its affiliated union.^{(Endnote 29)} The second site where tensions are currently running high is the transport sector in the city of Calcutta, where due to environmental pressures the state government wants to implement Euro I and Euro II emission standards as from 1999. This would mean taking nearly 80 per cent of private buses and taxis off the road. The transport workers' unions have threatened indefinite strikes to cripple the public transport system in the city. ^{(Endnote 30)}

Calcutta exemplifies the failure of state-controlled unionism where "class consciousness" is dominant. In contrast, the market-driven industrial relations system of Mumbai exemplifies a decentralized regime where the important role played by individual trade union leaders in the city's labour history has added to the "job conscious" nature of industrial relations. Although considerable growth has taken place in Mumbai, this regime has led to almost uncontrollable and corrupt links between politics, the real estate market and the organized labour movement. It is no wonder that post-reform capital has invested in newer industrial locations, and this logic is slowly but surely changing the industrial geography of India (Shaw, 1999). It will thus become increasingly difficult, if not impossible, to generalize about an Indian labour relations system.

3.2.8 Women workers and unionization

The percentage of women members in trade unions that submit returns rose from 7.3 per cent in 1951-52 to 10.3 per cent in 1985; in 1992 it was 11.6 per cent (see table 2). Detailed information on the extent of unionization among women workers is not available, although there are rich case studies of specific sectors/industries where women workers form a substantial section of the workforce (see the chapter on tea plantations by Bhowmik in Davala, 1992).

There is considerable controversy on whether there has been an increasing "feminization" of the workforce over the last decade, especially since the liberalization process began. Deshpande and Deshpande (1992, p. 1998) assessed the short-run impact of liberalization on female employment and participation. They found that: (a) in urban areas, both male and female participation rates increased after liberalization; (b) gender-based wage differentials widened among regular wage/salaried rural and urban workers; (c) women workers were increasingly taking to self-employment and to the informal sectors as their proportion in manufacturing declined even though women's share in the urban workforce rose slightly.

Although in several countries globalization has led to feminization of the manufacturing workforce, Banerjee (1997) argues that in India the reverse has taken place: women's opportunities in the secondary sectors have fallen drastically in all states. However, there has been a slight increase in work opportunities for rural women in agriculture and some gains were made in the tertiary sector. According to Banerjee (1997, p. 433), it would be unrealistic to expect a "mechanical reproduction of international trends in a country the size of India". Women workers account for only 17 per cent of the manufacturing workforce that in turn is only 13 per cent of the total workforce. She goes on to suggest that even if the entire export sector (commodities) were staffed by women, "it is doubtful that this would result in a feminization of the Indian manufacturing workforce as a whole".

Given that a substantial section of women workers in India today are engaged as "homeworkers" in several industries (such as bangle makers, cobblers, dye makers, flower workers, kite makers, lace makers, leather workers, etc.), it is encouraging that the Indian trade union movement, under the leadership of the Self-Employed Women's Association (SEWA)^{(Endnote 31)} has taken a lead in drafting an ILO convention on homeworkers (Mukul, 1998).^{(Endnote 32)} A bill was introduced in the upper house in 1988 that attempted to equalize treatment of homeworkers with other wage earners in terms of remuneration, health and safety, minimum wage and maternity protection, with tripartite boards as the enforcement mechanism. Although the bill was dropped, it did contribute to initiating a national debate, according to Ela Bhatt, general secretary of the SEWA (Mukul, 1998, p. 758).

The SEWA model, where poor working women in the informal sector are organized so as to improve their wages and working conditions, and also assisted with credit from banks and cooperatives, needs to be replicated elsewhere in India with considerable urgency. This is already happening in the Working Women's Forum in Chennai and Annapurna in Mumbai. Established trade union federations have to take a lead in fostering these organizational models and cooperating with local NGOs where the situation warrants, especially in states where gender equality is a serious problem (Seeta Prabhu et al., 1996). The CITU has made considerable progress in organizing women workers in the informal sector.

Trade unions should lobby central and state governments to improve education for women and increase state intervention in favour of women's employment. There is also considerable scope for increasing the number of women in leadership roles within the established trade union federations.

3.2.9 Changing public perception of trade unions

The memories of nationalism and independence struggles sustained the image of trade unions as the collective voice of the oppressed, dispossessed and the exploited during the first two phases of unionism. In the third phase, segmented and uneven economic growth fractured union voices, and the public was able to distinguish between them. The Mumbai textile strike of 1982-83 generated considerable solidarity, not just within the organized labour movement, but among the urban citizenry at large. At the same time, however, impatience with the declining work ethic and a growing lack of accountability in the public sector, especially in service and education, often reached critical levels.

Since the liberalization process in 1991-92, the print and visual media have clearly supported globalization. By and large, organized labour, especially in the public sector, has not received a sympathetic press; in fact, coverage of trade union matters has declined considerably. At the same time though, considerable media attention is given to the conditions of unorganized workers, child and women workers, and to the attempts by non-governmental organizations to improve the status of disadvantaged groups. Trade unions in India today face the challenge of convincing the public that they can act on behalf of all employees, unionized or not. This requires the formation of strategic alliances with community bodies, social movements, and other non-governmental organizations. Trade unions will have to come to accept that the credibility of political parties is at a very low level. There is considerable scope for the trade union movement to capitalize on potential alliances, and a concrete beginning can be made by first forging alliances among themselves. However, this could entail a weakening of links with their political parties.

Future role of trade unions in India: Organizing the unorganized

The future role of the trade union movement is linked with a broader concern for ensuring the social cohesion of working people in a large and diverse country. In this final section, we examine union strategies in the private corporate sector, in public sector enterprises, and in the informal sector. It is imperative for the trade union movement to concentrate on organizing the unorganized, so as to create secure incomes and safe working conditions for those with irregular and precarious jobs.

4.1 The private corporate sector

On average, private enterprises employ around 30 per cent of all formal sector workers in India; in manufacturing and trade, this proportion is around 70 per cent, whereas in transport, electricity and construction the figure is less than 5 per cent (Datta Chaudhuri, 1996).

In successful private companies enterprise-based trade unions (that may or may not be politically affiliated) will have to accept that their pay is partly (if not largely) determined by productivity. Rather than blind resistance to this kind of pay structure, a cooperative strategy may pay greater dividends in terms of gainsharing at enterprise level. Unions will have to use their "collective voice" effectively in collective bargaining when incentive structures are proposed and negotiated. While the independent unions will find this strategy quite natural, those which are affiliated to the centralized federations may find it difficult. In either case, the extent to which a union is willing to take a risk will partly determine the composition of pay (performance-based "risk" pay and "steady" pay).

While the majority of contracts in this sector are (and probably always will be) negotiated at enterprise- or plant-level, unions in some organizations, possibly in the multinationals, could concentrate on attaining firm-wide agreements in the face of considerable management opposition. Firm-wide agreements will strengthen union power at the corporate-level, and to achieve this, unions may have to trade off some plant-level gains. An example of this situation is being played out at Bata India. Management recognizes the enterprise unions in its various plants across the country, but the loosely united All India Bata Employees Federation is not recognized. It appears that management is willing to talk to the federation if it agrees to restructuring plans at the plant in Faridabad. If the federation agrees to these plans in exchange for management recognition, this would clearly reduce union influence at the plant.

In the older industries in the private sector, where industry-wide bargaining is the dominant structure and where inter-firm differentiation has grown considerably since liberalization, unions and employers are finding it difficult to reach industry-level agreements. Unions will continue to face obstacles to industry-wide solidarity in this sector.

What have been the effects of economic liberalization on the connections between unions and political parties, and what has this meant for the private corporate sector? To the extent that most of the centralized trade unions continue to oppose the basic implications of economic liberalization, there has been a surprising reconciliation of unions affiliated to opposing political parties on a range of issues at both regional and national level. There has been a gap between the preoccupations of political parties and the macro-objectives of trade unions since the reforms. This has created a dilemma for most of the unions in this sector: while the loosening of ties with the parent body inevitably leads to greater autonomy in decentralized decision-making, it also means a lessening of centralized lobbying power. Market forces will increasingly dominate union strategies in this sector.

4.2 Public sector enterprises

On average, the public sector employs around 70 per cent of all formal sector workers in India; in transport, mining, construction, electricity and services this proportion is high (>80 per cent), but it is considerably lower in agriculture (40 per cent), manufacturing (<40 per cent), and trade (<35 per cent) (Datta Chaudhuri, 1996).

In non-viable public sector enterprises that are ready for closure, most of which are in the East, the situation continues to be very grim. Workers have not been paid for several months and the endless talk of revival now sounds hollow. The closure of these firms seems to be the only solution and unions can do no more than see that lay-offs are implemented fairly and as generously as possible. In several state-owned enterprises and organizations unions have accepted that privatization is the only way of saving the unit, and that informed negotiation is required.

As a result of increased competition from both domestic and international producers, the output of public enterprises and services has improved substantially. Nowhere is this more true than in the state-run airlines. But unions in the public sector, especially those in services such as medicine, education, the police and municipal workers, can substantially increase their credibility by agreeing to enforceable accountability procedures. This would mean internal monitoring, which the unions are reluctant to accept.

Although the government has indicated a preference for decentralization, the centralized bargaining structures have not yet been dismantled. Unions could campaign for a restructured central system that allows for greater local autonomy and minimizes bureaucratic inflexibilities. For the public sector to deliver long-run productivity improvements in the post-liberalization period, unions will have to partly align their objectives with those of the end-user - the average voter/consumer - who has become an important voice in the labour relations system.

4.3 The informal sector

In terms of union density, India fares rather badly compared to other large developing countries. According to the ILO World Labour Report 1997-98, union membership as a percentage of non-agricultural labour dropped from 6.6 per cent in 1985 to 5.5 per cent in 1995 (the corresponding figure in 1995 for Argentina was 23.4 per cent, Brazil 32.1 per cent and Mexico 31 per cent). Union membership as a percentage of formal sector workers in India declined from 26.5 per cent to 22.8 per cent between 1985 and 1995 (the corresponding figures in 1995 were Argentina 65.6 per cent, Brazil 66 per cent, Mexico 72.9 per cent). If the figures are derived only from registered unions that submit returns, it is possible that they may somewhat underestimate union density in India. According to the above source, less than 2 per cent of workers in the formal and informal sectors in India are covered by collective bargaining agreements. Clearly, a large proportion of workers (certainly those in the formal sector) fall within the ambit of labour legislation, even though they are not covered by a collective agreement. Nevertheless, it is apparent that considerable organization of workers remains to be undertaken in the Indian economy.

If one were to assume that the formal sector corresponds with the unionized sector (in reality, the unionized sector is a subset of the formal sector), then the following figures give an idea of the extent to which unions in future can organize workers in the various sectors. In total, less than 10 per cent of all workers are in the formal sector. The proportion of workers in this sector by industry groups is: mining and quarrying (56.9 per cent), manufacturing (19 per cent), construction (17.5 per cent), trade (2.1 per cent), transport (38.7 per cent), and services (38.7 per cent). Clearly, there is enormous potential for organizing workers in construction, manufacturing and trade. In addition, detailed surveys in several industries have found that the existing unions do not sufficiently represent the interests of casual and temporary workers (see the studies in Davala, 1992). Finally, according to National Sample Survey Organization data, there is a "high incidence of women's involvement in unorganized sector activities, ranging anywhere between 20 to 25 per cent of total employment in urban areas and anywhere between 30 to 40 per cent of total employment in rural areas - figures which far outweigh women's recorded involvement in productive activities from Census sources" (Mukhopadhyay, 1997, p. 485).

In sharp contrast to the formal sector, "the unorganized sector has little by way of protective legislation or union representation" (Anant and Sundaram, 1998, p. 833). In this site, the "not so invisible" forces of demand and supply determine wages and working conditions. There are no automatic cost-of-living adjustments and substantial improvements are required in designing need-based minimum wages for unorganized workers (Jhabvala, 1998) and providing them with assured employment for a minimum number of days (Unni, 1998). In this regard, the government's recent signal about labour law reform consisting of "umbrella legislation for welfare of unorganized sector workers" as part of the agenda should be critically examined. The government proposes to relax contract labour laws so as to generate more jobs, arguing that this would ensure better overall security and welfare provisions for unorganized workers. Unions, however, feel that any such move will only undermine permanent jobs. In any case, there are strong economic reasons why the wages and working conditions of informal sector workers should be improved through welfare legislation. Such measures improve the capabilities of the disadvantaged and vulnerable sections of working people. In the absence of enhanced capabilities the economy suffers a net loss.

It is clear that unions have a whole range of workers to organize in the coming years, since the majority of labour market entrants will probably work as self-employed or casual/temporary/ contract workers. Visaria and Minhas (1991) estimate that nearly 80 million people will join the labour force between 1990 and 2000. A whole range of non-governmental organizations have successfully organized (not necessarily unionized) several informal sector occupations and sites in India during the last decade, but it seems that these interventions are resented by the established trade union federations as an intrusion into their terrain (often, it is claimed, with financial backing from abroad). These fears are probably unwarranted, and cooperation between trade unions and NGOs is required to level up working conditions in these relatively neglected labour markets.

One way of organizing workers could be through union mergers and a joint trade union front. However, the latter presupposes a certain number of shared objectives among the large centralized trade union federations, and this unity is not yet on the agenda. The All India Trade Union Congress, the Hind Mazdoor Sabha, and the Indian National Trade Union Congress have talked of mergers and unity, but the Centre of Indian Trade Unions has taken a different approach; they have proposed a confederation of central trade unions which will preserve the individual identities (Muralidhar, 1994).The large unions have considerable differences on the efficacy of a secret ballot system to generate the legitimate bargaining agent. However, one good sign is the increasing willingness of trade union federations to work together in spite of the differences between their political parties on reform. In many states acute differences are surfacing between political parties and their affiliated unions and these issues are now being openly debated.

It is well-known that the informal sector in India contributes significantly to employment generation and to value added in industry. It is also true that there are considerable links between the formal and informal sectors and that there is a crucial regional dimension to informal sector manufacturing (Shaw, 1990; 1994). The attempt to unionize the unorganized in India has been difficult, although some progress has been made in certain regions with sympathetic state governments. But in other states, the situation of informal sector workers remains grim, and will probably get worse unless there is a concerted effort by trade unions and NGOs, hopefully with the assistance of local and state governments, to level up the labour market institutions of the informal sector. In these endeavours, the organized labour movement should not view NGOs as competitors.

To what extent will alliances between trade unions break the links with their respective political parties? Or, will such alliances lead to the emergence of national unions without explicit political affiliation? It is too early to answer these questions but the tensions between some political parties and their trade union affiliates have come onto the regional and national stage since economic liberalization began in the early 1990s. In the private sector, these tensions emerged during the third phase of unionism and resulted in the proliferation of independent unions. It may be premature to suggest that independent unions, if they can be organized at industry/regional level, will lead to far greater "collective voice" effects and less "monopoly" effects than the existing industrial relations system. With economic liberalization and its effects on regional variations in economic activity, it seems that battles over working conditions will increasingly be fought at local and regional levels. Trade unions will have to forge deep links with neighbourhoods and communities, urban movements, environmental groups and regional NGOs to enhance their effective power. Ultimately, it all depends on "public action", participation in the process of social change. Public action refers not to what the state does for the public, but to action taken by the public (Dreze and Sen, 1989). The trade union movement could trigger this much needed "public action" through broad-based alliances.

References

Ahluwalia, I.J.1991. Productivity and growth in Indian manufacturing (New Delhi, Centre for Policy Research) (mimeo).

-. 1992. Redefining the role of the state: India at the crossroads (New Delhi, Centre for Policy Research) (mimeo).

Anant, T.C.A.; Sundaram, K. 1998. " Wage policy in India: A review", in The Indian Journal of Labour Economics, Vol. 41, Oct.-Dec., pp. 815-834.

Bagchi, A.K. 1982. The political economy of underdevelopment (Cambridge, Cambridge University Press).

Bain, G.S.; Elsheikh, F. 1976. Union growth and the business cycle: An econometric analysis (Oxford, Blackwell).

Banaji, J.; Hensman, R. 1990. Beyond multinationalism: Management policy and bargaining relationships in international companies (New Delhi, Sage).

-; -. 1995. "India: Multinationals and the resistance to unionised labour", in International Union Rights, Vol. 2, No.2, pp. 5-6.

Banerjee, N. 1997. "How real is the bogey of feminization?", in The Indian Journal of Labour Economics, Vol. 40, July-Sep., pp. 427-438.

Basu, K. 1995. Do India's labour laws hurt Indian laborers? A theoretical investigation, Center for Analytical Economics, Working Paper 95-12 (Ithaca, Cornell University).

Bhaduri, A. 1993. "Orthodox development theories and their application to less developed countries", in Vaggi, G. (ed.): From debt crisis to sustainable development: Changing perspectives on north-south relationship (London, St.Martin's Press).

-; Nayyar, D. 1996. An intelligent person's guide to liberalization (New Delhi, Penguin India).

Bhagwati, J.; Srinivasan, T.N. 1975. Foreign trade regimes and economic development (New York, Columbia University Press).

Bhattacherjee, D. 1987a. Union-type, labour market structure, and bargaining outcomes in Indian manufacturing. Unpublished Ph.D. thesis. (Urbana-Champaign, University of Illinois).

-. 1987b. "Union-type effects on bargaining outcomes in Indian manufacturing", in British Journal of Industrial Relations, Vol. 27, (July), pp. 247-266.

-.1988. "Unions, state and capital in Western India: Structural determinants of the 1982 Bombay textile strike", in Southall, R. (ed.): Labour and unions in Africa and Asia: Contemporary issues (London, Macmillan), pp.211-237.

-. 1989. "Evolution of unionism and labour market structure: Case of Bombay textile mills, 1947-85", in Economic and Political Weekly, Vol. 24, (May), M67-M76.

-. 1992. "Determinants of bargaining structure in India: An exploratory exercise", in Economic and Political Weekly, Vol. 27, (May), M71-M74.

-. 1996. "Economic liberalization, democracy and industrial relations: India in a comparative perspective", in Indian Journal of Labour Economics, Vol. 39, (Oct.-Dec.), pp. 1011-1022.

-. 1997. "Labour truths", in The Telegraph, (Calcutta), 19 Sep .

-; Datta Chaudhuri, T. 1994a. "Modelling trade union growth in India", in Indian Journal of Labour Economics, Vol. 37, (Jan.-Mar.), pp. 67-72.

-; -. 1994b. "Unions, wages and labour markets in Indian industry, 1960-86", in Journal of Development Studies, Vol. 30, (Jan.), pp. 443-465.

Borjas, G. 1996. Labor economics (Singapore, McGraw-Hill).

Breman, J. 1995. "Labour, get lost: A late-capitalist manifesto", in Economic and Political Weekly, Vol. 30, (16 Sep.), pp. 2294-2300.

Buchele, R.; Christiansen, J. 1992. "Industrial relations and productivity growth: A comparative perspective", in International Contributions to Labour Studies, Vol. 2, pp. 77-97.

Chakrabarty, D. 1989. Rethinking working-class history (Princeton, Princeton University Press).

Chakraborty, S.K. 1991. Management by values: Towards cultural congruence (New Delhi, Oxford University Press).

Chatterjee, R. 1980. Unions, politics, and the state: A study of Indian labour politics (New Delhi, South Asian Publishers).

Chaudhuri, S. 1995. "Government and transnationals: New economic policies since 1991", in Economic and Political Weekly, Vol. 30, (6-13 May), pp. 999-1012.

Corbo, V.; Fisher, S. 1995. "Structural adjustment: Stabilization and policy reform - Domestic and international finance", in Behrman, J.; Srinivasan, T.N. (eds.): Handbook of development economics, Vol.3 (Amsterdam, Elsevier.).

Costabile, L. 1996. "Institutions, social custom and efficiency wage models: Alternative approaches", in Cambridge Journal of Economics, Cambridge, Vol. 19, pp. 605-623.

D'Souza, E. 1998. "World employment: ILO perspective", in Economic and Political Weekly, Vol. 33, (26 Sep.-2 Oct.), pp. 2537-2545.

Datta, R.C. 1998. "Public action, social security and unorganised sector", in Economic and Political Weekly, Vol. 33, (30 May-5 June), L2-L5.

Datta Chaudhuri, M. 1996. Labor markets as social institutions in India. IRIS India Working Paper No.10 (College Park, MD).

Datta Chaudhuri, T.; Bhattacherjee, D. 1994. "A structural model of strikes and wages in Indian industry, 1960-86", in Indian Journal of Industrial Relations, Vol. 30, (Oct.), pp. 144-155.

Davala, S. (ed.). 1992. Employment and unionization in Indian industry (New Delhi, FES).

-. 1992. "Chemical and pharmaceutical industry in Maharashtra", in Davala, S. (ed.): op.cit., pp. 191-222.

-. 1996. "Independent trade unionism in West Bengal", in Economic and Political Weekly, Vol. 31, (28 Dec.), L44-L49.

Desai, M. 1975. "India: Emerging contradictions of slow capitalist development", in Blackburn, R. (ed.): Explosion in a subcontinent (London, Penguin), pp. 11-51.

Deshpande, J.V. 1998. "College teachers' strike: An assessment", in Economic and Political Weekly, Vol. 33, (19 Sep.), pp. 2430-2432.

Deshpande, L.K. 1992. "Institutional interventions in the labour market in Bombay's manufacturing sector", in Papola, T.S.; Rodgers, G. (eds.): Labour institutions and economic development in India. Research Series No. 97 (Geneva, International Institute for Labour Studies).

Deshpande, S.; Deshpande, L.K. 1992. "New economic policy and female employment", in Economic and Political Weekly, Vol. 27, (Oct.), pp. 2248-2252.

-; -. 1996. "New economic policy and response of the labour market in India", Paper presented to a seminar on "Labour markets and industrial relations in South Asia: Emerging issues and policy options" (India International Centre, New Delhi, 18-20 Sep.)

-; -. 1998. "Impact of liberalisation on labour market in India: What do facts from NSSO's 50^th round show?", in Economic and Political Weekly, Vol. 33, (30 May), L31-L39.

Doeringer,P.; Piore, M. 1971. Internal labor markets and manpower analysis (Lexington, D.C. Heath).

Dreze, J.; Sen, A.K. 1989. Hunger and public action (Oxford, Clarendon Press).

Edgren, G. 1990. "Employment adjustment and the unions", in International Labour Review (Geneva, ILO), Vol.129, No. 5, pp. 629-648.

Fallon. P.R.; Lucas, R.E.B. 1991. "The impact of changes in job security regulations in India and Zimbabwe", in The World Bank Economic Review, Vol. 5, (Sep.), pp. 395-413.

FES-IIRA. 1996. Social clause in trade: Trade unions' perspective (New Delhi, FES).

Fonseca, A.J. 1964. Wage determination and organized labour in India (Bombay, Oxford University Press).

Freeman, R. 1992. "Labor market institutions and policies: Help or hindrance to economic development?", Proceedings of the World Bank Conference on Development Economics, pp.117-144.

-; Medoff, J. 1984. What do unions do? (New York, Basic Books).

Ghose, A.K. 1992. "Economic restructuring, employment and safety nets: A note", in Social dimensions of structural adjustment in India (New Delhi, ILO-ARTEP), pp. 94-102.

Guptan, S.U. 1992. "Engineering industry in Andhra Pradesh", in Davala, S. 1992. op.cit., pp. 113-144.

Hensman, R. 1996. "Minimum labour standards and trade agreements: An overview of the debate", in Economic and Political Weekly, Vol. 31, (20-27 Apr.), pp. 1030-1034.

Hiro, D. 1976. Inside India Today (New York, Monthly Review Press).

Holmstrom, M. 1984. Industry and inequality: The social anthropology of Indian labour (Cambridge, Cambridge University Press).

Horton, S.; Kanbur, R.; Mazumdar, D. 1991. Labor markets in an era of adjustment. Policy, Research, and External Affairs Working Paper No. 694 (Washington, D.C., The World Bank).

Illustrated Weekly of India. 1985. "State of the unions", (10-16 Feb.), pp. 18-23.

ILO. 1996-97. World employment - National policies in a global context (Geneva, ILO).

-. 1997-98. World Labour Report, Industrial relations, democracy and social stability (Geneva, ILO).

Jackson, D.A.S. 1972. "Wage policy and industrial relations in India", in Economic Journal, Vol. 82, (Mar.), pp. 183-194.

Jaitly, J. 1997. "Organising the unorganised in Kerala: Case studies of Aruvacode and Kodungallur", in Economic and Political Weekly, Vol. 32, (12 July), pp. 1729-1736.

Jhabvala, R. 1995. "Invisible workers reach international heights", in Economic and Political Weekly, Vol. 30, (9 Dec.), pp. 3133-3136.

-. 1998. "Minimum wages based on workers' needs", in Economic and Political Weekly, Vol. 33, (7-13 Mar.), pp. 500-502.

Johri, C.K. 1967. Unionism in a developing economy: A study of the interaction between trade unionism and government policy in India, 1950-65 (Bombay, Asia Publishing House).

Jose, A.V. 1992. "Earnings, employment and productivity trends in organised industries in India", in Indian Journal of Labour Economics, Vol. 35, (Jul.-Sep.).

Joseph, K.P. 1997. "Piped music and telephone attendants: Report of Fifth Pay Commission", in Economic and Political Weekly, Vol. 32, (22 Mar.), pp. 563-565.

-. 1999. "Government pay and perks: Growing chaos", in Economic and Political Weekly, Vol. 34, (15 May), p. 1167-1170.

Joshi, V.; Little, I.M.D. 1994. India: Macroeconomics and political economy, 1964-1991 (Washington, D.C., The World Bank).

-; -. 1996. India's economic reforms 1991-2001 (New Delhi, Oxford University Press).

Karnik, V.B. 1966. Indian trade unions: A survey (Bombay, Manaktalas).

Katz, H. 1993. "Decentralization of collective bargaining: A literature review and comparative analysis", in Industrial and Labor Relations Review, Vol. 47, (Oct.), pp. 3-22.

Kennedy, V.D. 1966. Unions, employers and government (Bombay, Manaktalas).

Kingdon, G.G. 1997. "Labour force participation, returns to education and sex discrimination in India", in Indian Journal of Labour Economics, Vol. 40, (Jul.-Sep.), pp. 507-526.

Madan, B.K. 1977. The real wages of industrial labour in India (New Delhi, Management Development Institute).

Mathur, A.N. 1992. Employment security and industrial restructuring in India: Separating facts from folklore - The exit policy controversy (Calcutta, Indian Institute of Management).

-. 1993. "The experience of consultation during structural adjustment in India (1990-92)", in International Labour Review (Geneva, ILO), Vol. 132, No. 3, pp. 331-345.

Mazumdar, D. 1997. "Labour markets, trade patterns and workers' living standards", in Economic and Political Weekly, Vol. 32,(1-14 Mar.), pp. 463-474.

Milgrom, P.; Roberts, J. 1992. Economics, organization and management (New Jersey, Prentice-Hall).

Mitra, A. 1977. Terms of trade and class relations (London, Frank Cass).

Morris, M.D. 1955. "Labor discipline, trade unions, and the state in India", in Journal of Political Economy, Vol. 33, (July), pp. 293-308.

-. 1965. Emergence of an industrial labor force in India (Berkeley, University of California Press).

Mukhopadhyay, S. 1997. "Locating women within informal sector hierarchies", in Indian Journal of Labour Economics, Vol. 40, (July-Sep.), pp. 483-492.

Mukul, 1998. "Homeworkers: High hopes and hard realities", in Economic and Political Weekly, Vol. 33, (4-10 Apr.), pp. 758-762.

Mundle, S. 1992. "The employment effects of stabilisation and related policy changes, 1992-1994", in Social dimensions of structural adjustment in India (New Delhi, ILO-ARTEP), pp. 25-40.

Muralidhar, S. 1993. "Slipping through the holes in the safety net", in Economic Times, (Calcutta), 26 Dec., p. 7.

-. 1994. " In tune with changing reality", in Economic Times, (Calcutta), 13 Feb., p. 9.

Myers, C.M. 1958. Labor problems in the industrialization of India (Cambridge, MA, Harvard University Press).

Nagaraj, R. 1994. "Employment and wages in manufacturing industries: Trends, hypothesis and evidence", in Economic and Political Weekly, Vol. 29, (Jan.), pp. 177-186.

-. 1997. "What has happened since 1991? Assessment of India's economic reform", in Economic and Political Weekly, Vol. 32, (8-14 Nov.), pp. 2869-2879.

Nambiar, R.G.; Mungekar, B.L.; Tadas, G.A. 1999. "Is import liberalization hurting domestic industry and employment?", in Economic and Political Weekly, Vol. 34, (13-19 Feb.), pp. 417-424.

Nayyar, D. 1981. "Industrial development in India: Growth or stagnation?", in Bagchi, A.K.; Banerjee, N. (eds.): Change and choice in Indian industry (Calcutta, K.P. Bagchi), pp. 91-118.

Nelson, J.M. 1991. " Organized labor, politics and labor market flexibility in developing countries", in World Bank Research Observer, (Washington, D.C.), Vol. 6, pp. 37-56.

Noronha, E. 1996. "Liberalisation and industrial relations", in Economic and Political Weekly, Vol. 31, (24 Feb.), L14-L20.

Olson, M. 1971. The logic of collective action (Cambridge, Harvard University Press).

Papola, T.S. 1994. "Employment growth and social protection of labour in India", in Indian Journal of Industrial Relations, Vol. 30, (Oct.), pp. 117-143.

Patankar, B. 1981. "Textile workers and Datta Samant", in Economic and Political Weekly, Vol. 16, (Dec.), pp. 1981-1982.

Patil, B.R. 1982. "Coalition and convertive bargaining", in Indian Journal of Industrial Relations, Vol. 18, (Oct.), pp. 241-262.

Patnaik, P.; Rao, S.K. 1977. "1975-76: Beginning of the end of stagnation?" in Social Scientist, Vol. 5, (Jan.), pp. 120-138.

Pendse, S. 1981. "The Datta Samant phenomenon", in Economic and Political Weekly, Vol. 16, (Apr.), pp. 695-797 and 745-749.

Punekar, S.A. 1966. "Aspects of state intervention in industrial relations in India", in Ross, A.M. (ed.): Industrial relations and economic development (London, Macmillan).

Raj, K.N. 1976. "Growth and stagnation in India's industrial development", in Economic and Political Weekly, Vol. 11, (Feb.), pp. 223-236.

Ramaswamy, E.A. 1973. "Politics and organized labour in India", in Asian Survey, Vol. 13, No. 10, pp. 914-928.

-. 1984. Power and justice (New Delhi, Oxford University Press)

-. 1988. Worker consciousness and trade union response (New Delhi, Oxford University Press).

Ramaswamy, K.V. 1999. "The search for flexibility in Indian manufacturing: New evidence on outsourcing activities", in Economic and Political Weekly, Vol. 34, (6-12 Feb.), pp. 363-368.

Rameshan, P. 1998. "Employment reduction under reforms and the productivity of sick PSEs: A learning curve analysis", in Indian Journal of Labour Economics, Vol. 41, (Jan.-Mar.), pp. 105-118.

Rudolph, L.I.; Rudolph, S.H. 1987. In pursuit of Lakshmi (Bombay, Orient Longmans).

Seeta Prabhu, K.; Sarkar, P.C.; Radha, A. 1996. "Gender-related development index for Indian states: Methodological issues", in Economic and Political Weekly, Vol. 31, (26 Oct.), WS72-WS79.

Sen, A. 1996. "Economic reforms, employment and poverty: Trends and options", in Economic and Political Weekly, Vol. 31, (Sep.), pp. 2459-2478.

Sen, R. 1992. "Jute industry", in Davala, S. 1992, op.cit., pp. 41-62.

-. 1992. "Coal industry", in Davala, S. 1992, op.cit., pp. 63-84.

-. 1997. "Changing employment and industrial relations in the jute industry", in Venkataratnam, C.S.; Verma, A. (eds.): Challenge of choice: Industrial relations in Indian industry (New Delhi, Allied), pp. 65-104.

Sen, S.K. 1977. Working class of India: History and movement, 1830-1970 (Calcutta, K.P. Bagchi).

Sengupta, A.K. 1992. Trends in industrial conflict in India (1961-87) and government policy. Working Paper Series No. 174/92 (Calcutta, Indian Institute of Management).

Shaw, A. 1990. "Linkages of large scale, small scale and informal sector industries: A study of Thana-Belapur", in Economic and Political Weekly, Vol. 25, (17 Feb.), M17-M22.

-. 1994. "The informal sector in Indian manufacturing activities: A regional study", in Indian Journal of Labour Economics, Vol. 37, (July-Sep.), pp. 335-350.

-. 1999. "Emerging patterns of urban growth in India", in Economic and Political Weekly, Vol. 34, (17-24 Apr.), pp. 969-978.

Subramaniam, D. 1997."Bangalore public sector strike, 1980-81: A critical appraisal", in Economic and Political Weekly, Vol. 32, (12 Apr.), pp. 767-778.

Tulpule, B.; Datta, R.C. 1988. "Real wages in Indian industry", in Economic and Political Weekly, Vol. 23, (Oct.), pp. 2275-2277.

-. 1989. "Real wages and productivity in industry: A disaggregated analysis", in Economic and Political Weekly, Vol. 24, (Aug.), M94-M102.

Turnbull, P.J. 1991."Trade unions and productivity: Opening the Harvard 'black boxes' ", in Journal of Labor Research, Vol. 12, No. 2, pp. 135-149.

Unni, J. 1998. "Wages and employment in the unorganised sector: Issues in wage policy", in Indian Journal of Labour Economics, Vol. 41, (Oct.-Dec.), pp. 875-892.

Vaidyanathan, A. 1977. "Constraints on growth and policy options", in Economic and Political Weekly, Vol. 12, (Sep.), pp. 2105-2109.

Van Wersch, H. 1992. Bombay Textile Strike 1982-83 (New Delhi, Oxford University Press).

Venkataratnam, C.S. 1996. "Industrial relations in India", Paper presented to a seminar on "Labour Market and Industrial Relations in South Asia: Emerging Issues and Policy Options" (New Delhi, India International Centre, 18-20 Sep.).

-. 1997. "Economic liberalization and the transformation of industrial relations policies in India", in Venkataratnam, C.S.; Verma, A. (eds.): Challenge of Choice: Industrial Relations in Indian Industry (New Delhi, Allied), pp. 12-64.

-; Verma, A. (eds.) 1997. Challenge of choice: Industrial relations in Indian industry (New Delhi, Allied).

Visaria, P.; Minhas, B.S. 1991. "Evolving an employment policy for the 1990s: What do the data tell us?", in Economic and Political Weekly, Vol. 26, (Apr.), pp. 969-979.

Weiner, M. 1962. The politics of scarcity (Chicago, University of Chicago Press).

Williamson, O. 1985. The economic institutions of capitalism: Firms, markets, and policy control (New York, New York University Press).

World Bank. 1995. World Development Report, Workers in an integrating world (Washington, D.C., World Bank).

Zeitlin, J. 1987. "From labour history to the history of industrial relations", in Economic History Review, Vol. 40, No. 2, pp. 159-184.

Appendix: Tables

Endnote 1:
Both these reports have been reviewed in India in the influential Economic and Political Weekly. See Breman (1995), Mazumdar (1997) on the World Bank report, and D'Souza (1998) on the ILO report.

Endnote 2:
There is a substantial literature on the trade union movement in India during the pre-independence years. Most works on labour history cover trade union struggles in the Bombay textile industry ( Morris, 1955; 1965) and the Calcutta jute industry (Chakrabarty, 1989). For an orthodox Marxist approach to the early working class movement in India see Sen (1977).

Endnote 3:
Statistically, there are three types of trade union in India: those that do not register and are statistically invisible, those that register but do not submit returns to the Registrar of Trade Unions on membership size, and those that register and submit returns on membership figures.

Endnote 4:
In the case of the Bombay Industrial Relations Act of 1946, which imposed an industry-wide bargaining structure on the textile industry, the INTUC rationalized the check-off system in the following way: "…the choice of a union must be a well considered long-term choice, which will be possible only on the basis of an insistence on continuing paid membership of a union. Temporary excitement and propaganda will influence the result of an election by secret ballot" (Karnik, 1966, p. 165).

Endnote 5:
V.V. Giri, a veteran trade union leader, served as Labour Minister between 1952-54. He felt strongly that compulsory adjudication discouraged genuine collective bargaining, which he tried to make statutory.

Endnote 6:
Internal labour markets are administrative units within which all decisions on pricing, compensation, promotion, recruitment, and training are determined by a set of internally-generated rules based on custom, tradition, and past practice. "Ports of entry" are usually at the lower levels, with workers' careers rising through predictable paths with assured lifetime employment (see Doeringer and Piore, 1971 for the original formulation). Wages are attached to jobs rather than to individual productivity. Internal labour markets seem to thrive during times of prosperity (when product markets are stable and growing); however with increased globalization, international competition and computerization, these markets have become expensive to maintain and reproduce.

Endnote 7:
For a complete list of all statutes governing employment and employment security see Annex 1 in Mathur (1992).

Endnote 8:
This period is also associated with industrial concentration. An official survey revealed that in 1964, seventy-five top "business houses" (the representative unit of Indian capital), controlling less than 6 per cent of all non-banking firms in the country, owned 47 per cent of total company assets. By 1969, their share had gone up to 54 per cent, with the top twenty houses registering individual gains of 55 per cent to 196 per cent (Hiro, 1976). Some argued that the nationalization of major commercial banks in 1970 strengthened the existing monopoly houses (Bagchi, 1982).

Endnote 9:
This discussion on macroeconomic trends relies heavily on Joshi and Little (1994).

Endnote 10:
According to Joshi and Little (1994: 63): "The above liberalization was not all that bold looked at from the vantage point of the 1990s, but it was certainly faster than the funeral pace of 1975/76 to 1984/85. While it can be criticized as being lopsided in some respects, it was largely pointed in the right direction from the viewpoint of efficiency and growth".

Endnote 11:
On founding this party, Samant seemed to have voiced a real shopfloor concern of the times: "We have now realized that workers should have a political party controlled by them rather than be controlled by the parties". Interview in the Illustrated Weekly of India, 10-16 February 1985).

Endnote 12:
Trade union membership data in India are notoriously unreliable, as there is a political economy to its collection and presentation. In addition, the only union membership data that are available are from those unions that submit returns. Assuming that these are the large, centralized, politically-affiliated unions, and assuming that the biases do not change significantly over the years, Bhattacherjee and Datta Chaudhuri (1994a) estimated trade union growth in India for the period 1962-84 using a modified Bain and Elsheikh (1976) type equation. Some of their findings were: increased money wages increased union membership, but not for real wages (i.e. a strong "credit" effect and a weak "threat" effect), the unemployment rate negatively affected union membership, and union density (of unions that submit returns) negatively affected membership growth. The authors interpret the latter as possibly indicating a saturation effect of conventional unionism.

Endnote 13:
For an analysis of labour contracts that uses "region" as a proxy for union and employer preferences for the type of bargaining structure in India, see Bhattacherjee (1992).

Endnote 14:
Nagaraj (1994) suggests an alternative explanation for the observed wage increases in the 1980s. Distinguishing sharply between the wage rate and earnings, and using data from the Occupational Wage Surveys, he suggests that earnings per worker increased primarily as a result of an increase in the number of workdays per employee, i.e., an intensification of the labour process. The latter to him indicates a decline in union power.

Endnote 15:
In an influential academic intervention in Indian labour history, indeed in labour history in general, Chakrabarty (1989), studying the jute mill workers in Bengal between 1890 and 1940, heavily criticized the teleological Marxist narrative of class consciousness and eventual socialist emancipation (as was dominant in Indian labour history till then). He showed how "culture" enmeshed the worker's sense of identity, and these conflicting identities, derived from language, religion and ethnicity, only further strengthened these divisions. It is this book that has forced the issues of "working class culture" on to any serious discussion of labour relations in India today. In the light of what transpires in the fourth phase, i.e. the growth of ethnic and community-based worker organizations, the intervention seems to be all the more important.

Endnote 16:
In the orthodox view, structural adjustment is defined as "a process of market-oriented reforms in policies and institutions, with the goals of restoring a sustainable balance of payments, reducing inflation, and creating conditions for sustainable growth in per capita income. Structural adjustment programmes generally start with a conventional stabilization programme, intended to restore the viability of the current account and the budget, but they are distinguished from pure stabilization programmes by the inclusion of a set of microeconomic-institutional policy reforms" (Corbo and Fisher, 1995, p. 2847; cited in Nagaraj, 1997). For a completely different view of what economic liberalization should mean to India's political economy, one that avoids the extreme positions of both the left and the right, see Bhaduri and Nayyar (1996).

Endnote 17:
I am grateful to my colleagues, Professors A. Ray and A. Sinha, for permitting me to use expressions from their unpublished annual economy reports.

Endnote 18:
To get an idea of the magnitude of the problem, consider the following: Steel Authority of India (SAIL, the public sector giant) has about 170,000 workers producing 10 million tons of steel, whereas Essar Steel (a private sector firm) has 1,700 worker producing 2 million tons of steel. In SAIL, several skilled workers have opted for the voluntary retirement scheme in search of greener pastures in the private sector (Economic and Political Weekly, 12 June 1999, p. 1469).

Endnote 19:
The Business Telegraph, (Calcutta), 28 May 1999.

Endnote 20:
In the jute industry for example, (unionized) permanent workers often employ casual workers (bhagawalas) themselves to meet production targets.

Endnote 21:
The Business Telegraph, (Calcutta), 4 March 1999. This industry-wide scheme, drawn up by the Indian Banks Association, aims to cover over 10 lakh (1 lakh=100,000) employees in 27 nationalized banks.

Endnote 22:
Five Pay Commissions have been appointed: 1948, 1957, 1970, 1985 and 1995.

Endnote 23:
According to Joseph (1999, p. 1167), "…Indian government employees enjoy several perks and privileges that are unusual and would be the envy of their colleagues in other countries. Part of this is a legacy from the days of British rule in India and part what the employees have been able to extract from government after independence". Some of these unique benefits are: commutation of pension, annual accumulation of leave, reimbursement of petrol and telephone bills.

Endnote 24:
Managers from the coal sector, when questioned about the appalling accident rates in Indian mines, often state that workers refuse to carry oxygen cylinders on their backs and demand a wage increase for doing so.

Endnote 25:
ILO, 1997-98, Ch. 5 and 6.

Endnote 26:
According to the state labour department, 17 well known industrial units have pulled down their shutters during the past year, rendering over 36,000 workers jobless. The McKinsey Report on the state called "Manifesto for a Business Revolution" has questioned the depth of the state government's commitment towards industrialization. It has said that though the new industrial policies framed by the state are commendable, prospective investors are waiting to see whether pro-investor policies would be followed. Source: The Sunday Statesman, (Calcutta), 7 February 1999.

Endnote 27:
The Telegraph, (Calcutta), 12 February 1999.

Endnote 28:
The Business Telegraph, (Calcutta), 16 March 1999.

Endnote 29:
The Telegraph, (Calcutta), 25 May 1999.

Endnote 30:
The Telegraph, (Calcutta), 5 June 1999.

Endnote 31:
"The SEWA has been organizing home-based workers in India for 20 years. SEWA members include small artisans like weavers and carpenters and piece-rate workers like garment and bidi workers. SEWA is a union of 143,000 members and sponsors cooperatives to produce alternative employment. So far it has 63 cooperatives including a cooperative bank and cooperatives providing social protection like healthcare, childcare and housing to its members" (Jhabvala, 1995).

Endnote 32:
In 1996, the ILO voted in favour of an international convention on "homework", defining the latter as "work carried out by a person in his or her home or in other premises of his or her choice, other than the workplace of the employer, for remuneration, which results in a product or service as specified by the employer, irrespective of who provides the equipment, materials or other inputs used, unless this person has the degree of autonomy and economic independence necessary to be considered an independent worker under laws, regulations or court decisions" (see Mukul,1998).