A key challenge facing Indian policymakers today is to ensure that the poorer regions of the country catch up with the living standards in the richer regions. This is of central concern as it enables a more broad-based sharing of the benefits of economic growth.
A comparison of annual per capita growth rates in 16 major Indian states before and after the mid-1980s shows the extent of the challenge—as the chart depicts. In the former phase, most states appeared to be stuck with the “Hindu rate of growth”, in a 1-2% range. At the same time, the US and most major economies were growing at roughly 2-3%, which meant that virtually all the Indian states were falling further behind in global terms. The good news is that since the mid-1980s, the majority of states have markedly increased their growth rates, so that these states, and India as a whole, have been steadily “converging” on income levels in the rich countries. The less good news is that a significant minority of states have barely seen any increase in growth.
This leads to two interesting questions: Was the break in Indian GDP?growth policy-driven??Why was the shift not uniform across states?
In our recent research (http://www.isid.ac.in/~cghate/stephenver2b-Final.pdf), we examine these questions using newly assembled data on regional indicators from 1960–2005. Our data encompasses and extends all previous data sets relevant to macroeconomic analysis of Indian states.
The divergent growth patterns highlighted in our chart could in principle have arisen due to a wide range of factors affecting different states in different ways. But a key conclusion of our work is that there was essentially a single common causal factor. We call this the “V-factor” because the output of “V-states” and “V-sectors” affected by it tended to fall fairly systematically, relative to the major economies, up until the mid-1980s (and were thus on the downward part of the “V”), but have risen fairly systematically since then (i.e., have been on the upward part of the “V”).
Both “V“ and “non-V” states are very diverse in nature. States as different as Madhya Pradesh and Maharashtra both benefited from the impact of the V-factor, while the equally different states of Orissa and Haryana appear both to have been unaffected by it.
So, what made the V-states turn around? Our study suggests, in contrast to some past research, that India-wide policy changes probably provided the crucial impetus. Past research has suggested that the turnaround occurred possibly in the late 1970s, which is hard to rationalize in terms of policy. Our more disaggregated analysis suggests that the V-factor had a low point distinctly later, in the mid-1980s. This result is much more readily explicable in terms of changes in the policy shift in India during this time. For instance, the period saw a gradual effort to eliminate market-distorting policies—such as curbs on investment and price controls—and to shift towards market-correcting and market-creating policies.
The big puzzle is: If the V-factor was driven by policy changes that were common across states, why did it have such uneven effects? What made the V-states different?
In some striking ways they were not different. In particular, and in contrast to claims often made, it has not just been rich states that have benefited from the turnaround. “V-states” were not systematically richer than the non-V-states. In 1981, just before the low point of the “V”, net state domestic product per capita (in 1993 prices) for V-states was Rs5,337, and Rs5,497 for non-V-states. Both groups also had fairly similar populations, population growth rates and investment rates.
We do, however, find some interesting differences. V-states were on average more urbanized and more literate. They were also somewhat more industrialized, and somewhat less dependent on agriculture.
There are also three other striking features of the data.
The first is that state-level public revenue expenditures show strong evidence of being negatively correlated with the V-factor. For instance, since the mid-1980s, many V-states (such as Gujarat, Maharashtra and Tamil Nadu) have decreased public spending on manpower (public employment) and increased capital investments. This suggests that revenue expenditures are a symptom of poverty, as supply-constrained economies reduce capital investments and increase revenue expenditures.
The second is that there is evidence that supply constraints may have limited the impact of the V-factor in some states. It is striking that the impact of the V-factor in V-states also shows up in a very similar V-shaped pattern of electricity production. Electricity supply is for some activities close to being a binding supply constraint as production bottlenecks in electricity hinder overall economic activity. V-states have typically been better at progressively releasing themselves from this constraint.
The third striking feature is the sectoral impact of the V-factor. In contrast to what might be expected, we find that registered manufacturing was not typically a “V-sector”: i.e., it did not experience any obvious turnaround in the mid-1980s. Instead, the impact of the V-factor in V-states was strongest in private sector service industries. This goes against consensus thinking which attributes India’s growth turnaround in the mid-1980s to reforms in the manufacturing sector.
We do not claim to have found a deep causal explanation of the V-factor itself, but the fact that there seems to be just a single factor that has driven the great Indian growth turnaround, rather than multiple factors, is striking. It suggests that, despite the diversity of the growth patterns across states, India-wide changes have ultimately dominated. Differences across states, and also across industrial sectors, have been driven by the extent to which they have been able to respond to this common impetus. A key element in our future research is to investigate further the crucial question of why some states have been so much better at responding than others.
Chetan Ghate is an assistant professor in the planning unit at the Indian Statistical Institute, Delhi centre. Stephen Wright is a reader in economics at Birkbeck College, University of London. Comment at email@example.com