The impact of caste on economic mobility in India
The caste system is arguably the most distinctive feature of Indian society. The Indian population is divided into four hierarchical classes, or varnas, with a large sub-population of untouchables excluded entirely from the system. Within each of these classes, and among the untouchables, are thousands of castes, or jatis. The central rule in Hindu society is that individuals must marry within their own caste. Recent genetic evidence indicates that this rule has been followed for over 2,000 years. Spatial segregation on caste lines within the village results in a high degree of local social connectedness, with caste clusters in distant villages and select urban locations linked to each other through ties of marriage over many generations. This unique social structure has remained in place 70 years after Independence, in one of the world’s most dynamic economies.
Convergence between upper and lower castes in education and jobs
Economic mobility is a prerequisite for development. Given the continuing importance of caste in Indian society, an obvious question to ask is how occupational and spatial mobility in the Indian economy has been shaped by the caste system. The first thought is that the exploitation, prejudice, and discrimination that are associated with the hierarchical aspect of the caste system would have stifled mobility among the lower castes. It is certainly true that the lower castes remained locked in unskilled, low-paying occupations for centuries in the traditional economy. There is also evidence of continuing discrimination in the labour market, although this appears to be statistical—that is, employers use caste as a proxy for unobserved socioeconomic characteristics—rather than prejudice. Despite these obstacles, evidence from surveys of nationally representative samples indicates that there has been convergence between the upper castes and the lower castes on education and occupations over the past decades.
Some of this convergence may be due to the affirmative action policy that has been in place since Independence, reserving seats in institutions of higher education and the central government for former untouchables and other disadvantaged groups. Another force driving convergence could be the caste-based networks that facilitate economic activity and support the mobility of their members in an economy where markets function imperfectly. Individual members of a caste can be severely sanctioned if they renege on their obligations because information flows smoothly within the caste and because they are tied to their community in many different ways. This allows high levels of cooperation to be sustained. Numerous historical accounts document the important role played by castes in supporting the rural-urban migration that accompanied British rule and the growth of cities in the 19th century. Particular castes found particular niches in the urban labour market, and once networks in the city were established, they supported the movement of fresh migrants from the hinterland, often over the course of many generations.
Economic transitions as drivers of convergence
Structural change has created new economic opportunities over the past 25 years, but it has also brought new challenges. In particular, market imperfections, which give rise to networks, can be exacerbated in a dynamic economy. Harish Damodaran’s fascinating book (goo.gl/mBCBwa) on Indian entrepreneurs documents the movement of castes from agriculture and administrative occupations into business in recent decades. My own research on the diamond industry shows how a historically disadvantaged caste took advantage of a shock to the world supply of rough diamonds in the late 1970s to move from agriculture and then industrial labour into the export business, over the course of a single generation. The encouraging theoretical and empirical finding that emerges from this research is that once networks form, they will strengthen relatively rapidly in historically disadvantaged castes (with weaker outside options). The occupational convergence between upper castes and lower castes that has been previously documented may well have been driven by many such transitions, supported by underlying caste networks, especially among the lower castes.
Caste networks can also be a hindrance to mobility
The picture I have painted of the caste networks up to this point is entirely positive. However, these informal institutions have limitations of their own. The same networks that can be so effective in supporting the movement of groups of individuals across space and occupations can also restrict the mobility of individual members once they are established. Mark Rosenzweig and I studied schooling choice (goo.gl/S7y7Se), which is a strong predictor of future occupations, in Mumbai’s dynamic economy over the 20-year period between 1980 and 2000. Particular castes historically occupied niches in Mumbai’s mills and factories with the support of their networks. When the Indian economy restructured in the early 1990s, shifting economic activity in Mumbai from manufacturing to services, these networks had been in place for over a century. We provide evidence, based on the schooling choices of the children, that these blue-collar networks turned out to be a hindrance in this economy, keeping their members in the traditional (now less remunerative) occupations and preventing them from taking advantage of the new opportunities that became available.
Caste networks can have other unintended consequences for mobility. In a recent paper (goo.gl/gt1eEF), my co-author and I show that rural mutual insurance networks, which have smoothed consumption within castes for centuries, can restrict the (permanent) migration of individual members to the city. These networks are based on reciprocity. When a household suffers a negative income shock, it receives monetary transfers from caste members that allow it to consume at its customary level. In the future, it is expected to provide transfers to other households when they receive a negative shock.
A household with migrants will be less insured by its rural network for two reasons. First, it cannot credibly commit to reciprocate at the same level as households based entirely in the village because social sanctions against it will be less effective; it can always fall back on its new base in the city. The urban component of its income is also unobserved by the rural network. If the consequent loss in insurance is sufficiently large, then rural households could forego substantial gains in income from migration and keep all their members in the village. We use this argument to explain why rural-urban migration is unusually low in India, despite the presence of large rural-urban wage gaps. The restriction on mobility that we document, which leads to inefficiency in the labour market, arises because of inefficiency in another market; that is, because formal substitutes for the rural insurance network, such as private credit or government safety nets, are unavailable.
Community-based networks are active in all developing countries where markets are functioning imperfectly. However, these networks are exceptional with respect to their size and scope in India, because of the special caste-based structure of its society. Caste networks thus play an unusually important role in shaping economic mobility in the Indian economy. As discussed above, whether these networks support or hinder mobility will depend on the circumstances. In general, networks are effective in supporting the movement of groups, but they can restrict the mobility of individuals trying to follow a path of their own. The caste networks will disappear when the market economy starts to function efficiently. In the interim, which could be many decades, policies aimed at fostering growth would be more effective if they took account of the underlying caste networks that continue to shape educational, occupational, and locational choices in the Indian economy.
Kaivan Munshi is Frank Ramsey Professor of Economics at the University of Cambridge.
(Published with permission from Vox Dev (goo.gl/csiVhN), a policy portal for development economics.)