ally representative population-based study of dementia in the USA to include subjects from all regions of the country can provide essential information for effective planning for the impending healthcare needs of the large and increasing number of individuals at risk for dementia as our population ages.

A structural model of the demand for college attendance is derived as a selection problem in the theory of comparative advantage, in which individuals are endowed with different kinds of talents. Some talents and abilities are more valuable in the types of work associated with college education and others are more valuable for work associated with high school education. For example, mechanical abilities are less important for lawyers than for plumbers, whereas verbal abilities are much more valuable for lawyers. The market tends to sort people into work activities for which they have a comparative advantage, as indexed by expected earnings in each activity. The structural model also allows for the influence of parental background in the selection process.Estimates are based on NBER-Thorndike data and support the theory.There is marked heterogeneity in the population and expected financial gains from college attendance are distributed with substantial variance.Nevertheless, those with greater expected gains have a much larger probability of attending college. The elasticity of demand for college attendance with respect to the permanent college-high school earnings differential is 2.0. Parental background factors also influence demand.The data support the comparative advantage theory: Those who did not attend college would have earned less as college graduates than those who actually chose to attend. More surprisingly, those who attended college would have earned less as high school graduates than did those who actually chose high school. There is no "ability bias" in these data.

Cognitive impairment without dementia is more prevalent in the United States than dementia, and its subtypes vary in prevalence and outcomes.

Until the past decade or so, economists tended to believe that the determinants of fertility are largely noneconomic or, at least, that the analysis of fertility is outside the scope of economic theory.1 In part, these beliefs This is a substantially revised version of a paper of the same title, which, in its earlier form, is the first chapter of my University of Washington Ph.D. dissertation. I have accrued a substantial debt to too many individuals who have contributed to the evolution of the ideas in this paper to acknowledge them all individually. I would like especially to thank John Floyd, who first suggested to me that economic analysis might be applied to population, Jon Rasmussen, who contributed importantly to the mathematical development of the paper, and Warren Sanderson, whose contribution to my thinking on fertility behavior is so ubiquitous that I can only reluctantly absolve him from responsibility for any errors or inadequacies in this paper. My debt to Yoram Ben-Porath will be apparent from his paper in this book. T. W. Schultz, Gary Becker, and H. Gregg Lewis made valuable suggestions for improving the final draft of this paper. I would also like to acknowledge the excellent research and programming assistance of C. Ates Dagli.