Harsh Gupta
State-varying GST necessary for Indian federalism
This article originally appeared in CRI content has now been subsumed in The views expressed here are personal and do not necessarily reflect those of the editors of

State-varying Goods and Services Tax (GST) necessary for Indian federalism*

How to efficiently integrate the Indian market while maintaining the autonomy of states

Executive Summary

The current Goods and Services Tax (GST) proposals conflate two objectives. Firstly, rationalizing and merging various indirect taxes – like Value-added Tax (VAT), Central Sales Tax (CST), excise duty and others – for better economic efficiency. And secondly, to have a uniform rate for such a tax in all the states-which is ostensibly crucial to integrate the market.

But using information technology which allows for input tax credits (ITC) across states despite different tax rates (based on unique transaction IDs) would create a common market without sacrificing tax-rate setting autonomy of states. An integrated market does not require a uniform tax rate in all the states. Indeed, such an extreme tax harmonization across states would incentivize states to run higher fiscal deficits and would invariably result in tighter conditionalities by the Center on spending by states. This would further centralize power and homogenize welfare schemes across states, preventing inter-state health fiscal competition and experimentation in public policy.


After independence, consumption taxes in India used to be primarily raised by direct taxation at point-of-sale. But sales taxes were easy to evade because one point in the supply chain felt the burden of the entire taxation. Over time, economic theory and international best practices brought the VAT to India. In a VAT system, a business pays VAT on whatever inputs it buys, but could use that VAT input as a credit against the VAT output it charged to its buyers, hence preventing cascading taxes. But even post-VAT many taxes like excise continue to cascade downstream.

Moreover, Input Tax Credit (ITC) is not available across states. This makes buying from within a state more beneficial than from another state (on the margin) even if the two states had the exact same tax rate. Equally importantly, services were taxed differently (and often less) compared to goods, and in a rapidly modernizing economy- this was neither efficient nor fair.

Problems in the current proposals

Before looking at the specific problems with the current proposals, the good aspects must be acknowledged- all the proposals solve the across-state ITC problem. Moreover, despite pressure from states the latest proposals have narrowed down the tax categories to three – standard, low and special. These reforms – along with the phase-out of excise and octroi- represent significant improvements over the current tax code.

But Indian states cannot levy corporate/income taxes, and currently revenue from sales tax/VAT is a significant part of the states’ revenue. With GST subsuming other indirect taxes, this ratio will only go up. But according to the current proposals, a state’s elected representatives cannot unilaterally cut taxes or prevent tax hikes, then ceteris paribus they would be incentivized to further increase spending to win votes (according to public choice theory which rejects the assumption of selfless benign politicians)

That is, if state leaders cannot politically benefit from cutting tax rates (or preventing tax increases) across the board, then why will they forgo the political benefits of increased spending? State politicians would then present the Centre with the fait accompli of more budget deficits. With this structural problem, a uniform GST rate would have to be increased again and again, potentially harming the economy. Ultimately, the Center would reduce the autonomy of state spending further with more conditionalities and austerity provisions. A simple analogy here is – imagine 28 people going to a restaurant and sharing the bill (or in the ratio of their weights if you like). They are much more likely to order expensive food and frivolous deserts, as compared to when they pay most of their bill themselves. That is how state governments are likely to react- with inefficient spending -when taxation gets centralized.

Current dual structure is not consistent with fiscal federalism

Fundamentally, a single GST and a dual but uniform GST are the same as far as tax incidence and incentives are concerned. Therefore, while the constitutional ramifications of these proposals may be different both the current proposals (and other such new or similar proposals) of GST would centralize taxation.

As a result states with high growth and stronger revenue base might be forced to tax significantly more than their local spending and current horizontal redistribution needs. Moreover low growth states might not be allowed to attract industries with tax reductions and might be forced to spend more than they would otherwise have (what economists call “the flypaper effect” whereby even without conditionalities revenue received from higher administrative levels is disproportionately spent, and a very small part is used for reducing the taxation burden on the common man)

Suggested solutions

1. We should have a Central GST and a state GST-with the states free to have different rates for the state GST so that we can preserve the autonomy of states and enhance healthy inter-state fiscal competition – a race to the top. and not to the bottom. This proposal only needs an appropriate Information Technology (IT) infrastructure for inter-state tax credits (ITC). Nandan Nilekani of the “Aadhar” program (Unique ID authority) is helping prepare such an infrastructure for the non-varying GST proposal in any case. For example, the European Union has achieved a single, efficient market while still having different value-added tax (VAT) rates for its mem her states using the VAT Information Exchange System.

2. There should be a fixed number of rate categories in all states to prevent favouritism and crony capitalism. By having similar tax bases and model state GST legislations throughout the country, state politicians would not be able to offer arbitrary tax breaks. Federalism means allowing maximum flexibility in the policy of states so far as the rule of law is not violated-that is, any policy can be followed, so long as specific individuals or companies are not unduly privileged or discriminated against.

 3. To prevent excessive or destructive competition the EU does has a minimum GST rate for its member countries; to allay “tax war” fears, India too could have a EU-tvpe GSTNAT structure with minimum tax rates (of course. the minimum mandated consumption tax rate in the American federal system is effectively zero)

4. A state should tax the value-added within its own boundaries. That is, the GST reform should aim for a source-based rather than the currently recommended destination-based taxation system. A destination-based system would be unfair for states that have a “trade surplus” with other states as inputs provided by “net exporting” states would not be taxed by those states. This is how we can create a truly efficient and seamless “origin-based” system – towards which the European Union has also moved with some success from a pre-1993 exclusively “destination-based” system. Moreover. a source-based GST would make our export subsidies more transparent by making exporters the only final sellers who do not charge their buyers a GST (or by giving only our exporters a credit for VAT Output on exports), if we decide to continue to subsidize our exporters due to international competition.

Inter-state Tax war and Race to the Bottom? No.

Domestic skeptics of inter-state tax competition say a “race to the bottom” could take place, with states potentially cutting spending in drastic ways to cut taxes. But taxpayers, like consumers, operate on two variables-cost and quality. Tax competition makes state spending more efficient and does not necessarily decrease it: A state could choose to have great infrastructure and social insurance, and firms might be ready to pay higher taxes to locate there. Such policy experimentation in the states is useful because the successful policies are then copied by other states.

Accountability and efficiency are best served when the expenditure and income are done at the same administrative levels unless there are significant spillover effects. And for equity reasons, the Centre can always spend more on the backward states from its revenue share (and if the Central welfare schemes are means-tested, it will happen automatically anyway)

How will this work- An example

Net GST due for somebody in West Bengal will only be for his value-add in West Bengal. That is his net GST payable due would be equal to:

Unique West Bengal rate multiplied by difference of {output in West Bengal AND input from anywhere else-West Bengal/Kerala/China etc.}

(Of course, the Kerala/China purchases need to be confirmed by the new smart database)

For example, if somebody wants to make paper in W.B. and if he buys wood from Kerala at say Rs. 100 per unit. Let us say, that if we had different GST rates in different states- say 10 percent in Kerala and 20 percent in WB (including both Central and unique State rates). Then he would buy the wood for Rs. 110 in Kerala post-GST, and let us say he does a value-add ofRs. 40. Then when he sells paper in WB for 110+40+8 (Rs. 8 = 20 percent of 40) at Rs. 158, he pays the WB government Rs. 8 in taxes. The Kerala and WB govt. would not have any dues to clear. but they would have to verify the transaction information – and the Central database can act as a clearing house for that.

Related comments- States as Laboratories of Democracy

Good policies like Tamil Nadu’s school lunches, or Gujarat’s Chiranjeevi maternity vouchers, or West Bengal’s land reforms did not come by central diktat. They came about by the genius of state politicians, an intricate knowledge of local conditions and sometimes good old trial-and-error. Centralizing finances – which current GST proposals inadvertently lead to – would be harmful for such future experimentation and healthy competition amongst states.

A relevant example here is the bipartisan welfare reform passed in the United States by former Democratic President Bill Clinton, and a Republican Congress. Instead of micro-managing welfare policies centrally, the reform provided for block grants – and let states compete to come up with the best policies to help the poor and the unemployed. The results over the years resulted in lower welfare cases and better employment numbers, andproved the critics of the reforn wrong.

Of course, competition between states only works if there is free domestic immigration along with no other (non-tax or tariff) commercial barriers. That is, if people and goods are allowed to cross state borders without protectionist or migratory borders. Fortunately, the evolving Indian constitution and the Supreme Court have consistently stood for such free movement.

Policy innovation in some states forces the laggard states to shape up, as the revenue base erodes due to the migration of taxpayers and industries. According to M Govinda Rao, director of the National Institute of Public Finance and Policy, a thorough review of fiscal federalism for India’s increasingly market-based economy is urgently needed. He notes that the “fiscal arrangements in India have evolved in a quasi-federal system to meet the requirements of centralised planning in a mixed economy framework.”

The central government should of course fund capital investments like infrastructure in poorer states to “jump-start” their economies, and then let migration across states further equalise incomes. Moreover, central means-tested plans would automatically help the poorer regions more but will not create disincentives for fiscal prudence on the part of the states.

The top-down mentality of governance in India- New Delhi’s fiscal carrots and political sticks

States, however, are not meant to be just effective implementers of policies that are fashioned in New Delhi. But this is what they have been reduced to because he (the Central Government) who pays the piper calls the tune. One look at the budgets of state governments, and one can see New Delhi’s imprint all over.

On the revenue side for states, we have VAT and other indirect taxes whose rates are already substantially harmonized. Even the remaining independence to set rates will completely go if the current proposals of GST come into force. Other internal revenue sources are distortionary levies and octroi which are being gradually withdrawn, and the states cannot tax income or profits. Therefore, they end up relying on grants from the Union government.

On the expenditure side also, the state must increasingly spend on the New Delhi’s welfare programs. If they do not, they either stand to lose substantial matching grants (as in the Sarva Shiksha Abhiyaan education scheme, where 50 to 85 percent of the funding comes from the Union government) or end up paying fines and mandated benefits (for example, not implementing the National Rural Employment Guarantee Scheme means losing central grants, even if the mandated wage is above the local market wage). Such matching grants should be used by New Delhi only to remove existing distortionary policies (like they were used to repeal rent control laws), otherwise they create incentives for states to spend on locally unsuitable policies. Other state expenditures include salaries of state officials which must be raised in tune with the Center’s Pay Commissions.

If a state is actually allowed to set its own policies, it must perforce be economically viable and administratively efficient. If it has a deficit, it must raise taxes or reduce spending or both. States, unlike the Union, cannot print money, so state bonds would be effective only in rolling over limited debt.

Comparative advantage of states cannot be leveraeed in a centralized economy

Over-centralisation has kept many states poor by preventing them from exploiting their comparative advantage. For example New Delhi’s “freight equalisation” policies-the effective centralization of the transportation policy regarding mineral resources- partially contributed led to West Bengal’s relatively poor economic performance after independence. Moreover, it is common in India to allege that the Central government – if it belongs to a different party or coalition – is fiscally discriminating against one’s state. Centralization allows partisanship to further fester.


It is time India learns from the fiscal federalism of the USA, EU, China, Canada and Brazil – they have an integrated market without forcing same tax rates for all states. A huge economy and heterogeneous polity like India’s needs to be aware of the dangers of fiscal centralization. Why cannot we have a dual but differentiated GST? The only major economy which has had a same-rate GST or similar tax structure is Australia and certain provinces there (like Western Australia) are already expressing concerns about fiscal centralization. We too can route less money through our central government- and we too will have a better standard of living as a consequence.

With fiscal autonomy comes responsibility and efficiency. Rather than adopting more technocratic and centralised solutions, India must have a debate about the first principles oflndian federalism and then translate them into policy reality.

Harsh Gupta is a  Mint Columnist and  Co-author of upcoming book on financial derivatives.

This  idea was first written by Harsh in Mint  here