ABSTRACT
It appears to be widely believed that, in general, public sector agricultural R&D has paid handsome dividends for society as a whole, but even those who hold that view may be sceptical about some of the very high reported estimates of rates of return to research.1 An interest in the outcome might lead to biased estimates in some cases since rate-of-retum estimates are often intended to be used to justify past investments and shore up support for future investments. Both implausibly high and unfavourable results are less likely to be acceptable for this purpose. Rates of return are also likely to involve errors even when the analyst is disinterested because it is inherently difficult to identify which research investment was responsible for a particular productivity improvement (or, conversely, which parts of the productivity benefits are attributable to a particular research investment).
