Column: Rotting food, rotten arguments

Written by Surjit S Bhalla | Updated: Sep 4 2013, 08:37am hrs
Everything populist (popular) is wrong.

Oscar Wilde modified

(ever so slightly)

There have been some questions raised about my estimate of the Food Security Bill costing R3,14,000 crore or approximately 3% of GDPManmonias FSB: 3% of GDP (FE, July 6), http://goo.gl/QhcWMn. In their article Correct costs of the Food Security Bill (FE, August 17)http://goo.gl/9LuDWnthree distinguished economists (Ashok Kotwal, Milind Murugkar, Bharat Ramaswami, hereafter KMR) join together to demonstrate that I was barking up the wrong tree, that my calculations were all wrong and that the true cost of the FSB would only be 18% higher at R85,000 crore. (Full disclosure: I insisted that their article be published.)

The KMR estimates have been boldly endorsed in a spate of recent articles and/or comments, e.g. Planning Commission member Abhijit Sen, articles in Tehelka and Mint, and in cyberspace by leading economist Debraj Ray. These are senior and respected economists/commentators, and it is important that their (wild) allegations be seriously looked into.

In my article, I had provided a simplified short-cut method of estimating the approximate size of the food subsidy. In the table, I present stage-by-stage estimates of the FSBfrom procurement to distribution. Broadly speaking, there are four estimates of the cost of the FSB. The lowest of these is the one by KMR; at R85,000 crore, they are some R40,000 crore less than even the official estimate of R1,25,000 crore.

This is the first time I have witnessed any person, let alone respected economists, estimate leaky government subsidies to be less than even the leaky government of the day has claimed, and not just in India but anywhere in the world. And not leaky by a small amountby 47% (R1,25,000 crore versus R85,000 crore!) I would appreciate if any reader can provide evidence to the contrary.

The third estimateR2,27,000 croreis by a government food official and expert, Ashok Gulati, and his team (hereafter Gulati). Gulati is the chairman of the Commission for Agricultural Costs and Prices, and therefore in a natural position to know better.

The FSB provides a right to 860 million people (bottom 67% in rural areas and bottom 50% in urban areas) to avail of 60 kg of rice and wheat a year for R2.5 a kg and a subsidy on each kg of R20.3. This leads to a total distribution of 51.5 million metric tonnes (MMT) a year at a cost of R1,05,000 crore. In addition, the government has added recipients who will receive extra cash or grain (e.g. pregnant women, special groups) and Gulati estimates this subsidy to be R23,000 crore. The total, at R1,28,000 crore, is almost identical to the government estimate of R1,25,000 crore. But far, far removed from the left touted estimate of R85,000 crore.

Gulati has additional costs (agricultural production enhancement costs, infrastructure and logistics etc) which add up to R54,000 crore. So the base Gulati estimate is R1,82,000 crore.

How does a respected authority like Gulati get a higher estimate of R2,28,000 crore The above R1,82,000 crore estimate assumes that there are no storage costs, no rotting food, and no leakages in distribution. It is as if the government was playing God with efficiency. In their reign since 2004-05, the UPA buffer stocks have averaged 37 MMT which, at a cost of R20.3 per kg, comes to approximately R75,000 crore or R2,000 crore per tonne. Gulati says that no more than 20 MMT are needed, and thus the storage costs will be R40,000 crore a year. The Gulati final estimate, therefore, is R2,22,000 crore. However, if the average historical storage of the UPA is taken (why would they do things differently in the 10th plus years if for nine years they have felt it necessary to store annually 37 MMT), then the basic costs rise by an additional R34,000 crore.

But even the Gulati estimate assumes that the government is perfect, there is no rotting grain, and there are no leakages from the FCI to the ration shop (leakage 1) and no leakages by providing grains to the people not eligible (leakage 2). The governments own estimate of food that goes rot in the night (or leaks to the liquor trade) is R30,000 crore. Which brings a conservative estimate, with no leakages at all, of the FSB costing R2,86,000 crore, more than three times the low-ball, rogue, KMR estimate of R85,000 crore.

The leakages are yet to be accounted for. Do we have any estimate of these Yes, we have an expert on the subject who is none other than the R in KMR, Bharat Ramaswami. In an exhaustive analysis on the Indian PDS system for 2004-05, Ramaswami estimated that only 10% of the PDS money reached the targeted populationrest, all leakage! Curiously, in his new avatar, Mr Ramaswami believes that all leakages have disappeared into thin air, the UPA government came and reached nirvana for the poor, and for him, and these leakages are now zero!

Leakage 1: This occurs before the food even gets to the ration shop. This leakage has been estimated by various authors, for various years, to be approximately 45% of the food that leaves the FCI but does not reach either the poor or the rich. Given the official distribution of around 56 MMT, this leakage is 25 MMT or around R50,000 crore.

Leakage 2: Analysis of those not eligible for food subsidies in 2013 -14 (those above the 67th percentile in rural areas and the 50th percentile in urban areas) with 2011-12 NSS data indicates that there were 143 million non-eligible individuals who received an average of 4.5 kg of grain per month. Leakage 2: R17,500 crore.

Summing up, the Gulati estimate with no rotten food and no leakages was R2,56,000 crore. Add the following: extra storageR34,000 crore: rotten foodR30,000 crore; Leakage 1R50,000 crore; Leakage 2R17,500 crore. These four items add up to R1,31,500 crore. Maybe all these extra costs will be zero, and as the pigs fly, we will observe the minimum estimate of R2,56,000 crore to be right. Maybe the new government will increase efficiency, decrease corruption, and reduce these extra costs by halfa 100% improvement! In which case the FSB, if implemented according to both the spirit and the letter of the law, will be R3,20,000 crore, somewhat higher than my original estimate of R3,14,000 crore.

What all this means is the following. First, unfortunately, ideology goes a long way in explaining policy dialogue in India. Second, I am thankful to KRM for pushing me via their provocatively wrong estimates to set the record straight. I would suggest to KRM that they influence the respected (albeit ideological!) Economic and Political Weekly to publish further discourse on the subject (I would be happy to contribute) so the subject can be dealt with the finality and the truth that it deserves.

Surjit S Bhalla is chairman of Oxus Investments, an emerging market advisory firm, and a senior advisor to Zyfin, a leading financial information company. He can be followed on Twitter, @surjitbhalla