How are Budget Estimates Made?

By Vijay Kumar

THE Budget Estimates for 2014-15 were:

2014-15 Budget Estimates

But how on Earth did the Finance Minister arrive at these figures?

The Tax Administration Reform Commission (TARC), asked the two Revenue Boards to explain the tax forecasting methods adopted by them. And both the Boards explained.

CBDT : There are two factors which form the basis for estimation of BE. First is the revenue collection during the past three years. Second is the corresponding GDP growth during these three years. Further, the GDP growth for the relevant year, i.e., the year for which the Budget Estimates are being made is also considered. For the purpose revenue estimation, the rate of growth of direct tax, separately for Corporate Income Tax (CIT) and Personal Income Tax (PIT) is taken for the past three years. Buoyancy is worked out for each of these three years separately for CIT and PIT,which is defined to be:

Buoyancy = Tax growth rate / GDP growth rate

The forecast of GDP growth rate for the relevant year for which the estimates are being made is provided by the Department of Economic Affairs (DEA). The average buoyancy for PIT and CIT is multiplied by GDP growth rate to arrive at the tax growth rate for the purpose of projecting revenue estimates. This is basically done for gross collections of tax. An estimate based on the previous year's collections, which are major head and minor head wise, is also taken into account before arriving at the final estimate. The estimates made by the TPL division are subject to the final approvals of the same, taken by Budget division of the DEA.

The above method for revenue estimation is based on historical data and uses the GDP growth rate of the year in reference to arrive at the Budget Estimates. Over the past five years, the method has been tested against actual collection and in all years, the difference between actual collection and revised estimates has not been more than three per cent.

CBEC : the work relating to indirect tax revenue estimation/forecasting at aggregate level is carried out by the Tax Research Unit (TRU) for preparation of the budget estimates (BE) for indirect taxes. Before enumerating the exact methodology used for indirect tax revenue estimation it is necessary to understand the complexities involved in such estimation in the prevalent tax regime in India and why the standard text book economics and statistical methodology, based on pure economic criterion and a set of standardized assumption parameters, cannot solely be used for such estimation. Even though tax estimates based on buoyancy factor may turn out to be more realistic in the case of direct taxes than indirect taxes (sic). This is because direct taxes are progressive in nature in comparison to indirect taxes.

Indirect tax revenue at central level accrues from three taxes, namely, Customs, Central Excise and Service Tax, each of which is influenced by different factors which are described below:

a) Customs revenue in a particular year is a function of import and export volumes and the policy changes made in the fiscal policy. However, prediction of import and export volumes at the estimation stage is highly complex as these are a function of international prices of imported goods, monetary exchange rates, world economic and political scenario, FTAs entered into by various economic groups during the course of the year, trade restrictions imposed by importing/exporting countries, country specific tariff barriers imposed in terms of antidumping and safeguard duties and mid-year policy changes that may be implemented to address various concerns such as that of current account deficit. In recent times there have been huge volatility in the international prices of key commodities and also in exchange rates.

b) Central Excise duty is levied on manufacture. As of now about 60% of the total Central Excise revenue comes from specific rated commodities (like petroleum & products, tobacco product,sugar etc). Rest of the revenue comes from ad-valorem rates. A tax on manufacture cannot have a predictable co-relation to GDP numbers and tax buoyancy in the previous few years for the following reasons:

1) Manufacture does not have a predictable co-relation with growth in GDP. With increased FTA, there is likelihood of import substituting for manufacture. This is unlike consumption of goods and services, where there is predictability.

2) While a majority of excise revenue is a function of quantity of production, GDP numbers are in value terms. Therefore, revenue from these commodities cannot be based on GDP estimation. Further, production numbers of specific rated commodities cannot be predicted for the reason that this may vary for various reasons like ban imposed by the states on tobacco products or changes in consumption pattern.

3) While taxes imposed on income side have a predictable co-relation with the GDP, inflation and other economic indicators, the taxes on expenditure side are dependent on the consumption pattern, which in itself is a function of price elasticity of a commodity, the saving rate, etc.

c) Service Tax is a tax on consumption of services. Services constitute major portion of the GDP and consumption of services has a co-relation with GDP. Therefore fair estimation of service tax revenue is feasible by economic forecasting technique based on tax buoyancy/tax elasticity method with certain underlying assumptions. However, this estimation could be much more accurate if service sector is comprehensively taxed. Even with the introduction of the Negative List based tax regime a significant portion of the service sector (illustratively, a significant portion of public services/government services, health, education, public transport by road and rail, non-commercial construction, sports, entertainment and cultural services, financial services, agricultural sector related services and transport of goods) is outside the tax net.

The indirect tax revenue forecasting/estimation is done within the above stated constraints using the top down model. Broadly, the steps involved are,-

(i) The sectoral analysis of revenue estimates taking into account the past trends, growth prospect, the impact of policy level changes, sector analysis available in credible publication and the information provided by the respective ministries e.g., the growth projection in petroleum sector is provided by the Petroleum Planning and Analysis Cell (PPAC), Ministry of Petroleum and Natural Gas.

(ii) The import/export volume growth is estimated based on trends.

(iii) Feedback and revenue projection is taken into accounts from all zonal offices (CCs) in the 4th quarter.

(iv) Sectoral analysis of top 10 commodities/sectors w.r.t Customs, central excise and service tax.

(v) Based on the above inputs the numbers are consolidated and validated with nominal/Real GDP growth as reported by CSO, Ministry of Statistics, taking into accounts (sic) the likely outgo in refunds/Drawback, etc.

(vi) The estimates are finalized after due deliberation with the Budget and Economic Affairs Division taking into account the overall economic and fiscal scenario.