ON THE CONCEPT OF EFFICIENCY Prabhat Patnaik Concepts which have to be handled with great care are bandied about irresponsibly today It would not be inappropriate therefore to dwell on this issue and to illustrate it with reference to the concept of efficiency which is widely misused I A particularly common example of misuse is the statement the private sector is more efficient than the public sector If the evidence offered in support relates to overall profitability of the two sectors then the baselessness of the claim is obvious If two sectors producing non identical commodity vectors and facing arbitrary prices as opposed to shadow prices derived from an optimisation exercise have different rates of profit then this fact says absolutely nothing about their respective efficiencies no matter how the latter is defined unless it is defined to be identical with such empirical profitability in which case the proposition is trivial Even when we compare the two sectors producing apparently the same commodity say steel there are important differences in product mix because of which profitability comparisons are no index of relative efficiency In fact such a comparison between the public and the private sectors is exceedingly complex A possible way out is to take an engineering notion of efficiency examining the use of some key inputs per unit of comparable output e g consumption of power per unit of pig iron production On such a comparison however the public sector does not come out badly at all Bagchi et al 1985 I shall be concerned here however not with such sectoral comparisons but with the macro level where the concept of effciency commonly used is equally flawed One often comes across the view that the Japanese economy is more efficient that the Indian economy The proof of this statement is supposed to lie in Japan s superior export performance As an explanation of the relative export performances of the two countries this is a tautology reminiscent of the following anecdote When Paul Samuelson was asked why doctors earned so much more than sweepers his reply was that the former s contribution to society was greater than the latter s and when asked on what basis he could assert this his reply was look at their respective earnings If relative efficiencies assessed in terms of relative export performances are used as an explanation of the latter then we have an explanation of the same genre On the other hand relative export performances as an indicator of relative efficiencies have nothing to recommend them Like relative profitabilities between the public and the private sectors in the example above they have no rationale as an index of relative efficiencies And it would force us into absurd conclusions e g that colonial India prior to the First World War was a shining example of efficiency since it witnessed rapid export growth that it was more efficient than inter war India when the Depression arrested export growth and so on Fortunately however a direct meaning is given to the concept of efficiency in the economic literature namely if with the resource endowments available in an economy in any given situation in the sense of being offered by their owners the output obtainable is not in a vector wise sense larger than what is produced then the situation is one of efficient production Putting it differently if the resource endowments of an economy are given then a production possibility frontier can be drawn with respect to those endowments showing the alternative bundles of maximal amounts of commodities that can be produced with those endowments Any point on this frontier is a point of efficient production It by no means follows that any point on this frontier is better or socially preferable to a point inside the frontier i e that a movement from a situation of inefficient production to a situation of efficient production is necessarily a good thing Welfare economics spent decades looking for arguments that could establish that a point of efficient production no matter where it is located on the frontier is preferable to a point of inefficient production But a comparison between any two points necessarily requires a judgment about income distribution It follows then that not only can we not compare two points one inside the frontier and one on the frontier but producing vector wise noncomparable commodity bundles without making some judgment about income distribution but we cannot even compare two points one involving a larger output vector wise than another without such a judgment In other words even a move from one point to another where the latter produces more of all commodities cannot be defended on welfare grounds unless there is a judgment that it has not worsened income distribution Thus even according to conventional Welfare Economics an improvement in the efficiency of production is not per se desirable So all the talk one hears these days about the need for improving the efficiency of production which says nothing at all about income distribution lacks any theoretical rationale even according to conventional economics But I shall not be concerned about this aspect either My concern is with the concept of efficiency itself not with whether a move towards efficiency as conventionally defined is a good thing or a bad thing II The proposition that a vector wise increase in output from a given endowment of inputs constitutes an improvement in productive efficiency is an unexceptionable proposition But it is invariably taken to be synonymous with another quite distinct proposition which can be expressed as follows Consider any actual position of production It is characterised by an activity set that transforms a bundle of inputs into a bundle of outputs One can think of marginal rates of transformation substitution of inputs into outputs of outputs into outputs and of inputs into inputs associated with this activity set Now if there exists some activity outside of this set whose inclusion at the expense of some activity inside the set improves these rates of transformation substitution in the sense that one more unit of any input either gives more extra output in the alternative set than currently or saves more of some other input than currently or that one more unit of any output necessitates a lower fall in some other output than currently then the use of this alternative set instead of the current one entails an improvement in efficiency This second proposition can be
View Full Document
Unlocking...