As you all know, one of the biggest problems faced in India is the fiscal deficit. The consolidated fiscal deficit, of the centre and the states, has been at stubbornly high levels for around twenty years now.

The essence of this problem has been a stagnation in the tax/GDP ratio. From 1990–91 to 2003–04, we did obtain progress on direct taxes, which went up from 1.9 per cent of GDP to 3.5 per cent. The phasing out of customs duties has inevitably given poor growth in indirect taxes, which went from 7.9 per cent of GDP to 5.7 per cent. The fiscal difficulties at the states have given a fresh impetus to state level tax efforts, which have yielded some progress, with growth from 5.3 per cent of GDP to 6.3 per cent of GDP. However, the overall picture has been unchanged, with the tax/GDP ratio being stable at 15.5 per cent of GDP in 2003–04 and in 1990–91. The combination of large fiscal deficits with a stagnant tax/GDP ratio has given sharp growth in the debt/GDP ratio. From 1992 to 1998, the debt/GDP ratio was stable at 60 per cent of GDP, and that might have given some comfort. But after that, it has resumed an extremely rapid climb to the present level of 80 per cent of GDP. This has fuelled concerns about the possibility of India facing the problem of debt trap as interest payments have steadily become a bigger fraction of tax revenues.

Sometimes, India’s fiscal problem is seen narrowly in terms of debt sustainability or a debt trap. I think this is a narrow perspective. The fiscal problem can be damaging to growth in coming years, even if it does not come to a debt trap. The reasons for this need to be reiterated:

It is important to observe that the fiscal problems would have had an exacerbated impact on growth, by ‘crowding out’ private investment, if it had not been for the growth in household savings that was discussed earlier. Roughly speaking, Government has taken 10 per cent of GDP in 1990 and in 2003. However, household savings grew from 18 per cent to 23 per cent, thus supplying an additional five percentage points of GDP to non-government investment in the country.

In many countries, ‘downsizing government’, i.e., cutting government expenses, has been central to fiscal adjustment. In the case of India, central government expenses dropped from 18.9 per cent of GDP in 1986 to 15.6 per cent of GDP in 2001. These values do not appear to be particularly out of line by international standards, and are broadly consistent with the level of expenses that are required to produce public goods of the required quality and quantity.

Three difficult items of expenditure, i.e., interest payments, defence expenditures and subsidies make up near 100 per cent of tax revenues. In addition, there are highly inflexible expenses such as pensions, transfers to states, etc. Hence, it appears that there is little flexibility in obtaining a fiscal adjustment by compressing expenditures. There is a great deal that can be gained in terms of improving the extent to which existing expenditures are refocused away from subsidies towards providing public goods, and improving the efficiency of provision of public goods. However, it is hard to visualise a drop in expenses which would be large enough to significantly contribute to the required fiscal adjustment.

This leads us to focus on improving tax revenues as the central policy instrument in the required fiscal adjustment. Hence, efforts towards the fiscal consolidation, that have been undertaken, are focused on the following elements:

Have these efforts borne fruit? Many observers have pointed out that the tax to GDP ratio is still below the levels found in the late 1980s. This observation, taken in isolation, is sometimes interpreted as implying a failure of tax reforms in India. However, this aggregative fact masks important accomplishments in terms of obtaining change.

The table summarises changes in the structure of tax revenues from 1990 to 2001.

The most important accomplishment was in the area of direct taxes, which grew by almost 7 times over these 11 years. Direct taxes hence improved sharply from 18.9 per cent of collections to 35.7 per cent. This may be interpreted as a striking ‘Laffer curve’ outcome, where a sharp reduction in rates was accompanied by a sharp improvement in tax collections, by influencing incentives towards tax evasion and labour supply. Customs collections have lost ground, and will drop further, as India shifts away from protectionist policies. Taxing the services sector has now begun, in a small way.

These reforms anchor the fiscal consolidation envisaged in the Fiscal Responsibility Act,^[14] and the commitments of state governments, which require elimination of the revenue deficit: from 5.83 per cent in 2002–03 to 0 by 2007–08. It is important to envision what the sources of a 5.83 per cent improvement could be. One example of a feasible combination could be as follows:

The Interim Budget presented in February this year indicates that fiscal consolidation is proceeding on this line, as the revenue deficit for the year 2003–2004 has been projected to decline by 0.5 per cent. This has been due to combination of higher tax/GDP ratio and lower current expenditure.

This fiscal consolidation will assist GDP growth in many indirect ways, including:

A successful implementation of this transformation of the tax system, and an elimination of the revenue deficit by 2007–08, is perhaps the most important single issue in public policy in India today. The tax reforms that are currently underway will enable the economy to meet the objective of fiscal consolidation. Successful fiscal consolidation will enable the economy to achieve other important social goals such as better environment protecttion, greater investment in health infrastructure, Research & Development and the agriculture sectors.

A major problem that India faces is the large cross-sectional dispersion in economic development. There is a roughly 3:1 ratio in the per capita GDP, when we compare the richest states to the poorest states. Much attention has been focused on the ‘BIMARU’ states (Bihar, Madhya Pradesh, Rajasthan, Uttar Pradesh) which have high population density and low per capita output. The term ‘BIMARU’ is catchy because the word ‘bimaar’ means ‘sick’ in Hindi. The table above identifies the four large states where per capita SDP was over 33 per cent below the national average. These four states make up 33.9 per cent of India’s population.

Regional disparities in India have been present for at least a century, if not more. Under normal circumstances, the processes of the market economy should generate ‘equalising differences’, whereby firms move to low-wage areas in the quest for reduced costs thus equalising differences in wages and land prices. Similarly, individuals migrate to high wage areas, thus equalising wages, and increasing the land per capita in poor areas. These processes are expected to generate convergence of per capita GDP in the normal framework of growth theory.

It is important to emphasise that the forces of convergence depicted here are based on factor mobility. They operate over and above the conventional notions of convergence through trade, which are based on technological catch-up and trade in goods, without factor mobility. This worldview has faced a challenge from the empirical evidence of the 1990s, where there is some evidence of a lack of convergence. Some states, particularly the states of the West and the South, seem to have excelled in harnessing the opportunities of globalisation and the market economy. In other states, weaknesses in human capital and governance have generated reduced growth rates in the post-1990 period.

This has been a source of much concern on the part of many observers, from two points of view. First, it is argued that if the economic reforms of the 1980s and 1990s failed to ignite growth in Bihar, then there is a need to find a new policy mix which can achieve high growth in Bihar. Second, there are fears of mounting political stress that might come about if income disparities between rich and poor states widen further. There is a remarkable similarity between these problems in India and those that have been observed in China, where coastal provinces have progressed enormously compared with the interior.

These problems are undoubtedly important, and are going to be a central issue in Indian economics and politics in the years to come. While the above difficulties are real, there are also many forces at work which are steadily having an ameliorating effect.

Flexibility of the labour market: Factor mobility is a fundamental element of the process of equalising differences. As of today, roughly 90 per cent of India’s labour force is in the unorganised sector, which is a classical labour market, undistorted by labour law. In addition, unlike China, India has no government restrictions on inter-state or rural-to-urban migration.

This innate flexibility of the labour market will assist the process of convergence. In the historical data, migration flows do not (as yet) account for substantial movements of the population. The reforms of the 1990s ignited high growth rates in some states. It is likely that migration flows have a lagged response to high wage differentials. By this logic, the 2011 census may be expected to show larger migration flows than were observed in the 2001 census.

Impact of new infrastructure on ‘equalising differences’: The development economics literature has emphasised the problems of land-locked states, which are unable to harness gains from trade through high costs of transportation.^[15]

India’s growth experience suggests that geography is important. At the same time, there are exceptions. Coastal states in India have fared well; however, Orissa is a coastal state. Land-locked states have fared poorly; however, Punjab and Haryana are land-locked.

Gains from internal trade are clearly an important mechanism through which poor states can obtain economic growth. This is critically related to costs of transportation. This suggests that the recent successes in infrastructure policy — particularly in roads, ports, airports, and telecom — are highly significant in thinking about regional disparities. The new roads being built by NHAI imply that vegetables produced in Bihar or Orissa can find markets in Calcutta. This constitutes a new impetus for the forces of convergence, as compared with the preceding post-independence experience.

Fiscal transfers: India has a well-developed system of fiscal transfers, through which taxes collected in rich states are transferred to poor states. This constitutes an important channel for convergence — one that is perhaps reminiscent of the 60-year story of North Italy and South Italy.

While these rules have always been with us, the economic significance of these transfers improves in line with growth in GDP and in the tax/GDP ratio. Holding the fiscal rules intact, when the size of the pie goes up, larger per-capita flows are being sent into poor states. In the decade of the 1990s, India’s GDP was roughly $350 billion and the tax/GDP ratio was roughly 12 per cent. GDP has already risen to $620 billion, giving a quantum leap in the expenditures of government. Looking forward, in a few years, if we envision GDP of $1 trillion and a tax/GDP ratio of 15 per cent, then there will be enormously larger resource flows through existing fiscal institutions, which will generate much larger spending in poor states.

Policy innovations: One important insight derived from the experience of economic growth in East Asia is the importance of ‘regional role models’.^[16] East Asian countries learned from each other. Across these countries, there was a significant amount of experimentation and real-world trials of alternative ideas, including choices of effective institu­tions, policies, and technologies. There was a contagion effect within this region with countries learning from each other’s success stories.

In the decade of the 1990s, a similar phenomenon has begun with the states of India. States are now increasingly conscious of the importance of local public goods. The political leadership of many states is increasingly conscious of the need to find policy innovations which would improve the quality and quantity of local public goods.

The 1990s began with a certain heterogeneity of governance procedures in the states. The economic reforms of the 1990s have inevitably had a differential impact on various states; some states had policies which were more conducive to harnessing these opportunities. When the gap in per capita income widens, the political system has incentives to search for policy responses which would close the gap. Andhra Pradesh, Madhya Pradesh, West Bengal, and Kerala are all examples of states where there has been a distinct learning from the regional role models, and consequent changes in governance.

In parallel to this learning from regional role models, there are two important policy innovations which are going to fully play out in the coming decade. The first is the move towards smaller states. It is widely conjectured that smaller states are more effective at catering to local variation in preferences and technology, and at ensuring greater accountability for public goods outcomes. Uttaranchal, Bihar, Jharkhand and Chattisgarh are important experiments in this regard. It is, as yet, too early to tell whether the outcomes play out in line with the conjecture. If governance does prove to be superior in smaller states, then (a) it will generate convergence, given that these four states are all below the national average, and (b) it suggests one policy avenue for improving governance in other large states in the future.

The second innovation is the devolution to local govern­ments, as a consequence of the 73rd and 74th constitutional amendment. The underlying premise of Panchayati Raj is that when local citizens control public expenditures, there will be a greater likelihood of obtaining good outcomes in terms of producing public goods. There are three key elements of local autonomy: (a) Transfer of functions and schemes, (b) Transfer of staff, and (c) Transfer of funds, and autonomous financial decision making. As of yet, different states have made different degrees of progress on these three fronts. It is, as yet, too early to tell whether the outcomes play out in line with the underlying premise. If we do obtain improvements in governance by empowering local governments, then this would constitute one channel for convergence.

How are we faring? There is some evidence that these effects are already at work and are reshaping the nature of regional inequalities in India. There are two striking illustrations which show the changes which are taking place:

The actual outcomes, from 1993 to 2003, have been inconsistent with the prediction that IT-enabled services would primarily be located in the peninsula. When we look back at the last ten years, it is an undeniable fact that Gurgaon, Noida and Chandigarh have also emerged as the major centres of IT-enabled services exports. While locations like Bangalore, Madras, Hyderabad, Poona and Bombay have all also succeeded in this area, Gurgaon and Noida are probably the largest centres. This suggests that issues such as low labour cost dominated issues such as poor production of local public goods, which suggests that forces of convergence were effective.

The most interesting evidence about the question of convergence is found in data for investment projects outstanding.^[17]

The table above exploits the CMIE database which tracks investment projects at hand as of a point in time. It juxtaposes the projects under implementation as of April 1995 (the first point in the CMIE database) versus October 2003 (the most-recent date available). All values are expressed as rupees per capita.

This data is interesting from two points of view. Focusing on levels, we see states like Gujarat, which have above-mean output and above-mean investment. At the same, there also appear to be equilibrating forces at work. High growth in investment is seen in backward states like Kerala, Madhya Pradesh, Chattisgarh and Bihar. In a striking display of convergence, of the 10 states with above-average growth in per-capita investment, 8 had a below-average level of per capita investment as of 1995. In addition, Punjab shows the opposite phenomenon. Low growth in investment is found in high income states like Punjab and Gujarat.

These trends are indicative of the possibility of meeting the objective of regional equity with well defined policies at the Central level and at the State level. Regional equity is going to be perhaps one of the most important issues for the political economy of growth in a federal system like India. We will need to be continuously mindful of this aspect and keep policies under review so as to achieve equity in outcomes across the States of our Union.