No. 58, Sept. 2014

No. 58
(Sept. 2014):

A Middle-Class India?

A Middle-Class India?

IV. The Reality: A Deformed Structure of Employment

Some eulogists of the existing pattern of development acknowledge the growth of inequality. However, they say it is an inevitable part of development. After all, the Nobel-winning economist Simon Kuznets suggested in 195482 that, as agrarian societies undergo industrialization, inequality would grow. Firstly, industry being far more productive than agriculture, there would be a growing gap between the incomes generated by the two sectors. Secondly, the inequality of incomes within industry too would be higher at first, as workers would be plentiful and cheap and there would be few social checks on capitalists’ wealth. So industrial development and overall rise in productivity would lead – at first – to a rise in inequality. But later, as industry became predominant and led to various changes in society, inequality would diminish. Inequality would follow an ‘inverted U’ path, first rising and then falling: So ran Kuznets’ tentative hypothesis.

Those who wished to justify growing inequality embraced the hypothesis enthusiastically, discarding Kuznets’ qualifying remarks and caveats. Inequality, they said, had to grow before it shrank, but it inevitably would shrink, and at a much higher level of income for all: such was the necessary and happy course of capitalism.

“Premature deindustrialization”

However, development in the underdeveloped countries has not taken the course Kuznets envisioned.83 We should recall that economic development in what became the developed countries took the form of a structural shift away from agriculture, first to industry. Industry, which was not bound by natural constraints in the same way as agriculture, enjoyed increasing returns, and hence was able to keep expanding its output and investment. The subsequent shift to services in those countries took place after industry had attained a share of 50 per cent in output. Changes in output structure were accompanied by a corresponding shift of workforce first to industry and then to services. As a result, today, “developed countries have a product structure similar to their employment structure – in each of them agriculture accounts for less than 5 per cent, both in GDP and employment, and services contribute 70 to 75 per cent and industry 25 to 30 per cent, both in GDP and employment. Inter-sectoral productivity differences are thus rather insignificant.”84

The case of developing countries – even those now classified, like the BRICS, as ‘middle-income countries’ – is markedly different. Dani Rodrik has recently coined the phrase “premature deindustrialization” for their actual pattern of growth. As he reminds us, no country has developed without manufacturing leading the way. The early industrializers (i.e., today’s developed countries) grew and developed first by industrializing, and then by moving into services: For example, manufacturing employment in  Britain, as a percentage of total employment, peaked at 45 per cent prior before it began to decline (Sweden: 33 per cent; Germany: 40 per cent; etc.).85 By contrast:

  • Manufacturing employment in the late industrializers has peaked at much lower levels. The most striking case is India, in which manufacturing peaked at a meagre 13 per cent of total employment, and trended downward thereafter.

  • Moreover, manufacturing employment in the late industrializers (e.g., Brazil, China, and India) peaked at comparatively very low levels of per capita income. Here, too, India is the most extreme of Rodrik’s cases.

  • Thirdly, Rodrik argues, in the early industrializers, the modern, high-productivity segment of the economy absorbed the traditional, low-productivity segment and unified the economy86; but in the late industrializers, dualism – the coexistence of  high- and low-productivity sectors – is actually getting strengthened, with ever more workers crowding in the low-productivity  sectors.87 (For example, it appears that while the ‘unregistered manufacturing’ sector in India accounts for 78 per cent of the employment in the manufacturing sector, its Gross Value Added per worker is less than 7 per cent of the figure for the ‘registered manufacturing’ sector.88)

Rodrik’s observations are not entirely novel: development economics, particularly its Marxist stream, has long been concerned precisely with the distortions and divergence in the development path of the Third World. The roots of these distortions lie in the legacy of colonialism and the continuing phenomenon of neocolonialism. Nevertheless, Rodrik’s observations usefully highlight the persistence of this distorted pattern in the period of rapid growth in the BRICS countries, which are touted as the new industrial powers. In his words, “structural change has become increasingly perverse: from manufacturing to services (prematurely), organized sectors to informality, modern to traditional firms, and medium size and large firms to small firms.”89

India’s labour market: the expectation...
This perversity is reflected in the actual pattern of India’s employment, admirably summed up in the recently published India Labour and Employment Report 2014 (ILER), prepared by the Institute for Human Development. What follows is largely extracted from that publication.

By conventional wisdom and by historical experience, most of the ‘pre-modern’ features characterising the labour markets of a developing economy are expected to give way to more rational, clearer and less inequitable attributes of a ‘modern’ labour market in the process of economic development. This is also expected to be accompanied by a gradual replacement of ‘traditional’ institutions and practices by modern ones. As a result, identification, description and analysis of concepts and categories are also expected to get crystallized and become easier and less ambiguous.

Thus the transformation of an economy from predominantly subsistence-oriented economic activities, agriculture and informal modes of work and production organizations to one with its major part in the capitalist, industrial and formal sectors is likely to lead to a change in labour categories and employment structure. In this process, there are changes not only in the shares of different sectors in employment, but also in the modes of work and employment status of workers.

Workers would increasingly become either fully employed or fully unemployed, and the extent of under-employment would decline as the economy develops. More and more workers would work as wage and salary earners, the proportion of the self-employed would decline. Most workers would be employed on a formal basis; informal enterprises would increasingly become formal and formal enterprises would employ most workers as regular employees. Thus most workers would have formal time-based contracts, clear rules and procedures to govern conditions of work and employment tenure. Employer and employee identification would be unambiguous and employer-employee relationships well-defined. One would also expect an increase in the degree of unionization and greater role of unions and collective bargaining in industrial relations. The labour market would be unified, and wage differentials would be determined mainly by economic factors.

In other words, labour markets are expected to become more rule-based and labour institutions more visible and effective, as the economy makes a transition from the subsistence, agriculture-dominated and informal character to the capitalist, industrial and formal status.

...and the reality
However, in India we find:

  • The share of self-employment has declined extremely slowly, from 61 per cent in 1972-73 to 52 per cent four decades later, in 2011-12; even in urban areas it is 42 per cent in 2011-12.

  • The proportion of regular wage/salary earners in total employment has hardly grown – from 15.4 per cent in 1972-73, it has grown to 17.9 per cent in 2011-12.

  • What has grown is the proportion of casual workers in total employment, from 23.3 per cent in 1972-73 to 29.9 per cent in 2011-12. Thus there is a partial shift from self-employment to casual work, but not to regular work.

  • The share of the organized sector in total employment reached 17 per cent by the mid-1970s; it declined thereafter to about 12 per cent by 2004-05 before returning by 2011-12 to 17.9 per cent.

  • More importantly, within the organized sector, the share of informal workers (those without employment stability and social security, such as ‘casual’, ‘contract’, ‘apprentice’, etc.) rose from 41 per cent in 1999-2000 to 58 per cent in 2011-12. That is, the majority of the organized sector workforce consists of informal workers.
  • Indeed, practically all new employment in the post-reform period, particularly from the late 1990s till 2004-05, has been either in the unorganized sector or in the form of informal jobs in the organized sector. Thus there has been a continuous deterioration in the overall quality of employment.

  • While the share of agriculture in GDP has declined by 45 percentage points between 1950 and 2011-12, its share in employment has declined by only 24 percentage points over that period.

  • The subsistence mode and the unorganized nature of production of goods and services prevail. Out of the 472 million workers in 2011-12, about 392 million or 83 per cent were estimated to be working in the unorganized sector (defined to comprise all privately-owned unincorporated enterprises employing less than 10 workers). Even the organized sector, which comprises 80 million workers, had 46 million workers (58 per cent) in the informal category of employment. In this way informal workers account for about 92 per cent of the total workers. Employment, therefore, is of low quality, devoid of the characteristics of decent employment.

Structural changes in employment have thus not been in line with either what is postulated by the conventional theory of development or was experienced historically by today’s developed countries. India stands apart, to some extent, from the experiences of even the developing countries of East and Southeast Asia. It seems to have jumped from the first stage (agricultural) to the third stage of a service-dominated economy, without an intermediate industrial phase. This has been accompanied by large asymmetries between GDP and employment shares in agricultural and services sectors.

Table 6 below covers the period of post-1991 ‘liberalization’. There has been significant decline in the employment share of agriculture. However, it remains the single largest sector. There has been a small increase in the employment share of manufacturing, which remains exceedingly low. Some service industries – trade, hotels and restaurants, transport, storage, telecommunications, finance, real estate, business services – have increased their employment share by a combined 7.3 percentage points. These are sectors which would include a good number of ‘middle class’ jobs, but also include large numbers of insecure, low-wage jobs. At the same time, another source of middle class jobs, namely, ‘community, social and personal services’ (largely public sector, and therefore relatively secure employment) saw its share fall by 1.2 percentage points. Only one sector has seen a dramatic increase, namely, construction. This is a particularly sweated sector, in which nearly the entire workforce is informal and insecure, and there is a near-complete absence of trade union organization. 

Table 6: Share of Employment by Industries 1993-94/2011-12





Agriculture & allied activities




Mining & quarrying








Electricity, gas & water supply








Trade, hotels & restaurants




Transport, storage & communications




Finance, real estate, business services




Community, social & personal services




Employment by Usual Principal & Subsidiary Status, based on National Sample Surveys for the relevant years.
India Labour & Employment Report, 2014, p.208.

Employed and poor
The ILER points out that low earnings from employment, rather than open unemployment, are the main source of poverty. This disconnect between work and income is primarily caused by the predominance of informal employment. Furthermore, the distinction between the employed and the unemployed is also often unclear; a large number of Indian workers cannot really be classified exclusively in either of the two categories. Many of them are employed for part of the year and unemployed for another.

That under-employment is a much larger problem than unemployment in developing countries like India has been widely recognized by economists in India and abroad. It was, however, postulated to decline with economic development. Development was expected to have been accompanied by changes in the technology and organization of production that would bring about the employment of most people on a full-time basis. It would lead to a situation where those working would be fully employed and those not working would be fully unemployed.

Such change does not seem to have taken place. As is typical in poor, agrarian, developing economies, open unemployment rates in India are low, irrespective of the methodology used to measure it. The poor cannot afford open unemployment, and do eke out an existence through a multiplicity of subsistence activities.

Finally, not only do shortages and surpluses of labour co-exist, wage differentials which cannot be explained by economic factors persist and even increase despite increasing occupational and geographical mobility. The lack of information and increasing occupational segmentation prevent movement towards ‘market clearing’ of any sort. Labour is segmented not only geographically, but by caste, community and gender. Among those officially defined as ‘poor’, there is a higher proportion of Scheduled Tribes (STs), Scheduled Castes (SCs), Other Backward Classes (OBCs), and Muslims. That is, historically disadvantaged castes and communities continue to suffer from gross inequalities.

The largest proportion of regular formal jobs (which are considered the ‘best quality’ jobs) both in the public and private sector are appropriated by upper caste Hindus. Such slight improvement as has taken place is in the share of regular formal jobs held by OBCs. The proportion of SCs and STs in regular formal work in the public sector is much higher than in the private sector90; but, with neoliberal policies, the share of public sector employment in total employment is declining. Thus one finds “an increased importance of social aspects in the economic dimensions of inequality during the period of market liberalization and economic reforms in the country.... class and caste inequalities have been on the rise....”91 The available data indicate that, even when the socially oppressed castes manage to get college degrees, this does not necessarily translate into better jobs and higher incomes.92

As the ILER sums it up, there is in effect one labour market for the poor and another for the better-off, and there is little traffic between the two.

The condition of the petty bourgeoisie

One conclusion we may draw from the above observations of the ILER is that the ‘middle class’ of Marx and Engels’ Manifesto, that is, the petty bourgeoisie – small tradespeople, shopkeepers, handicraftsmen (which can be extended, today, to petty manufacturers) and peasants – survive in vast numbers in India, even as they do not thrive. They possess some meagre means of production of their own, but their conditions of work and levels of income are generally worse – poorer and more insecure – than those of formal workers in organized industry. The peculiar place of the petty bourgeoisie in the social organization of labour, as both labourers and owners of means of production, delays the development of their class consciousness. (Muslims, who are under-represented in the organized sector workforce, and are over-represented among the self-employed, are among the poorest communities, with incomes comparable to Dalits and Adivasis.93)

Unable to find employment in organized industry, the vast majority of the petty bourgeoisie eke out a subsistence in the prison of the low-productivity activities to which they have access. Thus they fit neither the Economist’s description (cited earlier) of the middle classes as a “motor of economic growth” nor other economists’ description of them as an “aspirational” new consumerist class.

This can be illustrated with a few data from the National Sample Survey’s reports on unorganized sector enterprises (in manufacturing, trade, and other services, but excluding construction).94

The NSS estimated that there were 58 million such enterprises in 2010-11, employing 108 million workers: i.e., less than two workers per enterprise. Of these ‘enterprises’, the great majority (49 million) were ‘Own Account Enterprises’, with no hired workers (i.e., just family members); even the remainder employed only an average of 4.3 workers per enterprise (family members + wage workers).

These units are starved of capital. In the year preceding the survey, the owners added the following paltry sums to their fixed assets other than land and buildings: plant and machinery: Rs 701; transport equipment: Rs 1,064; tools and other fixed assets: Rs 838; software and hardware: Rs 11; information & communication technology equipment: Rs 122; capital work in progress: Rs 55. Equally, they face a problem of demand, as most of their customers too are poor peasants and workers. The meagre surpluses of the single largest group of consumers, the peasantry, are drained away in land rent, usurious interest payments, traders’ margins, high input costs, and extortions by Government officials. The survey found that 51 per cent of the enterprises were either stagnating or contracting over the previous three years.

Of course these are merely averages, and a disaggregated picture would no doubt show that some among these 58 million are prosperous enterprises. There are well-off self-employed shopkeepers, doctors, chartered accountants, lawyers, consultants, engineers, beauticians, tuition providers, yoga instructors and the like. However, the very fact that the averages are so low tells us of the condition of the vast majority of petty proprietors.

The value added per worker in these enterprises was a meagre Rs 58,000 per year, just 7 per cent of the figure for the organized factory sector.95 Indeed, the unorganized sector enterprise value added per worker is just 67 per cent of the average wages of a worker in the organized factory sector. The condition of the owners of these unorganized sector enterprises was not much better than that of their own hired workers, since the salaries paid to the workers were almost as large as the value added. The value added per worker in these unorganized sector enterprises is also only slightly higher than the corresponding figure for agriculture and allied activities (around Rs 55,000 for 2010-1196). In fact, in rural unorganized sector enterprises, the value added per worker is only Rs 37,000, i.e., much below the corresponding figure for agriculture . In a celebrated paper of 195497, the economist W. Arthur Lewis predicted that in countries such as India, the ‘capitalist’ economy (broadly identified with industry) would grow, absorb the workers from the ‘subsistence’ economy (broadly identified with agriculture), and finally eliminate dualism. Since the share of workforce in India’s agriculture has been falling, albeit slowly, in recent years, and since the latest (2011-12) data at last show a fall in the absolute numbers employed in agriculture, some economists claim that the economy has finally reached a ‘Lewisian turning-point’. However, the discussion in the earlier paragraphs makes it clear that the ‘capitalist’ economy (modern industry or services) is not ‘pulling’ workers out of agriculture; rather, the crisis in the agrarian sector is so acute that it is pushing workers out, and they are forced to seek a living in low-productivity unorganized ‘industry’ or ‘services’, no matter how abysmally low the income.

Hidden reserve army
These workers, spread over different low productivity sectors, are part of the reserve army of labour. In developed countries, the reserve army is composed of openly unemployed workers; the capitalist is able to discipline workers and keep their wages from encroaching on his profits because there are unemployed workers ready to take the place of the employed. In India, however, most of the reserve army is hidden within the officially ‘employed’ category.

Organized (formal) sector

Formal workers

Unorganized (informal) sector

Informal workers

We can view the relation between the unorganized (also known as informal), sector and the organized (also known as formal) sector as the above matrix. Here the unorganized sector is defined as all privately-owned unincorporated enterprises employing less than 10 workers. Informal workers are workers without employment stability and social security, no matter whether they work in the formal or informal sectors. (To avoid confusion: we use below the terms ‘organized/unorganized’ for sectors, and ‘formal/informal’ for workers.)

(i) The organized sector produces goods and infrastructural services (e.g., electricity, petroleum products, transport, communications) consumed by formal workers, informal workers, and the unorganized sector. However, the production of the organized sector mainly (and increasingly) caters to demand from the upper income segments. This is seen strikingly in the emergence of the automobile industry as one of the largest organized sector industries, and air conditioning as one of the major consumers of electricity. Given the backward production conditions of the unorganized sector, the organized sector is not a major source of inputs for it. For example, units that did not even use electricity for production purposes accounted for the majority of informal sector workers in manufacturing.98

(ii) The organized sector employs both formal workers and informal workers; indeed, as we noted above, it now employs more informal workers than formal workers.99 Not only does this practice allow them to keep their wage costs down directly; it also helps them to do so indirectly, because formal workers know that they can be replaced with informal workers, and so curtail their demands. (It is this trend which the new labour ‘reform’ aims to carry further.)

(iii) A segment of unorganized sector production is directly integrated with certain activities of the organized sector. For example, bidis are manufactured in the unorganized sector, often in households, but are branded and sold through organized sector firms. A share of textiles and readymade garments are similarly produced in the unorganized sector and sold through the organized sector. (Data on unorganized sector manufacturing units show that 20 per cent of them do at least some work on a contract basis.) Small retail outlets distribute the goods of organized sector firms. And so on.   

(iv) The unorganized sector also produces goods and services for consumption by both formal and informal workers. Unorganized manufacturing units make food products, beverages, tobacco products, textiles, apparel, leather products, wood products, paper products, vessels, pots, and furniture. Small restaurants and roadside stalls prepare food; tutors give workers’ children tuitions; and so on. These unbranded, ‘downmarket’ products are sold at prices much lower than the comparable organized sector products. Informal workers are also employed as domestic workers, drivers, child care, and so on in the homes of the upper layer of formal workers. By keeping down the cost of subsistence of workers, the unorganized sector effectively enables the organized sector to pay its workers lower wages.100

Since the organized sector in India is integrated with global Capital, this resultant distorted pattern of development in India, and its associated hidden reserve army of labour, expand the surpluses accruing to global Capital. Given the character of large industry in India, it can make large profits without transforming Indian society; rather it stands to gain by preventing any such transformation.

The purpose of this long excursion has been to highlight the absence of any fundamental transformation in the social organization of labour in India. The partial changes that have taken place have merely redistributed some workers among the overcrowded low-productivity sectors (e.g., from agriculture to self-employed manufacturing and low-paid services).

Industrial revolution and social revolution
History tells us that thoroughgoing industrialization, involving a major shift of workforce to industry, is linked with social revolution, whether capitalist or proletarian. A critical part of those social revolutions has been the sweeping away of the old agrarian order, the creation of an internal market, and the establishment of the political power of a dynamic native class determined to bring about national development.

Indeed, Kuznets, writing in the immediate wake of the Chinese revolution, feared that such an outcome might seem attractive to people in the underdeveloped countries. He thus warned against “repetition of past patterns of the now developed countries, patterns that, under the markedly different conditions of the presently underdeveloped countries, are almost bound to put a strain on the existing social and economic institutions and eventuate in revolutionary explosions and authoritarian [read: Communist] regimes”.

Can the political framework of the underdeveloped societies withstand the strain which further widening of income inequality is likely to generate? This query is pertinent if it is realized that the real per capita income level of many underdeveloped societies today is lower than the per capita income level of the presently developed societies before their initial phases of industrialization. And yet the stresses of the dislocations incident to early phases of industrialization in the developed countries were sufficiently acute to strain the political and social fabric of society, force major political reforms, and sometimes result in civil war.... If, for many groups in society, the rise is even partly offset by a decline in their proportional share in total product; if, consequently, it is accompanied by widening of income inequality, the resulting pressures and conflicts may necessitate drastic changes in social and political organization.

Hence Kuznets warned against thinking that

completely free markets, lack of... progressive taxation, and the like are indispensable for the economic growth of the now underdeveloped countries. Under present conditions the results may be quite the opposite – withdrawal of accumulated assets to relatively ‘safe’ channels, either by flight abroad or into real estate; and the inability of governments to serve as basic agents in the kind of capital formation that is indispensable to economic growth. It is dangerous to argue that, because in the past foreign investment provided capital resources to spark satisfactory economic growth in some of the smaller European countries or in Europe’s descendants across the seas, similar effects can be expected today if only the underdeveloped countries can be convinced of the need of a ‘favorable climate.’

Today, the policies against which Kuznets warned, and the alarming inequality which he feared would result, prevail. And this, without even the thoroughgoing industrialization and consequent shift of employment structure he assumed would take place. This seems indeed a social mix even more explosive than he envisioned. However, due to the changed political conditions today (the absence of the earlier socialist regimes, and the current weakness of revolutionary political forces), social explosions are as yet not taking the revolutionary form Kuznets feared, but remain confined to mere reordering within the existing system.




82 .Simon Kuznets, “Economic Growth and Income Inequality”, The American Economic Review, March 1955. (back)

83. Thomas Piketty, in his celebrated Capital in the Twenty-First Century has shown that Kuznets’ hypothesis does not hold even for the developed countries, but here we address a specific feature of underdeveloped countries’ economic growth. (back)

84. India Labour and Employment Report 2014: Workers in the Era of Globalization (ILER), Institute for Human Development, 2014, p. 34. (back)

85. Similarly, Sweden peaked at 33 per cent in the 1960s, and Germany at 40 per cent in 1970. (back)

86. This is no doubt an overstatement, since it ignores the fact of large emigration from Europe. (back)

87. Dani Rodrik, “The Perils of Premature Deindustrialization”, Project Syndicate, 11/10/13; “The Growing Divide within Developing Economies”, Project Syndicate, 11/4/14; “On Premature Deindustrialization”, Within his theoretical frame, Rodrik’s solution is for the State to aggressively promote large industry with steps such as deregulation, tax incentives, special economic zones, and currency devaluation. (back)

88. Sources: Gross Value Added per worker, and number of workers, in registered manufacturing from Annual Survey of Industries, 2011-12; manufacturing GDP (registered + unregistered) from National Accounts; manufacturing workforce (registered + unregistered) from National Sample Survey. We have subtracted the value added and workforce of the registered sector from the total for the manufacturing sector in order to get the figures for the unregistered sector. ‘Registered’ manufacturing units are those with 10 or more workers using power, or 20 or more workers without the use of power. (back)

89. Rodrik, “The Growing Divide within Developing Economies.” (back)

90. In the public sector, SCs make up 16.3 per cent of the regular formal workers, and STs 8.1 per cent; but in the private sector, the percentages are 10.2 and 2.2 percent. ILER, p. 80. (back)

91. Wei Zhong, Vamsi Vakulabharanam,Sanjay Ruparelia, Yan Ming, Ashwini Deshpande, Lopamudra Banerjee, “Wealth Inequality: China and India”, 2010, (back)

92. ILER, p. 85. (back)

93. Social, Economic and Educational Status of the Muslim Community of India: A Report (Sachar Committee Report), 2006, pp. 91-100, 153-54. (back)

94. National Sample Survey, Report No.s 546: Operational Characteristics of Unincorporated Non-agricultural Enterprises (Excluding Construction) in India, and 549: Economic Characteristics of Unincorporated Non-agricultural Enterprises (Excluding Construction) in India, NSS 67th Round, 2010-11. (back)

95. The figures cited are for Gross Value Added. The Gross Value Added per worker in the organized factory sector was Rs 8.33 lakh in 2010-11, according to the Annual Survey of Industries, 2010-11. We have rounded off all the value added figures to the nearest thousand. (back)

96. Taking employment in agriculture for 2010-11 at the average of 2009-10 and 2011-12, for which there are data. (back)

97. “Economic Development with Unlimited Supplies of Labour”, Manchester School, 1954. (back)

98. National Sample Survey, Report No. 557: Informal Sector and Conditions of Employment in India, NSS 68th Round, 2011-12. (back)

99. Further, the organized financial sector also engages a large number of commission-based agents to gather deposits and premiums from vast numbers of people. (back)

100. This role is similar to that of the unpaid labour of women in household work, child-bearing, and child-rearing, which keeps down the cost of reproduction of labour power, by feeding and taking care of the labourer, and rearing a new generation of labourers. (back)


All material © copyright 2015 by Research Unit for Political Economy