producers' and consumers' durable goods, yielding a stream of benefits which are compared with a stream of costs. Births occur if the present value of the benefits exceeds the present value of the costs. The theory has been used by Malthusians to rationalize the casual observation that population growth rates increase during the early stages of economic development and then, perhaps, fall later on as

Dear Old Friend: As you have doubtless observed, it has been many a week now since I last stepped out of the shadows, put the barrel to your ribs and divested you of cash, trousers, and credit cards. You have probably asked yourself, Where is my faithful old stickup man, John? Doesn't he like me any more?-The truth is, sir, that I'm terrified you'll take it amiss and move out of New York if I overwork my welcome. John [Quoted in Russell Baker's column, New York Times, December 9, 1975]

Warren Robinson dismisses the strength of the on the grounds of Gorin Ohlin's data which suggest that children are a costly way of securing old-age support. Quite apart from Robinson's confusion of marginal and average costs and returns, suppose the rate of return on children were so low as to be even negative. Should we then conclude that the pension effect has no force? Surely it would be naive to do so. Rates of return cannot be appraised in vacuuo. For example, suppose nature offered people negative rates of return on children and everything else, and that intergeneration transfers are ruled out. Then parents, planning for their retirement, must invest at that negative rate, which will then be the observed rate, quite apart from their time preference. Only if children turn out to be bad investments, relative to other assets, can we conclude that the pension effect is weak. Even then, we might observe relatively low rates of return on children as assets, if they are thought to be relatively safe investments in diversified portfolios. But if Robinson's conclusions still hold, if children have no place in a rationally selected retirement portfolio, then I am grateful to him for having highlighted the argument in Section IV of my article when I suggested that a dominating . . good asset (bonds) drives out the bad asset (children) (p. 387) so that population will wither away unless there are other motives for having children (p. 388). Robinson goes on to question my assumption that human motivation has something to do with human fertility. Nothing he says convinces me that the desire to reduce family size is not an essential ingredient in the problem. The costs of fertility control can be viewed as costs associated with achieving a desired portfolio. If condoms, intrauterine loops, or other devices reduce these costs then I am in favor of them as much as is Robinson. Finally I do not think Robinson and I really disagree about the nature of the fundamental market failure in my model economy. It was precisely intergeneration effects, within the family, I had in mind. But surely we do future generations a disservice if we ignore time-wise generation effects. If the postwar baby boom turns out to be a bulge on the population pyramid, then Robinson is shortsighted in discounting the relative magnitude of the intergeneration transfers when that bulge of babies becomes a bulge of dependent grandparents. Perhaps Robinson belongs to an unconcerned generation. After all the crunch will not come until 2010 A.D. or so. * University of British Columbia.

A three-sector general equilibrium model is employed to analyze the consequences of changes in aggregate demand, union power, workers' time horizons, and labor force size on "unionized" industries, industries with competitive labor markets,and unemployment. Workers are "voluntarily" unemployed while searching for jobs,attempting to arbitrage a wage gap created by partial "unionization." Many reduced-form elasticities are ambiguous in algebraic sign. For example, decreased aggregate demand may increase employment in industries with competitive labor markets. However, a larger and younger work force with enhanced aspirations for higher-paying jobs will have unambiguously higher unemployment.