India. Craig Jeffrey

over time brings about institutional change (Acemoglu et al. 2005). This helps to explain what we mean when we say that economic growth is subject to circular and cumulative causation. There is an important feedback relationship between the pattern of growth and its political drivers, but this also sets a path of change from which it is difficult to break away.

      An example of this from India is that policy decisions taken in the 1950s, the outcome of the relationships that then prevailed between political and economic elites, set the country on a course of industrial development that has favoured capital-intensive industry in spite of India’s apparent comparative advantage, given an abundance of cheap labour power, in labour-intensive industry. In 2017 Arvind Panagariya, then the outgoing Deputy Chairman of the Niti Aayog, the body set up in 2014 to replace the old Planning Commission, lamented what he referred to as the ‘Brahminical mind-set’ of India’s big business groups, because of their evident preference for capital-intensive sectors rather than for investing in labour-intensive manufacturing. But in his remarks Panagariya also referred to the significance of the reservations that were instituted early in the history of independent India, restricting the production of many basic consumer goods to small-scale industrial units, so keeping out the big companies: ‘We have not been able to shake out of these historical legacies. In a natural way, the future structure of the economy gets a bit driven by the current structure. And our current structure has been determined by … the old licence permit, small-scale reservation period’ (reported in The Hindu, 13 August 2017).

This history accounts for the pronounced dualism that distinguishes Indian manufacturing, and the economy as a whole, given the size and significance of the unorganized sector (or ‘informal sector’, as it is sometimes described). This is estimated to account for 58 per cent of all domestic product, and 45 per cent of non-farm domestic product; and for 83 per cent of all non-farm employment (Kotwal et al. 2011). In manufacturing there is, it is argued, a ‘missing middle’. In India, according to data from the Asian Development Bank (Joshi 2017: table 5.3), 84 per cent of all manufacturing employment is in micro-sized and small units, employing up to 49 workers, and only 5.5 per cent in medium-sized firms (employing 50–199 workers). By comparison, in East Asian economies, the distribution is much more even across firm sizes (or in China it is weighted towards the largest units). This matters because there is a close relationship between firm size, labour productivity and wages, and it is argued that in India growth is held back by the high proportion of workers in very small firms. Generally, successful manufacturing of consumer goods such as garments and shoes, for export, is carried on in factories employing large numbers. And, as Joshi has it, in common with Panagariya, ‘In India, one of the obstacles to the formation of such companies was the reservation of hundreds of labour-intensive products for exclusive production by small firms. Small-scale industry reservations have been drastically pruned in recent years. But many obstacles to the growth of labour-intensive activities still remain’ (Joshi 2017: 72). These include constraints on credit supply: even medium-sized firms experience difficulty in securing credit (Kotwal et al. 2011).

      Research shows, therefore, that institutions (which include regulations such as those restricting the production of a range of commodities to small industrial units, as well as, for example, those that define property rights) exert an important influence on patterns and rates of economic development. But we also have to recognize that there may be a big gap between what institutions lay down, and what actually happens. Pritchett and Werker, for instance, refer to the rules regarding the issue of a driving licence in Delhi. These state that everyone should take a driving test, but research showed that only 12 per cent of those who had employed a tout, an intermediary, ended up taking the test. The touts, for a ‘fee’, found ways around the rules. In practice, the two authors argue, ‘in a regime of weak capability for policy implementation – that is, weak capability … to enforce rules – the actual practice is “deals”, and there are ubiquitous and widespread deviations of actual practice from “rules”, that create winners and losers’ (2012: 39). Rules, if they are properly implemented, are impersonal, while the term ‘deal’ refers to actions that are the result of the characteristics or the actions of specific individuals, and that are ‘one-off’. They have no implications for future transactions between other individuals. Deals may be ‘open’ or ‘closed’. In the latter case the possibility of a deal depends upon the identities of those concerned; in the former they depend only upon the actions of the agents involved. And deals may be ‘ordered’ – they will be honoured once negotiated – or ‘disordered’, when there is no certainty that the arrangement will be delivered upon. We may surmise that a movement from disordered to ordered deals is likely to be conducive to growth – investors will be more likely, for example, to invest in the capital equipment that is crucial for growth but that involves a large sunk cost; and that a shift from closed to open deals may be conducive to sustained growth, because it implies movement away from cronyism and increased competition. But there is no necessary, linear movement from closed, disordered deals to open and ordered ones. It will depend upon the settlements made between economic and political elites.

Kar and Sen (2016) have elaborated upon these arguments, in an analysis of the political economy of India’s growth episodes. The two authors also refer to the significance of the differences between various economic agents, distinguishing between ‘rentiers’, those who operate in export-oriented industries from which they can secure high rents (for instance, from having monopolistic control over a particular natural resource); ‘power-brokers’, those who also operate in sectors (like real estate, construction, utilities and telecommunications) in which high rents can be secured, but who serve domestic markets; ‘workhorses’, who operate in competitive industries (including smallholder agriculture, and informal manufacturing and services) supplying domestic markets; and ‘magicians’, who work in competitive but export-oriented sectors such as the garments industry, or the supply of IT services. Kar and Sen argue that the interactions of these actors, their relations with political elites, and the sorts of deals in which they are engaged have exercised a very important influence on trends in growth in India.

2.2 India’s History of Economic Growth

      Joshi’s broad overview (2017: ch. 2) of the history of economic development of independent India, showing that it falls into two halves, with the break coming around 1980 (see table 2.1), reflects a common understanding – though some think that the break came a little earlier, in the late 1970s. Kotwal et al. (2011) confirm that between 1951 and 1979, average growth rates remained below 4 per cent, and they say that the departure from the trend line after 1980 is both clearly visible in the data, and demonstrated in econometric testing. If an adjustment is made, however, for the sharp decline in GDP that took place in 1979–80, then the break is seen to have occurred in 1975–76. What is also noteworthy is that the variance in the growth rate declined sharply: from 15.8 in the 1970s, to 4.6 in the 1980s, and to 1.5 in the 1990s.

Bosworth and Collins (2015) have calculated a set of growth accounts for India, based on data from the Indian national accounts, in order to assess the contributions to growth of changes in inputs of labour and capital, and of TFP. They estimate (see table 2.2) that over the twenty-year period from 1960 to 1980 output increased, on average, by 3.4 per cent per year; output per worker increased by 1.3 per cent; and the contribution of TFP was a negligible 0.2 per cent. But then, over the thirty years 1980–2010, output increased by 6.2 per cent per annum; output per worker by 4.7 per cent; and the contribution of TFP was 2.5 per cent. Their data also show (table 2.2) a marked step up in rates of growth of output, output per worker and in TFP in the decade of the 1980s by comparison with the previous two decades; a small increase in output growth per year but also a drop in the growth of TFP in the decade of the 1990s; and then a very marked increase in all three variables in the 2000s.

      What seems to have underlain the improvement in TFP from the 1980s onwards – taking place especially in services (see Bosworth and Collins 2015, table 1) – was a sharp increase in private investment in equipment. This, in turn, appears