By The World Bank
Inequality may be greater in India than often thought. It affects poor and rich communities alike. Although comparisons based on consumption data have been used to argue that inequality in India is low by international standards, the research shows that when income, rather than NSS consumption data, is used, inequality in India appears to be in the same league as that in Brazil and South Africa, both high-inequality countries (figure 14). Why the gap between India’s consumption and income Gini measures of inequality is so large remains to be explained, but this finding at a minimum casts doubt on the often-rehearsed notion that inequality is low in India. (It also serves as a useful reminder of the difficulty of making international inequality comparisons, a difficulty too often overlooked when cross-country comparisons and regressions are undertaken.)
Inequality affects poor and rich communities alike. The research challenges conventional wisdom at the local as well as the international level. The poverty mapping exercise mentioned earlier shows that consumption inequality seems to be at least as high among poorer rural communities as among better-off ones. Indeed, in Andhra Pradesh inequality seems to be even greater in poorer rural communities than in better-off ones. If local inequality of consumption is also an indication of concentration of power and influence, then resources allocated to poor communities—for example, under “community-driven development” approaches—will not necessarily reach the poor and might instead be at risk of elite capture.
Consumption inequality has fallen over the longer term in India but is now on the increase. Turning from levels to trends, inequality is on the rise in India. This is a recent phenomenon. As figure 7 illustrated, the last five decades show a long-term trend in rural areas of declining inequality; a decline in inequality in urban areas until the 1980s, and since then a rise; and a long-term upward trend in the urban-rural gap. What this would mean for total inequality depends on how adjustments are made for urban-rural cost of living differences, but given that the great bulk of the population still lives in rural areas, a long-term downward trend would be expected.
Focusing in greater detail on the more recent past, however, tells a different story. Rural growth switched from being pro-poor (largely benefiting the poorer) between 1983 and 1993–94, to being largely distribution neutral between 1993–94 and 2004–05. In urban areas, over the same period, growth went from being distribution neutral to being pro-rich. And the gap between rural and urban areas continued to widen. Again, aggregate comparisons are difficult, but this set of findings would suggest an upward trend in national inequality. When one uses the official urban and rural poverty lines to correct for cost-of-living differences over time and between urban and rural areas, for most inequality indicators, no increase or a decrease in national inequality is apparent between 1983 and 1993–94, and a small increase is seen between 1993–94 and 2004–05. Figure 15 illustrates the Gini coefficients.
These results understate the increase in inequality, likely because the household consumption surveys are missing increases in top-end incomes. Increases in wealth holdings are also driving perceptions of increased inequality. Although the survey data we examine show an increase in inequality, it is not a dramatic increase. We have already noted, however, that the survey data likely under-report consumption at the top end. It is certainly popularly perceived that inequality has increased sharply, very likely driven by the observation that rich Indians did extraordinarily well during the boom of the 1990s. According to one study, in 1999–2000, the gap in per capita income between the 99th and 99.5th percentile was almost four times as large as the gap between the median person and the 95th percentile. Incomes of the super-rich at the 99.99th percentile grew by over 285 percent between 1987–88 and 1999–2000 (Banerjee and Piketty 2003). Wealth inequalities are also on the rise. Between 1996 and 2008, wealth holdings of Indian billionaires are estimated to have risen from 0.8 percent of GDP to 23 percent (Walton 2010).
Growing divergence across states in mean incomes does not explain the increase in inequality observed in the survey data. Divergence across states is often pointed to as the main source of rising inequality. Indeed, inequality in mean incomes across states is increasing, according to national accounts data (figure 16). Rich states used to have average incomes twice those of poor states in the 1970s; now the ratio is closer to four times. However, despite the clear evidence of divergence across states in incomes as measured by the national accounts, a decomposition analysis of inequality, using survey data between states, or between high-growth and low-growth regions, reveals that only a very small, albeit growing, share of overall consumption inequality can be attributed to differences in mean consumption levels between states. In other words, inequality of consumption within states, and within regions, dominates.
Increased returns to education appear to be an important factor. A similar inequality decomposition exercise (figure 17) shows that in urban areas, the share of inequality explained by a simple division of the population into those with and those without a primary education shows very little change. But the share of inequality explained when the population is divided into those with and those without a graduate education doubles to almost 20 percent in 2004–05, up from only 11 percent in 1983. The rural analysis tells a slightly different story. There, the share in inequality using both decompositions rises, more so for the graduates, but from a very low base.
This evidence fits well with the story of the growing non-farm sector told earlier, as we know that the less the countryside is dominated by agriculture, the more important education is. Even completing primary education increases the chances of escaping the farm. That education is a source of rising inequality appears paradoxical inasmuch as access to education is becoming much more equitable over time. However, inequalities in learning are high in India—among the highest in the world, and rewards to skills are becoming more unequal (Dutta 2006; Kijima 2006).
Some types of inequality, but not all, are harmful for growth and economic development. The link between inequality and poverty is far from straightforward. Everything else being equal, a rise in inequality will dampen the poverty-reducing impact of an increase in mean incomes. But everything else is not equal, and some growth accelerations might not be possible without an increase in inequality. The analysis suggests that the recent experience of India might fall into such a category, with increasing returns to education a necessary requirement for its recent rapid growth.
Even so, rising inequality can be of concern for other reasons. Some inequalities may be more structural and exclude groups from the development process.
This article is an excerpt from a technical report - “World Bank. 2011. . World Bank. © World Bank. https://openknowledge.worldbank.org/handle/10986/2299 License: CC BY 3.0 IGO.”
Download the Report - LINK