I want to be on the side of the angels, with the people rooting for the new employment guarantee Bill.
And I am mindful that people like me criticised MG Ramchandran's free mid-day meal scheme for schoolchildren in Tamil Nadu a quarter century ago; the result of that was not fiscal impoverishment of the state, as many warned, but dramatic improvement in Tamil Nadu's literacy and population growth rates. Critics of populist politicians have to be careful.
So I am prepared to ignore the facts and statistics dished out very effectively by Surjit S Bhalla in this newspaper (Business Standard) to argue that the Bill is founded in lies concerning unemployment and poverty.
I am prepared to concede that inequality in India may have increased -- perhaps because we do nothing about the quality of governance in our poorest states, while the dynamic states pull ahead. I also concede that many poor people will welcome the chance to do a day's manual labour for Rs 60 in the slack agricultural season.
My problems lie elsewhere.
Jean Dreze of Sonia Gandhi's National Advisory Council (the chief proponent of the Bill) has in The Times of India responded to commentators who have been sceptical about the Bill; as it happens, most of them have aired their views in Business Standard.
Dr Dreze is initially dismissive, but he does eventually examine whether the critics have a point. And he makes the breathtaking admission that in five of the six states studied, there were no muster-rolls under the food-for-work programme.
You know what that means. Does Dr Dreze have a solution? Yes, we need civil society activists, as exist in Rajasthan. Is it realistic to expect the replication of Rajasthan's NGOs? We are not told. So is there a substantial risk that wholesale leakage might mar the employment guarantee programme?
The riposte from another defender of the Bill is that we don't stop defence purchases because of corruption, so why stop a jobs programme?
I think that is conceding the point to the critics: there will be large-scale corruption. If so, will it address poverty and unemployment? To some extent, for sure. But, I suspect, to a very limited extent. It was Rajiv Gandhi who talked of 15 per cent efficacy levels.
Sonia Gandhi, asserting in Parliament her ownership of the Bill, declared that an economy growing at 7 per cent can afford an employment guarantee programme. I share her sentiment, India has to begin addressing social security issues.
But on the scale of Rs 40,000 crore (Rs 400 billion) right away, without ensuring it will be well spent? I didn't know that rapid growth provided an excuse for wasting resources. One response has been that banks are giving bad loans to rich businessmen, and that money should be recovered before criticising the new Bill.
This is a poor level of debate: if you give loans, some will go bad. Does the Indian banking system's bad loans level compare poorly with the levels in other similar countries? Not so.
The government has put in a safety clause: if corruption is detected, the programme can be stopped in the concerned area. Dr Dreze has attacked this. Assume that it survives the attack and becomes part of the law. What will this mean in practice?
We know that the number of people reflected in the total number of ration cards issued exceeds the country's population. Evidence of corruption, but is there corrective action?
We also know that the bulk of what is handed out under the subsidised public distribution system is not in the areas where there is a preponderance of poor people. Is there corrective action? The system is interested in the government spending, not on the quality of spending.
The truth is that the leading lights of this government don't believe in the wisdom of this new law. They have tried every trick in the book to delay it, to minimise its scope, to put in escape clauses.
They have been blocked at every turn by the NAC. I fear that scams will follow, and the intended results will not be achieved. Sonia Gandhi may prove to have MGR's level of wisdom. But I rather doubt it. I hope I am wrong.