article in the preceding journal issue raises a number of points of detail and one major question of principle. To take the detailed points first, Verry argues that because teaching and research are produced in joint supply, it is possible only to compare the total costs of higher education with its total benefits. If it were possible to measure the economic benefits of university research, or even to define its output in physical terms, this would be a useful exercise, but no satisfactory way of doing this has so far been devised. Admittedly it is difficult to separate the costs of teaching and research, and assumptions have to be made which appear arbitrary, but it is infinitely harder to measure research benefits.
Concerning the treatment of students' foregone earnings, Verry's point is valid. Although the cost figures in columns (2) and (8) exclude foregone earnings, the rates of return have been calculated using cost figures that do include them.
His claim that we equate the return to the government with that to society at large, however, is not correct, and is based on an incomplete quotation of our text. The "five dots" in his quotation conceal the critical word "recovered". What we are saying is that although the benefit to society as a whole is initially concentrated in the pockets of the exgraduates because it is they who receive the extra earnings, 30 percent of this is recovered by the government in the form of additional tax revenue. This involves a return of part of the~benefits of~niversity education from the graduate to the society-at-large which paid for his education. The amount in question shows a 5 per-cent return on the cost borne by the government. The economic criterion for judging additional educational expenditure is, as Verry says, a comparison of the full social rate of return with the social rate of discount, approximated by the rate on government