Oxus Sux!, O!!: No Proof Required.

Indian Express carried an article in Saturday’s, 04 February 2012, edition titled: “The real scandal: Non-Congress states have performed better in the MGNREGA corruption index (Same column also appeared in Financial Express as: Corruption by Another Name)”. It was contributed by Surjit Bhalla, Chairman of Oxus Investments, who runs a regular column that is generically called “No Proof Required”. The 66th round of NSSO has classified the rural households that sought employment under Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS) into *poor* and *non-poor* based on Per Month Consumption Expenditure (PMCE) as reported by the sample population surveyed. Based on this data, the columnist has constructed the following table.

The simplistic corruption index (column 6 above) conceptualized here is the sum of “difference between percentages of poor & non-poor rural households that sought work (columns 4a-4b)” and “difference between percentages of wage expenditure on poor & non-poor (column 5a-5b)” under MGNREGS. Conceptually such addition perpetrates double accounting. This arithmetical gymnastic of adding the two differences of percentages of “rural households participation” and of “wage expenditure” under MGNREGS does produce a new set of numbers; but besides calling them grandiosely *corruption index*  these are really useless for comparing real performance of the states considered in the column. For example, compared to the percentage of rural poverty, the participation of poor in MGNREGS is very low or low in case of Bihar (50/12), Maharashtra (28/7), or Jharkhand (43/20); whereas it is very high or high in case of West Bengal (31/54), Rajasthan (23/72), Andhra (17/51), Tamilnadu (19/50). These significant deviations do not reflect anywhere in the analysis nor is any attempt made to account for them. The corruption index is neither to scale nor does it have internal consistency in highlighting the diversion of MGNREGS funds from poor to non-poor; the latter, the columnist holds is a case of corruption.
Before one considers the validity of the charge of corruption, it would be useful to put the same data to more conceptually coherent analysis to see if some genuine sense can be made out of it. Same data with some modification has been presented in the following table:

Originally, the NSSO data gave the percentages of poor & non-poor households that sought work under the MGNREGS out of the total rural sample surveyed. The column 3 in above table gives percentage of total (poor and non-poor) rural households that participated in MGNREGS out of the sample surveyed. Based on this figure, the percentages of poor and non-poor households participating in MGNREGS have been converted to base 100 (column 4).  Now the new corruption index is either “Actualwage expenditure on poor less the expected wage expenditure on poor based on their % household participation” or “Expected wage expenditure on non-poor based on their % household participation less the actualwage expenditure on non-poor”. In other words, it is either what poor are paid less than their participatory share would dictate or what non-poor are paid more than their participatory share entitlement. The table is ranked by re-casted corruption index. The rationale behind new corruption index is as follows. Theoretically, wage expenditure on poor and non-poor should be directly proportional to the % participation of poor and non-poor households- Equal Work, Equal Pay. Moreover, in the absence of bias, over a large number of cases the quality and quantity of work done by poor is expected to be the same as that of non-poor. But if poor have got less than their share and non-poor have got more than their share, then that is a case of clear discrimination and diversion of funds. Here one can legitimately question if any malpractice has occurred in the administration of the scheme. Index value of Zero (0) indicates fair administration of the scheme, +ve value shows bias in favour of poor, and –ve value shows non-poor have been favoured.  Since, except in the case of Chhattisgarh (where index is close to Zero) the index is in the negative territory for every state under consideration; it is quite clear that poor have been discriminated against everywhere else. It blows the bottom out of the carefully constructed theory of the columnist that the Non-Congress governments are less corrupt in the administration of MGNREGS- Karnataka goes to the bottom of the heap, Orissa does worse than Maharashtra, Uttar Pradesh & Jharkhand who were tied to the same spot now find separated by Gujarat & Orissa. The *neat picture* that the columnist had *constructed* stands utterly ruined.
There is another useful way of looking at the data by linking it to the state wise per capita income. It would have been meaningful to have state wise per capita income of rural areas, but such figures could not be found. Consider now the following table.

This table is ranked by per capita income from lowest to highest. The rows marked in green have states where percentage of poor households participating under MGNREGS is more than rural poverty, and the opposite holds true in case of rows marked in red. The numbers for Rajasthan have been marked in RED, because its data as given by the columnist is erroneous. Except again Chhattisgarh, all the states that were given +ve values on the original corruption index indicating *no corruption* according to the columnist are grouped together at the top. These states are also marked by lower per capita income when compared to others. Can one therefore impute a conclusion that lower per capita income results in no or low corruption?  This shows how illogical and arbitrary the corruption index constructed by the columnist is. It is as if he handed out the indictment first and later went out to construct the edifice of numbers to prove it.
Coming to the qualitative criterion used for defining corruption, it is based on columnist’s presumption that the non-poor households who participate in MGNREGS constitute corruption. This presumption is quite ill founded. The objective of the MGNREG Act is quite unambiguous: “to enhance livelihood security in rural areas by providing at least 100 days of guaranteed wage employment in a financial year to every household whose adult members volunteer to do unskilled manual work”. Every household means every household, irrespective of its PMCE. The act is blind to per month consumption expenditure (PMCE) of rural households (included in NSSO survey) based on which columnist argues to make out a case of corruption. That means he has not understood either the act or what constitutes corruption.   The act provides bare minimum survival safety net by offering wage, mostly below the minimum wage prescribed under relevant acts in different states, an anomaly the Supreme Court has asked the government to rectify, for just 100 days in a year. NSSO survey has thrown up useful data linking PMCE to MGNREGS, and it would be legitimate to investigate as to why even some of those households in rural areas with PMCE, which suggests access to better livelihood opportunities, still choose to participate in MGNREGS that offers much less.  But instead of choosing to honestly investigate, the columnist chooses to blindly (?) castigate based on an erroneous presumption. Why does he choose to write a column that earns  a *pittance* even in a major national daily, when his own PMCE is expected to be abnormally high compared to that of even better off Indians? Is it to influence & corrupt the minds of the policy makers and drive the debate in some *desired* direction suited to his class? If not, what is it?
In fact the hidden agenda behind such *scholarly analysis* is revealed right upfront in the second paragraph: “Less talked about, actually hardly, is in the name of the poor corruption. This is corruption in programmes meant for the poor. Indeed, when reference is made to the possibility of corruption in schemes like MGNREG, the refrain is “Why are you criticising money spent on the poor? What about telecom licences, diesel subsidies, and other expenditures in the name of the rich?” These questions are worthy of an answer; perhaps not, because corruption is corruption no matter in whose name it is pocketed”.
Reams have been written about corruption in schemes like MGNREGS or PDS (Public Distribution System). The concept of social audit, whereby intended beneficiaries or affected villagers would audit the system, emerged because of the corruption in the administration of MGNREGS. Therefore, it is insincere to suggest that corruption in the name of the poor is less talked about when every pretender to remedy for *growth & development* talks of it unfailingly. It is true that corruption anywhere & everywhere is inexcusable. But when corruption in MGNREGS or PDS is under scrutiny, the hidden agenda almost always is to have one way or the other such schemes abolished or to have them converted into mere show pieces. Providing wage security or food security to poor is held as a millstone around the neck of GDP growth; but subsidizing the rich is held as rocketing GDP growth. Though the article avowedly aims to target corruption in the name of poor in MGNREGS, its real target is quite obviously the pro-poor scheme itself. It is a clever attempt at influencing the discourse to have the baby thrown out with the bath water.

One Response to “Oxus Sux!, O!!: No Proof Required.”

  1. Rahul Banerjee Says:

    nicely done. corruption is the diversion of funds away from the workers to the administrators and politicians but this is not reflected in the NSSO survey.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: