The implementation side

The implementation side

The two major tasks that the finance minister has taken upon himself are the implementation of the direct tax code and the goods and services tax (GST), both by 1 April 2011. The draft tax code is already in the public domain, and the adjustments in personal taxation made in this year’s Budget will make it easier to implement the new code. Of these, the GST roll-out will fundamentally change revenue realization, the transparency of tax collection and the reduction of administrative costs. It is likely that there would be significant revenue gains to the Centre as well as the states once GST is in place, most of it coming out of capturing revenue leakages into the system. Implementation of GST, though, appears to be fraught with several obstacles.

There are two reports on GST in the public domain—one put out by the empowered group of ministers and the other by the 13th Finance Commission. I have already commented that the latter has probably been contributed to and prepared by people who have no knowledge and experience of goods and services taxation and its implementation. Further, the recommended rates for GST—at 5% for the Centre and 7% for the states—are unrealistic now that the service tax, the Central excise and customs duty are all at 10%. The recommendations to include liquor excise and power tariff into GST are also unrealistic. Thus, it’s the discussion paper of the empowered committee that is the one that needs to be looked at from the point of view of implementation.

That paper recommends a dual (the Centre and the state) model of GST, to cover all goods and services except those exempt, with separate payments to the Centre and the states. The paper recommends a uniform procedure for filing GST returns across the country.

This outline needs to be fleshed out. The proper way to do this would be from the bottom up, so that implementation is smooth. For example, first, at the level of the assessee, if there are two forms to be filled (the state and the Centre) and two returns to be submitted, the first question is whether he would get two assessment orders for the same tax—and if so, whether there would be two appellate bodies for dispute resolution. In cases of misclassification, it is important to prescribe which of the two would have overriding jurisdiction.

Second, if the procedures are to be uniform, then the state governments as well as the Centre need to agree on these formats and procedures and to enact rules for them; I am not sure whether any significant attempt has as yet been made towards this goal. All the states as well as the Central excise department need to come on board for this, since it means a massive rewrite of all existing procedures.

Third is integrating the tax credits. The GST scheme, once in place, would require Central excise and service tax credits to be exchanged for state service tax and value-added tax (VAT) credits. For this exchange to occur seamlessly throughout the country, there is the need to create immense information technology (IT) architecture. This is neither difficult nor infeasible—we have the financial market architecture as well as the real-time gross settlement, or RTGS, system in banks already in place, something more complex—but we need to make a start.

The important criterion should be to capture all input credits, so that tax evasion and tax avoidance disappears. The model must then be automatic, with easy user interface, incapable of being manipulated by the officials, and transparent. If we think through the IT requirement, one can visualize one or two central servers in each state to handle the transactions—much like the central server for income tax coming up in Bangalore. The important challenge is the interlinking of transactions and of credit, not just between the states but also between the Centre and the states. It is important that a new IT set up is created, rather than tinkering with existing ones—since there are ample differences among the states’ tax structures, not to mention between the Central excise and service taxes.

Once the concept is in place, putting in the actual tariff numbers would be easy. And since Central excise as well as state VAT personnel are already using IT in several states, training the staff would also be easy. Working backwards, one should have a pilot project in place in a district within three months, which means that the IT suppliers need to start working now.

Now that the Budget has made all Central tariffs converge at 10%, we are possibly looking at another 10% or 12% for the states, and a GST rate of between 20% and 22%. This rate is not unreasonable at all, given that a number of cesses and local levies would be subsumed.

This narrative has become necessary, because everyone is talking only in terms of legislation, lists of goods and services and rates. While these are important, no one appears to be looking at how the system will actually work. Hence the need to sketch a feasible roll-out plan.

S. Narayan, a senior research fellow at the Institute of South Asian Studies, Singapore, is a former finance secretary. We welcome your comments at