The numbers game
Written by Prashant // September 28, 2010 // National Politics // 6 Comments
In this post I would like to merely touch on the role of numbers, and by that I mean empirical data, in the Indian policy making process and its discussion in the Indian media.
Most people would agree that the recent debate surrounding the nuclear liability bill was exceptional by normal Indian standards of debate. However an aspect which is often over-looked the lack of any relevant debate on the actual costs and expenses that would have been involved in fixing the liability for operators initially at Rs. 500 crores and later at Rs. 1500 crores. Ideally one would have expected the liability of the operator to be fixed at an optimum rate which balanced the cost per unit of electricity generated by the operator with the insurance premium that would have been required to be paid by the operator. It however appears that the initial liability cap was fixed at Rs. 500 crores because Indian insurance companies had the capacity to insure only upto that amount and no more! There does not seem to be any reasonable nexus between the liability cap and the cost per unit of electricity generated through the nuclear plants. The lack of any such nexus is highly troubling since the Government’s sales pitch suggested that nuclear power would be unaffordable in the case of an unlimited liability. In such a case it is expected that the Central Government would have done its homework to create some kind of economic model linking the quantum of the liability cap to the affordability of the power generated by it. However given any lack of national debate on this aspect it is reasonable to presume that the Government never did have any such economic model which could conclusively establish the liability cap at such a limit which would still ensure the affordability of nuclear power generated by the operator in question.
Other national debates on reservations for backward classes and SC/STs suffer from the very same fallacy. The actual percentage of the OBC population in India has been a subject of intense discussion with numbers varying from 27% to 52%. Once again the political class fails to focus the national debate on the numbers game. The best way for politicial parties to ensure success in both vote-bank politics and implementation of the reservation scheme is to commissioning the collection of such appropriate data either through the national census or alternate means.
The lack of hard numbers is also partially responsible for the slow-pace of judicial reforms. A few years ago noted lawyer Fali Nariman who was a member of the Rajya Sabha had introduced in the house a Judicial Statistics Bill which would create an appropriate bureaucratic machinery to collect judicial statistics from Courts around the country. Frontline carried an excellent article in this regard a few years ago. Such data would help policy makers in pin-pointing the actual reasons for the delays in the Indian legal system. It is possible that simple amendments to problematic provisions in penal laws such as S. 118 of the Negotiable Instruments Act which deals with cheque bouncing cases can considerably ease the burden amongst criminal courts. The same stands true for several other penal and civil laws. However until we have hard data in our hands it is impossible to suggest any concrete reforms in this regard.
All of the above scenarios raise a very pertinent question: If we don’t base our policy on hard numbers, what are we basing it on? Gut Instinct?
6 Comments on "The numbers game"
Hi Prashant. I was thinking just the same thing when reading Sanhita’s article on conditional cash transfers: we need more information to inform our decisions! But just to be difficult, I’m going to take a slightly contrarian position here
You say that we must base our policy on hard numbers. I wish it were so easy. In your example of SC/ST reservations, what would hard numbers tell you? Presumably, you mean a cost-benefit type study of the equity gains and efficiency costs from introducing quotas into, say, higher ed institutions (?). The results of such an analysis (aside from immense technical problems and huge assumptions involved in conducting it) would still depend on how politicians and policy-makers define the equity gains they want to achieve. Or what they intend to benchmark these “efficiency losses” against – entirely a political/policy question.
Nuclear liability is an interesting one. The model used to determine insurance premia and liability caps ideally shouldn’t have anything to do with the cost of power generation, but should be driven by the probability of a disaster. But what is the probability of a nuclear spill? Our actuarial hard numbers consist of Chernobyl and little else (thankfully). If, as a policy maker, you really, really want to open up the nuclear power sector, then you can only set the liability hurdle as high as possible without discouraging all private investment. It requires taking a leap of faith.
Of course, I can’t disagree with the substance of what you’re saying: more information is always better than less. But, in my opinion, while numbers can inform government, they can’t substitute for clarity in political and policy objectives.
Hi Anisha,
Thank you for that insightful comment.
You give me too much credit in your comment! I was aiming for something a lot more simplistic.
My point in regards the SC/STs and OBCs reservations was only in regards the percentage of reservation that they should be given in educational institutes and jobs. If caste is not a component of the census how do we figure out the actual percentage of SCs in the population? The corollary being, if we don’t have accurate percentages, how do we end up fixing arbitrary 27% reservation for these groups? The OBC calculation is a lot more complicated because of different criteria and this maybe the reason for the huge difference in the different surveys conducted by different organizations. Given the population explosion in the last 3 or 4 decades it makes little sense to fix quotas on the basis of antiquated data. From what I understand the Mandal Commission based its findings on pre-independence data. Similarly the SC/ST quotas have remained unchanged from the time of independence. If a politician wants to ensure equity he needs to have the right data to determine what is equitable. In India however we keep ending up on unreliable outdated data.
In regards the nuclear liability case, as uncomfortable as I am in arguing economics with an economist, I’ll still throw in my two cents worth. I think the very purpose of capping liability is an admission that private industry cannot meet the full costs or even half the costs of an accident. The actual cost of the accident is therefore only tangentially relevant. In such a case I presume you should start work the other way round and see how much the industry can afford to pay while remaining competitive instead of calculating how much the insurance companies can themselves afford to insure, which was the case with the Indian bill. The entire point of a cap is to ensure the profitability of an industry, right?
Regards,
Prashant
You’re right about the fact that the data cited by the courts, media, politicians and so on is incredibly old and most exasperating. But I honestly think that even if it was up to date it wouldn’t make any difference to the people who want reservations. Take this friend of mine. I tried to argue with him the same point that you raised. His response: current demands for reservations have little to do with the proportion of SC/ST people. The end objective is certainly representation in proportion to their population. But, he argues, it may be necessary to have more than proportional reservations in the present in order for the centuries-long discrimination to be erased. Bottom-line: extend reservations to as far as is politically possible in the near term and damn the numbers.
On the nuclear liability issue, I think you and I agree: the cap will have to be based on the business model of the interested MNCs. My guess as to why the government didn’t examine this model is that the MNCs did it themselves and declared they weren’t going to pitch up with a higher cap. I wonder if this liability bill has called the MNC bluff (assuming it is a bluff)? According to the media, the issue is still kicking around in Indo-US talks…
this post is very usefull thx!
Hi Prashant,
The problem, I believe, in policy making in India, on the contrary, is an over-reliance on numbers. Numbers can be quite tyrannical at times and I think it may serve us well to bring politics back into policy, especially on issues such as poverty analysis. For instance you have a multi-dimensional poverty index that has been introduced to reflect the deprivation of the poor from a more holistic perspective. Yet, it continues to remain as far-fetched in its indication of the deprivation as its preceding models.
I suppose the key is to arrive at some sort of balance between ideology and hard facts, but I think at times there is excessive importance placed on numbers, which has a tendency to shift the focus from what is actually socially relevant.
Hi Suhrith,
That’s an interesting example. Could you give us any examples in India of an over-reliance on numbers and data?
Cheers,
Prashant