Calculating the per capita cost of corruption in India
Gautam Pingle Gautam Pingle | 08 Dec, 2011
Calculating the per capita cost of corruption in India
They (officials) even seek to predominate over the king, and accepting bribes and practising deceit, obstruct the business of the State. They cause the State to rot with abuses by falsifications and forgeries
—Bhishma to Yudhishthira
BRIBERY
The extent and incidence of corruption in India is a matter of acute interest and agitation. The general view is that it is widespread and endemic, yet there is very little quantitative data on a national scale. Transparency International (TI) has collaborated with the Centre for Media Studies (CMS) to carry out two sector-specific, population-targeted corruption studies on a national basis. These are the only quantitative studies available at this time.
GENERAL SURVEY
The TI-CMS (2005) survey sampled 14,405 respondents spread over 20 states, covering 151 cities and 306 villages. It was conducted on an exit basis—that is, the sample was composed of those who had used or attempted to access the service in question. The respondents’ feedback also consists of the amounts they claim to have paid as bribes.
Table 1 shows that the proportion of respondents perceiving corruption in government departments is generally much more than the proportion of the same respondents claiming to have paid bribes to that department. Only in the case of the police and rural financial institutions did perception and reality come close to each other. In both user perception and bribes paid, the police rank highest—far outstripping other sectors.
Table 1 gives estimates for the entire population extrapolated on the basis of the sample results. On this basis, nearly 30 million households have paid bribes estimated at Rs 2,017 crore in 2005 for accessing hospital services. Similarly, schools are estimated to have collected the largest amount of bribes (Rs 4,137 crore)—though only from 15 million households. The police comes a close second in bribes collected (Rs 3,899 crore)— from 25 million households. Total bribes paid by the population are estimated by TI-CMS at Rs 21,068 crore annually.
INCIDENCE
To put the estimated total bribes paid in context, we need an estimate of the number of households and their expenditure. The National Sample Survey Organisation (NSSO) estimated consumption expenditure directly for 2004-05, where nearly 208 million households (number of persons divided by household size) had an estimated total consumption expenditure of Rs 805,476 crore.
Taking the TI-CMS (2005) figure of the amount of bribes paid (Rs 21,068 crore) and the NSSO estimate of total households as a base, we get a crude figure of Rs 1,013 per annum paid by an average household with total annual consumption expenditure of Rs 38,725. Further, bribes thus account for the equivalent of 2.6 per cent of annual household expenditure. While this may not be of great magnitude, the overall effect is to undermine the governance system and de-legitimise the State. If this is the condition of the general population, then what of the poorest?
BPL HOUSEHOLDS
The TI-CMS study on Below Poverty Line(BPL) households in 2007-08 covered 22,728 randomly selected BPL households. Table 2 indicates the perception and experience of BPL households.
Total bribes paid by BPL households was worked out on the basis that there were 53.7 million BPL households. In 2007-08, BPL households seem to have paid bribes of Rs 164 per household per annum or only 16 per cent of what the general population paid in 2005. This figure seems rather low, but bearing in mind the lower incomes of BPL households, there may be some validity in the estimate.
COMPARISON
We are also in a position to compare perceptions of corruption and the reality of bribes paid between general and BPL populations. Despite the problem of the data being about three years apart, a rough idea may emerge of the difference between the two groups. With this caveat, Table 3 consolidates data from Table 1 and 2 to indicate the differences in perceptions and experiences of general households—as recorded in TI-CMS (2005)—and of BPL households—as in TI-CMS (2008).
BPL household perceptions of common institutional/sectoral corruption are lower than those of general households—except for banking and the police, where they are similar. As far as direct experience of respondents having paid bribes goes, again the proportion of BPL households claiming to have paid for services is much lower than the proportion of general households claiming to have done the same. Only in land records do both groups share a similar order of magnitude in both perception and experience of corruption.
CONCLUSION
The TI-CMS results analysed here raise more questions than answers. This private initiative should be taken forward by a similar exercise by the NSSO. This will give the Government and public more robust figures than are currently available and which can be used to direct anti-corruption measures and monitor them on year-to-year basis.
The author is director of the Center for Public Policy and Governance at the Administrative Staff College of India, Hyderabad
More Columns
Lefty Jaiswal Joins the Big League of Kohli, Tendulkar and Gavaskar Short Post
Maha Tsunami boosts BJP, JMM wins a keen contest in Jharkhand Rajeev Deshpande
Old Is Not Always Gold Kaveree Bamzai