The First Optimality Theorem claims that in such equilibriums, the traded good or service is allocated among buyers in such a way that it would be impossible through any reallocation to make someone happier without making someone else less happy. In other words, such allocations are Pareto efficient.
The Second Optimality Theorem states that "any particular Pareto optimum can be achieved through competitive markets by simply prescribing an appropriate initial distribution of factor ownership and a price vector". In other words, specific politically desired social/economic outcomes can be achieved in a welfare-maximizing manner using the market mechanism by an appropriate initial distribution of incomes and wealth. In Prof Reinhardt's words,
"If on ethical grounds society wished to distribute a good or service (for example, education or health care or food or beach houses) among people in a particular way — like egalitarian principles — it need not have government directly involved in producing or distributing that good or service. The desired distribution could be attained by redistributing income and wealth among the citizenry in a way that would drive the perfectly competitive private market to achieve the desired allocation of the good or service among the people. Better still, it would do so in the welfare-maximizing way."
Let me illustrate this by taking the classic example of a market for any goods/service which is regulated - either in the form of dual-pricing (food grains through PDS) or price controls (petrol or diesel). The incentive distortions and inefficiencies in such markets have been discussed in earlier posts. Prof Arrow's theorems would have it that the specific social/political objectives (say, food security) can be achieved through the market mechanism and by transferring the proportionate cash subsidy directly into the hands of the targeted beneficiaries.
Such subsidies are ubiquitous in India - public transport, public utility services, farm inputs, weaker section housing etc. Assuming Arrow's theorem to be correct, then the way forward would be to dismantle the price controls and target the equivalent subsidies directly to the bank accounts of the intended beneficiaries.
The benefits would be two-fold. One, the market distortions caused by price controls would immediately disappear and the suppliers (including private ones) could compete to deliver services efficiently. Second, benefits would now be more effectively targeted, thereby avoiding much of the wastage and pilferage that characterize the current arrangement.
However, I foresee two major problems with this approach - one on the implementation side and the other on more theoretical considerations.
1. Targeting presumes perfect knowledge about the identity of the beneficiaries. Maintaining the accuracy of beneficiary identification has traditionally been the 800-pound gorilla in India's welfare administration. The complex nature of India's society and polity, heavily politicized welfare administration, and widely-pervasive and excruciating poverty exacerbates the beneficiary identification challenge.
2. The assumption of competitive markets in these sectors is questionable. Many infrastructure segments are classic monopolies and capital intensive and therefore not easily amenable to efficiency promoting competition. Further, given the under-developed markets, the often weak supply-side may not be able to always match up to the massive and increasing demand.
However, there are answers to such challenges. If the UID/Aadhar gets rolled out according to plan, then targeting suddenly becomes easier. In fact, UID-linked bank accounts have the potential to be a game-changer in the transfer of subsidies directly as cash. Further, the only thing worse than a market with monopolistic structure is one that is not only monopolistic but is also inefficient (due to price controls). To that extent, there is a very strong case to be made that the subsidies should be reimbursed to the targeted beneficiary instead of having a dual-price market. In any case, cushioning supply shocks will always require governments to often step-in with an active role with various policies, including maintenance of buffer stocks and aggressive open market operations on it.
Let me describe the current subsidies-based arrangement as S and the proposed cash-transfer based one as C. The balance sheet for S is that while it is easy to implement, it creates considerable market distortions and massive pilferage/wastage. In contrast, though C is difficult to implement, it avoids many of the market distortions, besides being more cost-effective in terms of the net public expenditure. On the balance, it cannot be denied that C produces far less incentive distortions than S. In simple terms, C is superior to S if its net benefits exceeds that of S.
Let C(d) be the incentive distortions caused by C and S(d) that caused by S.
Let C(w) be the wastage/pilferage generated by C and S(w) that by S.
Le C(i) be the measure of logistics of implementation of C, and S(i) that of implementation of S.
On the balance, the net cost of implementation of S is a measure of S(d) + S(w) + S(i), and that of C will be C(d) + C(w) + C(i). Further, as is clear from the aforementioned reasoning, C(d) < S(d), C(w) < S(w), and C(i) < S(i). Adding them
S(d) + S(w) + S(i) > C(d) + C(w) + C(i)
(S(d)+S(w)+S(i))-(C(d)+C(w)+C(i)) > 0
Therefore, the total cost of S will always be larger than C for any S or C interventions.
The choice of policy alternatives becomes very clear with this formulation. With UID-linked bank accounts, C(i) stops being a major problem, and to that extent the cost differential between S and C widens. As already discussed, the incentive distortions and wastage/pilferage is much less with cash transfers than with subsidized regimes. On the balance, cash transfers score over price controls and dual-pricing.