Bureaucratic productivity: The case of agricultural research revisited

Conclusions and implications

Two questions related to the financing of agricultural research merit further study. First, to the extent that selective access to agricultural research is possible, there is no presumption that research results should be distributed in a way which gives everyone equal access. Moreover, the returns for a large part of agricultural research can be appropriated by the developer through patents, copyrights, and other means. Thus, there is no presumption that publicly financed research services should be equally available to everyone since, ‘in general, equal access to government services is neither necessary nor efficient’ (Goldin, 1977: 54). More attention should be devoted to the method of financing agricultural research — i.e., by taxation versus market prices.

A second important question concerns how agricultural research services should be produced. In the case of a public good, it is necessary to provide the good collectively since, by definition, there is no way to make the service available selectively and, hence, private production is not feasible. There is evidence that a large part of agricultural research services are not public goods and, consequently, can be provided selectively by private producers. The advantage of private production is that goods and services are then subject to the ‘incorruptible judgment of that unbribable tribunal, the account of profit and loss’ (Mises, 1969: 35). Thus, the question of who should perform agricultural research — the private or the public sector — warrants more study.

If, as appears to be the case, agricultural research is largely a private good, one would expect entry of new research firms until the rate of return is comparable with returns from other investments of comparable risk. Consequently, the apparent high rates of return from agricultural research should be viewed as suspect if (as seems likely) there are no significant entry barriers. The methods used in estimating ex post rates of return from agricultural research can be questioned on a number of grounds. Regardless of the accuracy of ex post rate of return estimates, however, it is not appropriate to assume that decision makers should base ex ante expectations on ex post rates of return. Moreover, the outside observer has no way to measure the ex ante cost and returns which motivate decision makers as they weigh the opportunity cost of additional research funds in agriculture in terms of the sacrificed alternatives associated with potential increases for agricultural price supports, prisons, defense, roads, welfare, etc. Furthermore, the discount rate is likely quite high for public expenditures due to the short time horizons of political decision makers. In view of these considerations, the conclusion that there is ‘substantial underinvestment’ of publicly funded agricultural research remains in doubt.