It is widely expected that this Budget will chart a roadmap and indicate which model of Goods and Services Tax will be introduced in the country. GST has been declared to be the goal for 2010 in the last Budget. There are several models of GST, each with its own merit and demerit. A look at some of the models in circulation:

Australian Model: In Australia GST is a federal tax, collected by the Centre and distributed to the states. But India is a heterogeneous country and there is no chance that states will allow the Centre to collect all the taxes while they become just spending institutions.

Canadian Model: The GST in Canada is dual between the Centre and the states and has three varieties:

(i) Federal GST and provincial retail sales taxes (PST) administered separately - followed by the largest majority.

(ii) Joint federal and provincial VATs administered federally (Harmonious Sales Tax - HST).

(iii) Separate federal and provincial VAT administered provincially (QST) - only for Quebec as it is like a breakaway province.

The first variety is fundamentally the Canadian model, which is similar (though not the same) to the existing situation in India.

Kelkar-Shah Model: This model of a unified goods and services tax, is based on a "grand bargain" to merge Central Excise, Service Tax and State VAT into one common base. Two different rates of tax are to be levied by the Centre and the states. The collection will be by the Centre. This is like the HST model in Canada.

Ajay Shah wants the collection from big industries to be done by the Centre while the states collect it from the smaller industries. They want the model to be introduced straightaway.

This exposition does not even take into account the Constitution in the country. The Constitution allows Central Excise and Service Tax to be collected by the Centre and the VAT (sales tax) to be collected by the states. If the Constitution does not allow such an amalgam, how the model can be operative is not indicated in the model.

Bagchi-Poddar Model: This model, just like Kelker-Shah's, envisages a combination of Central Excise, Service Tax and VAT to make it a common base of goods and services tax to be levied both by the Centre and the states separately. This means that the Central Excise Act will be abolished and the goods tax will be only on the sale of goods. It will merge in it the service tax.

To put this in legal lingo, the " taxable event" for the GST will be the act of sale of goods and services. The concept of manufacture will simply vanish. The difference between the Bagchi-Poddar and Kelker-Shah models is that in the former, the collection is at two levels, by the Centre and the states, while in the latter the collection is only by the Centre.

So while the Kelkar-Shah model is like the Canadian HST, the Bagchi-Poddar one is like the Quebec model.

Although the model says that it is based on the Quebec model, it is actually not fully so as this model envisages collection both by the Centre as well as the states, whereas the Quebec model envisages collection only by the state of Quebec.

The Bagchi-Poddar model also clearly envisages that a Constitutional amendment is necessary to bring the taxing powers on goods and services under the Concurrent List and to abolish the present division of taxing powers between the Centre and the states. My arguments against the combined GST in the Bagchi-Poddar model are the following:

(i) The amendment of the Constitution as envisaged in this model will not be practicable immediately and it will in any case create a huge political upheaval in the country.

(ii) The Supreme Court also may not allow the change in the basic structure of the Constitution because it compromises the fiscal federalism ingrained in the Constitution.

(iii) Revenue: Total collection of revenue will remain the same. There is no reason why we should create such an upheaval for merely a theoretical pursuit.

(iv) Combining the service tax with VAT at the state level is fraught with innumerable problems.

(v) At present the service tax collected at any change is given credit with goods at the central level against duty on goods cleared from any other states. Such interchangeability will not be possible in case the service taxes are collected by the states. The free flow of a common Indian market will be hindered.

(vi) The states may impose different rates of service tax on the same service.

(vii) A service having an all-India character will get different treatment in different states. Even the model itself admits that its success assumes complete similarity of rates and procedures in all the states, which is most unlikely to happen. The states could not even enact the same VAT laws recently.

The Practical Model: The same result with no upheaval/without upsetting the present setup can be achieved by a dual VAT or parallel GST at the central as well as the state levels. At the central level we can have, as we have now, a combination of Cenvat and Service Tax. At the state level we can have VAT alone without Service Tax.

There is no need to combine Cenvat and VAT which envisages the complete abolition of Central Excise Act, which gives the power to the Centre to charge tax on manufacture. At the Centre the merging of Cenvat and Service Tax has been already done to a large extent by allowing interchangeability of input credit for both goods and services.

The rate of tax can be made 14 per cent for both goods and services in the next Budget or the one after that. At the state level, VAT can be perfected by abolishing CST and allowing inter-changeability of input credit between states. This will work administratively as well as revenue-wise.

Realising that India is a federal country with disparate states, this dual or parallel GST and VAT is a most practical proposition. There is no need to serve a doctrinaire approach and create a mayhem to get the same revenue.