Other countries, such as the member states of the European Union, have more complex VATs with many exemptions and multiple rates. Still, other countries — Brazil is an example — have even more complicated VATs with a manufacturing-stage VAT at the federal level (similar to India’s central excise duties system, called CENVAT), origin-based VATs (exports taxed, imports not taxed) at the state level, and separate taxation of goods and services.
Choosing an inefficient GST could easily cost India at least 1 to 2 per cent of national income due to unwarranted administrative and compliance costs, and avoidable economic distortions. The success of newly-elected Prime Minister Narendra Modi’s fiscal policy rests on making the right choice.
Modi’s vision for market-led growth will require the formation of a single Indian common market in which businesses can source anywhere, manufacture anywhere and sell anywhere. But India’s current indirect tax system is a serious impediment to achieving this goal. Its complexity is baffling and its incidence highly capricious and indeterminate. Of the many expert analyses of the system’s shortcomings, most call for the introduction of a broad-based dual GST levied concurrently by the Centre (CGST) and each of the states and territories (SGST).
Experience indicates that a modern GST without serious flaws can be introduced only once — the first time. Mistakes made at introduction are hard — indeed, almost impossible — to undo. A modern GST requires that all goods and services should be taxed at a single rate. The GST should not be used to influence the distribution of the tax burden, to protect domestic industry, to favour labour-intensive production processes or for any other ‘worthy’ socioeconomic goal. A modern GST is simply a revenue-raising instrument, though this isn’t to say that the revenue it raises can’t be spent on worthy social or economic goals.
A number of principles should be heeded in introducing a modern GST in India.
Firstly, tax empowerment (assignment) should be separated from tax design issues. Both the Centre and the states should be allowed to tax all goods and services under their GSTs, including imports and excisable goods. This would ensure equal treatment, economic efficiency and administrative feasibility.
Following the resolution of the tax empowerment issue, the Centre should proceed to introduce its own GST (modelled after, say, the New Zealand or South African GST) on as broad a base as possible and apply a single, uniform, low rate of, for example, 5 per cent. It would then be up to the states whether to introduce their own GST, with the federal government collecting the revenue and disbursing it, or to retain their own (modified) VAT. This is exactly the way the constitutional logjam was addressed in Canada.
Unprocessed foodstuffs should also be taxed, but the subsidy on publicly distributed foodstuffs could be increased to compensate the poor for the increase in price. If this is not possible, exemption (perhaps in conjunction with the zero-rating of selected agricultural inputs) would be preferable to a zero rate on foodstuffs. Following the introduction of a broad-based GST, the federal government and the states could impose their own excises on smoking, drinking, gambling, energy consumption, polluting and driving, or agree on who should tax what.
Secondly, uniformity of GSTs levied by the federal and state governments is not necessary or desirable. Although uniformity might minimise conflict, ease interpretation issues and generally make life somewhat easier for registrants, doing so would ignore the fact that states have already introduced their own VATs, which differ from each other.
Perhaps more importantly, best-practice GSTs should differ from one state to another. Some diversity (which is not the same as a free-for-all) would still be consistent with a common market if care is taken that interstate exports are freed of state-levied GSTs. Identical GSTs might lock federal and state governments into a system that would become undesirable over time, but which would be subject to a de facto unanimity requirement like the European Union’s —increasingly anachronistic — common VAT directive.
Thirdly, an integrated GST to tax interstate transactions is not necessary. State GSTs should be levied on a destination basis. States would continue to zero-rate exports. Obtaining a refund for the SGST paid would solely be a matter between the business and the state government. In essence, there would be no break in the GST collection chain — the CGST would still be applied. There would be a break, however, in the state cross-border audit trail, but this would be repaired by the Centre’s nation-wide audit trail, as in Canada. In India, the CGST would serve as a central audit agency, which can monitor interstate transactions.
Finally, the new GST needs an accounting approach to monitor compliance. This approach is quite different from the control system currently in place for the CENVAT, which is mainly about classification and valuation: defining manufactured goods by reference to some nomenclature, imputing a presumptive retail value to them and applying the tax rate.
By contrast, under the GST, the tax would be based on actual, not presumptive, prices and compliance control would be exercised through checks on books of account. In other words, auditors are needed, not excise tax officers. The implications of this profound change in organisation and methods should not be lost on policymakers.
Reforming India’s indirect taxation system will no doubt be complicated, but getting it right should be a priority for the incoming government.
Sijbren Cnossen is Academic Partner of CPB Netherlands Bureau for Economic Policy Analysis and Professor of Economics at the University of Pretoria; he is Emeritus Professor at the University of Maastricht (Economics) and Erasmus University Rotterdam (Tax Law) and has held appointments at the universities of Harvard, New York, and Florida.