TAX EVASION AS CORRUPTION

From Business Standard September 10, 2011
INDIA’S social fabric has been dented in the past month over the issue of corruption. Corruption has many faces, one of the more important of which is tax evasion. Tax evasion can be unilateral, on account of taxpayers alone, or bilateral, through collusion between taxpayers and the tax collector, thus robbing the state of its due revenue, or multilateral, where a society participates in it through resistance to remedial measures.

Tax evasion leads to inefficient and inequitable outcomes. Because tax evasion erodes tax revenue, the lost revenue is attempted to be made up through high tax rates. An example is real property tax rates in India that range around seven per cent, as opposed to developed countries where it approximates one to two per cent. Note, therefore, that paradoxically, an estimate of tax evasion based on this elevated or artificial or compensating tax rate would yield an exaggerated number. That is, if the tax rate were lower, tax evasion could also be lower. Thus, the consequence of those who get away with it is an extra burden of taxation on those who pay tax. This is an inequitable impact across taxpayers.

The sectors or activities that are amenable to tax evasion will fetch more investment, which would be redirected from other areas. And this reallocation of resources takes place often from more to less productive sectors. Thus, if investment in land is prone to evasion, more resources would go into construction than in the production of rail engines or food. Construction as an activity ends when a building or bridge is completed. Manufacturing and agriculture, by contrast, are more long-term and generate a longer time horizon of employment and supply of the product. Economists call this the outcome of inefficient allocation of resources.

Through time, scholars and governments have attempted to measure tax evasion. Estimation methods have used national accounts or expenditure survey information to assess the size of economic activity and compared it with reported activity. The difference between estimated and reported activity could then be related to the loss of tax revenue associated with it.

In India, tax evasion has not been directly estimated. In Latin America this was carried out widely in the 1980s and 1990s (Aguirre and Shome for Mexico, Serra for Chile, Shome for Colombia, and others). In the case of individual income tax, the following method would apply.

To personal income declared for tax purposes, adjustments should be made for those components of income that are included in the concept of income in the national accounts but are deductible for tax purposes, for example, personal exemptions, deductions, savings incentives. Adjustments need to be made for individual tax brackets if the tax structure is progressive. The resultant series of gross taxable declared income (by income class) should then be compared with gross taxable income from the national accounts to obtain an estimated undeclared income.

Similarly, for corporate income tax, adjustments to national accounts-based income for deductions for the array of tax incentives that are legally applicable, as well as for special conditions for small and medium industry, would have to be made, and then compared with declared taxable income from returns. To simplify the exercise, instead of national accounts, data by the Centre for Monitoring Indian Economy could be used to obtain more reliable conclusions, albeit from a more truncated information base.

For the VAT and CENVAT (or central excise), using an input-output table for the economy would be the ideal option. The widest theoretical VAT base is all purchasable goods and services, or GDP plus imports minus exports. However, given that the VAT base usually has a plethora of exemptions for not just the supply of outputs but also for the inputs of exempted outputs, it is not sufficient to use only GDP data. Instead, one has to go upstream and start with production data against which taxable inputs would have to be credited out since the VAT base is value added, or output minus input.

Tax evasion should be estimated in India. Another matter is how to limit tax dodging (Tanzi and Shome). The first means should be to focus on those taxes that the tax administrator can monitor more cheaply or easily. For example, to begin, it may not be fruitful to spend a lot of administrative resources on monitoring small taxpayers. Thus, the service tax did away with 80 per cent of taxpayers from the bottom. The second is to design taxes intelligently. The VAT has a good design since input tax credit is matched between buyer and seller, and allowed only on the basis of physical invoices. The third is to use third-party information for individual income tax as has been introduced in India through Annual Information Returns (AIR) for third parties such as mutual funds, credit cards, property transactions and others who must report transactions they carry out with customers (using PAN numbers) above stipulated thresholds. However, it is crucial that such information is correctly used and interpreted by the tax administration so that a taxpayer does not suffer from erroneous matching of such information.

A final word. Tax evasion is not the only form of corruption. Corruption has more tentacles. According to Transparency International, India’s Corruption Perception Index (CPI) ranking worsened from 72 in 2007 to 87 in 2010 amongst 180 countries. It is worse than Brazil at 69 and China at 78, but better than Mexico at 98 and Russia at 154. Though CPI does have flaws, it gives an indication of impediments to economic and social growth and equity. A worsening index is detrimental to attracting foreign investment and political stability. On a broader canvas, it is important to investigate what has been motivating and sustaining corruption before designing draconian measures to combat it. If the cart is put before the horse, the outcome may remain shy of success.

References:

Aguirre, Carlos and Partho Shome (1988), “Mexican VAT: Methodology for Calculating the Base”, National Tax Journal, Vol. 41

Serra, Pablo (1991), “Estimación de la Evasión Tributaria en el IVA", Servicio de Impuestos Internos, Publicación No. 91/05/C, Universidad de Chile

Shome, Partho (1995), Colombia: Fundamental Tax Reform, IMF Occasional Paper No. 123

Tanzi, Vito, and Partho Shome (1993), “A Primer on Tax Evasion”, IMF Staff Papers, Vol. 40.

The author is director and chief executive, Icrier. Opinions are exclusively those of the author. His book, Modernising Tax Administration: Championing Analysis and Specialisms, is forthcoming in the autumn

India needs to develop ways to estimate the cost of tax dodging, says PARTHASARATHI SHOME