Services tax benchmark should be 'nil list'

The long-awaited ‘concept paper’ on taxation of services based on a ‘negative list’ is indeed refreshing. Rather than merely stating the government policies already decided, the paper provides a balanced discussion of the issues that impinge on the choice between the current approach of selective taxation of only those services identified in the ‘positive list’, and the alternative of taxing all services except those in the ’negative list’. The paper provides a good base for an open and healthy debate.

The starting point in the debate is the recognition that the ultimate benchmark for a good service tax is not one with a ‘negative list’, but that with a ‘nil list’ of exemptions. Thus, a negative list must necessarily mean fewer exemptions. Under the value-added structure of the service tax, exemptions at best provide only a partial relief, and often lead to more tax. Paradoxically, the more you are exempted, the more you pay in tax. This is because exemptions create ‘tax cascading’ through blockage of tax credits on inputs that go in the production of exempted services.

Cascading, which affects only domestic suppliers, creates a bias in favour of imports, and discourages investment and economic growth. To avoid cascading, exemptions must be fewer, and full and immediate credits must be allowed for inputs used in the provision of taxable services. The negative list approach would serve little purpose if the government continues the current practice of denying or delaying input credits, simply to meet its revenue targets. For example, currently the service tax is applied to commercial rentals of real property, but credits are denied for inputs that go in the construction of the property.

The other canons of a good tax are neutrality, simplicity and fairness.

For neutrality, tax should apply at a uniform rate on all competing goods and services. This, in turn, requires that the tax be applied at a single rate to all goods and services. The current tax, while applied at a single rate, fails the second test because of multiple services remaining outside the tax net.

Simplicity also requires a comprehensive base and seamless taxation of both goods and services. The service tax is fraught with ambiguities about the definition of taxable services, giving rise to numerous classification disputes. These disputes can be avoided substantially under the alternate approach, provided the negative list is small.

Still, the goal of seamless taxation of goods and services cannot be achieved under the current Centre-State division of taxation powers. Increasingly, goods and services are supplied as composite bundles, referred to as ‘works contracts’. Under the current constitutional arrangements, such bundles are bifurcated into goods and services, with the former attracting the state VAT, and the latter the service tax of the Centre. This bifurcation is a significant source of complexity, and would continue pending the Constitutional Amendment for the levy of the Goods and Services Tax (GST).

Pending the GST, the governments should resist temptation to adopt policy measures which amount to double taxation of the same base. The recent amendment by the Centre to extend the service tax to meals served in an air-conditioned restaurant, licensed to serve alcohol, is an example of such policies. Notwithstanding that the Constitution deems such supplies to be sale of goods within the exclusive domain of the states, the Centre has staked claims on the same base on the grounds that they also include an aspect of service.

The canon of fairness calls for the tax to be progressive and minimising disproportionate burden on lower-income households. Historically, governments have responded to this objective by providing exemption for items consumed by lower income households. However, exemptions fail to alter the overall distribution of a consumption tax. This is because an exemption benefits all consumers, regardless of their income status. Moreover, they benefit upper-income households even more. Given their higher purchasing power, their spending on exempted items is more than those by the poor. Thus, under modern consumption taxes, the fairness objective is addressed instead through targeted income supplement or expenditure programmes.

In this context, it is debatable whether the negative list should include all of the items that it does. There could be compelling grounds for exempting them, but the exemptions would not enhance the fairness of tax or make it progressive. Lower income families depend on government-provided healthcare and education services, which are already subsidised or provided free of charge. Moreover, their inclusion in the negative list would mean that no credit would be allowed for the tax charged on their inputs. This blocked input tax would be embedded in the cost of the services provided, thus leading to a regressive outcome. If relief is to be provided for such supplies, it should be in the form of zero-rating (i.e., no tax on output and full credit for input taxes), not an exemption.

Administratively too, exemption to health and education services would generate definitional issues and classification complexities, given the diverse nature of the services offered.

And finally, the criterion of revenue potential of the negative list approach is misplaced. The debate on whether or not India should go for the negative list approach should be guided not by revenue considerations, but by the canons of efficiency and neutrality in taxation. The revenue goals should affect the tax rate, not the tax structure.

The concept paper argues that there is significant potential for additional revenues from service taxation, which account for 63 per cent of GDP. However, consumer services, which are the proper base for a non-cascading tax, account for only 20 per cent of GDP.

The author is a tax partner at Ernst & Young and the views expressed are personal