In another decade from now, it is likely that the per capita income of the richest state will be more than four times the per capita income of the poorest state

Finance minister Arun Jaitley wants to roll out the Goods and Services Tax (GST) on 1 June. Growing income inequality among India’s large states should be disconcerting especially in a complex federal polity such as India. Photo: Pradeep Gaur/Mint

The average Tamilian today earns four times more than the average Bihari. Just three decades ago, the average person in the then richest large state, Maharashtra, earned less than twice the average person in the then poorest large state, Bihar. In another decade from now, it is likely that the per capita income of the richest state will be more than four times the per capita income of the poorest state. The richer states are getting richer while the poorer states are being left behind. This stark and continuing income divergence is well established and understood but puzzling. It is puzzling because it runs counter to economic theory and observed reality in other large federal economies in the world. What is perhaps even more worrying is that we do not have a rigorous explanation yet for this phenomenon.

The Economic Survey of 2016-17 presents a scholarly and masterful narrative of the current state of India’s economy as well as some fascinating insights and ideas for the future. Chapter 10 of the Economic Survey provides an in-depth analysis of this very issue of income divergence of Indian states. Figure 1 in the chapter is an arresting depiction of this phenomenon that shows how per capita incomes in the various states have diverged over the past three decades. In a similar study published last year, we used a crisper data set to leave out small states that can potentially skew the analysis and included just the 12 largest states (Uttar Pradesh, Maharashtra, West Bengal, Tamil Nadu, Bihar, united Andhra Pradesh, Bihar, Madhya Pradesh, Karnataka, Gujarat, Rajasthan, Odisha and Kerala) that account for more than 85% of India’s population and nearly 80% of annual gross domestic product (GDP). (India’s Curious Case Of Economic Divergence, IDFC Institute, October 2016,). Our analysis compared India with other large federal economies such as the US, China and the European Union and revealed exactly the same story of stark divergence among India’s large states which put India at odds with the rest of the world.

Chart 1 represents how decadal per capita incomes of different regions have diverged relative to the poorest state’s income in 1974 in each of these federal economies. Straight lines reveal tight income convergence among the different regions while right-tilted lines reveal income divergence. The US has remarkably straight lines, while India’s states have diverged extraordinarily between 2004 and 2014, much more than China’s income divergence. Quite simply, India’s richer states are racing ahead of the poorer ones, making India an outlier in the world.

Economic theory postulates that, just as with countries, there will be a convergence of richer and poorer sub-national regions, driven both by diminishing returns to capital, so that investment is more productive in poorer than richer regions, and by mobility of factors of production, which tend to equalize factor rewards across regions. In other words, if the rate of return on investment is higher in Bihar than in Maharashtra, and/or if labour is cheaper in Bihar than in Maharashtra, new investment will move to Bihar to capitalize on a higher rate of return and lower labour costs. Eventually, this will lead returns on investment and factor rewards, including wages, to converge between Bihar and Maharashtra, if not become exactly equalized. Economic theory also suggests that convergence driven principally by differential rates of return on investment will occur slowly, while convergence driven by labour migration should be much more rapid. So, if there are restrictions on free movement of labour or capital or goods within a nation, it can potentially thwart convergence trends.

In our paper, we conjectured that India’s complex cultural and linguistic diversity, to say nothing of sheer physical distances, served as a natural barrier to free movement of labour and trade, impeding a rapid convergence of income among Indian states, as is observed in other federal economies. That is, it is much more difficult for a Hindi-speaking Bihari to move to Tamil-speaking Tamil Nadu for a job than, say, a New Yorker’s ability to move to Seattle in the US. So, this will keep wages high in Tamil Nadu vis-à-vis Bihar. What is more, the patchwork quilt of different excise taxes has widely been believed to retard inter-state trade, and this, indeed, has been one of the main motivations behind the goods and services tax (GST).

As it happens, however, new research, again presented in the survey (chapters 11, 12), casts doubt on these conventional explanations. Deploying novel “big data” analytical methods, the survey shows that both inter-state trade, and inter-state movement of people, are much more robust than we have conventionally assumed. The survey effectively argues that the cultural, geographical and linguistic diversity of India’s states are not the walls blocking free inter-state trade, as conventional wisdom held. In the language of the survey, these data belie the “caricature of a barrier-riddled economy”.

In fact, this only deepens the puzzle of economic divergence. For, if the arteries of inter-state commerce of goods and people are healthy and not sclerotic, our prior hunch that such barriers might account for the lack of rapid convergence—indeed for outright divergence—does not seem to hold. In summary, neither the survey nor our research has a rigorous explanation for this puzzling phenomenon of richer states getting richer more rapidly than the poorer states of India.

However, our analysis highlighted one other curious observation. India was indeed experiencing income convergence among its large states until 1990. Beginning 1992, the divergence among India’s states began to skyrocket. For evidence of this, see chart 2, which plots “second order” sigma convergence—in simple terms, the acceleration of income divergence, which, as marked, clearly takes off in 1992. It is inevitably tempting to hypothesize that the economic liberalization of 1991 played a role in this income divergence of India’s large states. Amid lack of substantial analytically grounded evidence of the causal impact of the 1991 reforms on income divergence, it is at best a hypothesis currently. But it is certainly worth pondering if the GST regime that will further ease inter-state trade barriers will exacerbate this divergence or reverse the trend to cause eventual convergence. In other words, will 2017-18 prove to be the second breakout year, after 1991 for income divergence among Indian states?

As we discuss this, there is an interesting case study that is unravelling to exactly this debate. Apple Inc. wants to set up a manufacturing plant in India, a very welcome development for investment- and job-starved India. Apple is negotiating for certain tax exemptions and in a pre-GST era, states would have competed with each other to offer these exemptions and throw in other incentives to attract Apple. In a GST regime, the ability of poorer Uttar Pradesh to wean Apple away from richer Karnataka using tax tools is even more diminished. To be sure, there are various other factors of governance, law and order, land and labour costs that will influence Apple’s choice of state. Perhaps states can still circumvent the GST spirit of one market one tax by offering cashbacks in lieu of GST as incentives to companies. Whichever way one expects this to play out, it is indubitably clear that with GST, Uttar Pradesh is in no better a position to attract Apple vis-à-vis Karnataka than it is without GST. So, GST at best will not have an impact on the current disturbing trend of income divergence of states or at worst will exacerbate it by removing a powerful fiscal tool of states.

We conjecture that the confluence of existing evidence on income divergence among states, and the survey’s new evidence on robust inter-state mobility of goods and people, points toward GST worsening economic divergence among states. For, if it is underlying deeper factors such as higher private sector productivity, and better quality of governance (the latter is also highlighted by the survey) that distinguish leading from lagging states, these are factors that will not disappear simply because goods and people are on the move. Rather, these could be self-reinforcing trends, driven by economies of scale, agglomeration effects and network externalities, such that mobile investments move to already prosperous regions which have good infrastructure and good governance, a ready pool of skilled labour, and proximity to major markets. This is a recipe for widening divergence, not convergence over time.

Growing economic divergence among India’s large states should be disconcerting especially in a complex federal polity such as India, with all its fissiparous tendencies.

It is prudent to keep a watch on this trend in the years post the roll-out of GST. Unfortunately, it is not obvious that the conventional economic policy tools of the Central government can do very much about this problem. While there is a process for devolution of funds to states based on economic prosperity, a sustained widening of the gap between rich and poor states will render such devolution methodologies moot. Perhaps, in the fifth year of GST rollout, we can check to see if economic divergence has sharpened or narrowed and take corrective steps. Such steps might need to include region-specific subsidies to attract investment and jobs. Yet, caution is required. The law of unintended consequences is the perennial bane of public policy.