The dissension among the states on the constitutional amendments for a goods and services tax (GST) is neither surprising nor unexpected. The changes proposed are of far-reaching importance, and would have a deep and substantial impact on the evolution of the Indian federation in both the economic and political spheres.

They relate to the taxation powers of the Centre and the states, which, in a democratic set-up, are as basic as human rights. It is these powers that allow elected governments to define the revenue sources for their essential programmes and policies, and make them accountable to their constituents.

The amendments have three components: (i) grant of new powers to the Centre and the states to levy a GST, (ii) deletion of powers to levy taxes that are to be subsumed by the GST, and (iii) empowering the finance ministers to collectively determine the design of the GST.

The states have expressed reservations mainly about the third componentthe restriction on their ability to unilaterally change the GST design and the grant of veto power to the Union finance minister.

The key issue emerging in the entire debate is the need for balancing fiscal autonomy with harmonization of GST design. A dual GST levied by both the Centre and the states is workable only if the design is harmonized.

The unfettered exercise of fiscal autonomy by individual states can lead to a sub-optimal outcome for all. It is like unregulated driving by one person, which can become a hazard for others on the road. Similarly, the rules requiring fiscal harmony are necessary for the collective enjoyment of fiscal autonomy by all.

In this context, Canada, Australia and the European Union provide interesting examples. In Canada, no amendments to the constitution were required for the levy of the dual GST, as it already empowered both levels of government to do so. GST harmonization was then achieved through binding inter-governmental agreements outside the constitution. The agreements provide for common GST design and administration. Provincial participation in the GST is elective, but they have no autonomy to change the basic tax parameters.

In Australia, the constitution empowers only the Centre to levy a GST, and no amendments were made to rectify this vertical imbalance in Centre and state taxation powers. Instead, the GST is levied by the Centre as a single national tax, but the entire GST revenue is given to the states.

The states agreed to subsume most of their transaction taxes which were giving rise to significant economic distortions. This arrangement, binding on all states, is governed by an inter-governmental agreement outside the constitution. The GST design is determined by a ministerial council, by consensus of all governments.

In the EU, individual members have full fiscal autonomy to levy taxes as sovereign states. However, each member state is required to levy a value-added tax (VAT), which conforms to the EU VAT directive. The EU directive, approved by unanimous consent of all member states, gives some flexibility in VAT implementation, but mandates significant harmonization of tax base and rules. Member states have flexibility to set rates, but not below the specified floor rates.

The common thread in the three models is the requirement of inter-state harmonization, which necessarily entails a constraint on fiscal autonomy of individual states. Interestingly, the sacrifice of fiscal autonomy is balanced by the requirement of consent of all participating governments. The fact that fiscal autonomy is constrained through an agreement or arrangement outside the constitution may be of symbolic importance, but does not affect the ultimate outcome.

In this respect, the GST framework proposed for India is similar to international models. It empowers both the Centre and the states to replace their inefficient indirect taxes by a more efficient GST. To guard against the risk of uncoordinated taxation by 36 governments becoming a tax jungle, it transfers the tax design decisions to the GST Council.

The states are not in agreement with the proposed set up, so what could be the possible alternatives?

The answer to this would lie in two modifications.

First, requiring the agreement of the Centre for GST Council decisions could be limited to the determination of the Centres GST rates. Other aspects common to both Centre and state GSTs would require only a two-thirds majority of the council. The Centre could propose unanimous consent of the council members for all GST design elements. While this is the rule in the EU and Australia, it runs the risk of extreme difficulty in reaching consensus on any future changes in GST design.

Second, flexibility can be given to the states to levy supplementary taxes on goods other than petroleum and alcohol, e.g., pollutants or goods bad for human health.

At the same time, the provision enabling the levy of GST should be broadened so that it can be levied by both levels of government on all items of consumption, including petroleum, alcohol, electricity, and real estate.

The decision to exempt any sector from the tax should not be within the Constitution, but in the GST legislation, which can be reviewed periodically, without undergoing the complex process of amending the constitution.