Published Editorial

One tax to end all tax?

  • Share

The Hindu Business Line


D.K. Srivastava
Chief Policy Advisor, EY

Talks of introducing a banking transaction tax (BTT) that will remove the need for all other taxes except customs duty have rekindled debate amongst policymakers, industry and the public on the way forward in the country’s stalled tax reforms.

There have been experiments with such tax earlier in Australia, Argentina, and Brazil. In Australia, a bank account debits tax was levied during 1982 to 2002 on customer withdrawals through cheques. Argentina had a bank transaction tax during 1984-1992. Brazil implemented a bank debit tax during 1993-2007. In all these cases, the tax served as a temporary revenue booster but had to be eventually abolished.

Such transaction taxes are usually criticized for the potential regression, cumulativeness, and incentive to transact outside banking channels. Competitiveness of industry is also adversely affected, particularly for export, as zero-rating is not possible. Its basic attractiveness lies in the diffusion of tax burden. The tax rate has to be kept very low and payments become automatic. Even at an extremely low rate, it can produce an attractive volume of revenue. The Brazilian experience showed that at the low rate of 0.38 per cent, the BTT raised a little more revenue than the value-added tax on industrial products.

However, revenue increase is not the only objective of taxation, which is also a potent fiscal instrument for redistribution of income, savings promotion, and curbs on undesirable consumption or use of demerit and polluting goods. India’s tax reform story remains unfinished, both for direct and indirect taxes. In particular, a comprehensive goods and service tax is still out of bounds as State governments fear significant revenue losses. Existing options include dual- or multiple-rate structures, high revenue-neutral rates, and rate bands. Such a GST structure, even if implemented, will be complex and prone to evasion. The direct taxes code, too, is unable to significantly lower marginal rates for fear of revenue loss. High rates lead to exemptions and multiple rates, leading to evasion and extensive classification disputes. Only a low, single rate complemented possibly by un-rebated excise and/or sales tax on demerit and polluting goods in the case of GST, and low marginal rates in the case of personal and corporate income taxes, and the abolition of taxes such as octroi, entry tax and tax on inter-State sales, which create fiscal barriers in achieving an all-India common market, would help achieve a tax system that does not encourage evasion, minimizes classification disputes, augments productivity and promotes industrial competitiveness.

If the fear of revenue loss has obstructed tax reform in India, the case for BTT can be examined. A range of transaction taxes have been proposed theoretically and some of these have been tried empirically. Economist John Maynard Keynes was one of the first proponents of a securities transactions tax to discourage excessive speculation. James Tobin and Paul Bernd Spahn have proposed some versions of a currency transactions tax. An automated payment transactions tax (APT), which includes the banking transactions tax, was proposed by Edgar L. Feige way back in 1989. This includes all transactions through automated payment systems, including those in stocks, bonds and financial derivatives. With the largest possible base, it can provide a given amount of revenue at the lowest possible rate. Available information indicates that financial transaction taxes are being practiced in one form or the other in about 40 countries.

If some form of transactions tax is to be implemented in India on a grander scale than tried before, we should be mindful that by its very nature it will be a centralizing tax with significant revenue implications for our federal structure. Autonomy of States and local bodies in deciding the size of government and tax rate is itself a source of efficiency.

A suitable way forward would be a comprehensive APT at a very low rate, say 0.25 per cent, with the revenue used to complete all the stalled reforms of direct and indirect taxes. We can then examine the empirical features of the system to see whether it would be worthwhile abolishing the traditional taxes or whether the transactions tax should remain.