Growth and Social Outcomes in India: Part One

Op-ed in the Business Standard, New Delhi

July 25, 2012

The frenzied downgrading of India's growth prospects, at the same time as the government has stepped up its social spending, raises an important question about the consequences for social outcomes. How important is growth for social outcomes? This question deserves some careful and dispassionate analysis, over longer and shorter time horizons. One virtue of India being a union of states is that it serves as a laboratory with multiple policy experiments, helping social scientists draw conclusions. This within-India framework can usefully supplement international comparisons.

Consider state-level data on four key social outcomes: poverty, life expectancy, child malnutrition, and inequality. The relationship between income and each of the social outcomes is depicted in the four figures (attached [pdf]). The long run nature of the relationship can be assessed by plotting the level of the social outcome, say the level of poverty, (on the y-axis) against the state's level of economic development (measured on the x-axis by state per capita GDP). The figures are for the most recent period for which data are available.

The line in the figures depicts the average relationship between income and the social outcome (the flatter the line the less strong the impact of income on the social outcome). These figures do not prove that development caused the social outcome, but the associations are nonetheless suggestive.

With the exception of inequality, the figures show clearly that the more developed states have better social outcomes, just as the richer countries do internationally. That is, states that have a higher standard of living on average tend to have lower levels of poverty, higher levels of life expectancy, and lower levels of child malnutrition. The magnitudes are also interesting. A state like Karnataka that is about 125 percent richer than Uttar Pradesh (UP) will have a poverty ratio that is about 13 percentage points lower (Karnataka's headcount poverty ratio was 24 percent, UP's 37 percent; figure 1). Similarly, Himachal Pradesh, which was one-and-half times as rich as UP, will be associated with a life expectancy higher by seven years (67 versus 60 for UP; figure 2).

As in every relationship, there are a number of interesting outliers. For example, Kerala stands out as India's Scandinavia in having much lower levels of poverty and better health outcomes for children and the population as a whole. These achievements are greater than what would typically hold given Kerala's income level (Kerala is far away from the line in figures 1, 2, and 3).

Another standout state is Jammu and Kashmir (J&K), where poverty levels are amongst the lowest in India despite J&K being one of the poorest states. Clearly, history—and the early land and educational reforms under Sheikh Abdullah—played a key role in this regard, insulating the most economically vulnerable from the subsequent turmoil.

Surprisingly, Gujarat, which is India's China in growth terms, has been less than stellar in reducing child malnutrition. In 2005–06, nearly 45 percent of Gujarat's children under five were malnourished. In comparably rich states—Kerala, Tamil Nadu, and Punjab—the numbers range between 23 and 30 percent, significantly better than Gujarat's. Gujarat should also be doing better in terms of life expectancy given how rich it is (it is below the line in figure 2).

Arguably the worst performer is Madhya Pradesh (MP), which has the highest level of child malnutrition and lowest life expectancy, a poor performance even allowing for the state's relatively low income. MP also has the second-highest poverty ratio, surpassed only by Bihar.

In fact, Bihar's performance on poverty is unusually bad. Had Bihar just been a "normal" poor state, its poverty ratio should have been closer to 40 percent than the 50 percent it actually registered. And since the poverty numbers are for 2009–10, Bihar's miserable showing is a salutary reminder that this state probably needs another decade of Nitish Kumar-type governance before any kind of victory can be declared. Put differently, Nitish Kumar would need to defer his prime ministerial ambitions by a decade if his campaign slogan can convincingly be: "I will do for India tomorrow what I did for Bihar yesterday."

The performance of the states on inequality defies the general pattern of higher incomes being associated with good social outcomes. In fact, just the opposite is true: Richer states, on average, are associated with higher levels of inequality of consumption (figure 4).

To some extent, the relationship within India mirrors the experience internationally. But there are also India-specific factors at work, namely the country's unusual pattern of development. Not just in services but also in manufacturing, India tends to use scarce skill-intensive labor rather than abundant unskilled labor, which is inherently disequalizing. This has created the "Precocious India" phenomenon whereby India's pattern of economic specialization mimics that in countries that are much richer. A more unequal India may have been not so much the consequence of higher growth, but the collateral damage of a Precocious India.

The corollary seems to be that poor and agriculturally dependent states such as MP and Bihar exhibit much lower levels of inequality, indeed lower even than what should be expected given how poor they are (MP and Bihar are below the line in figure 4). Surprisingly, Kerala is most un-Scandinavian in having very high inequality, much greater than for the typical state (figure 4).

Overall, and allowing for the inequality related caveat, the positive correlation between long-run income of states and their social achievements makes sustained economic growth and its pursuit desirable, a lesson that this UPA (United Progressive Alliance) government has sadly not acted upon.

But is this association also true over shorter horizons and in particular during the period of reforms after 1991? Tomorrow's column addresses that question.

A more technical note [pdf] elaborates on this column.