The pattern of diversification within U.S. manufacturing between 1963 and 1977 are examined. Firms didn't diversify at random; they were more likely to enter rapidly growing industries, and industries that were related to their primary activities through supply relationships or marketing similarities. Research and development (R & D) expenditures also influence the observed patterns. R & D intensive industries generate outbound diversification and attract inbound diversification. However, the strongest influence is directional; R & D intensive firms channel their diversification toward R & D intensive industries. Much diversification reflects the transfer of sharable organization capital among related activities.

In this paper we explore the role of transitory income in the housing purchase decision. Because of moral hazard considerations banks typically require downpayments to be financed internally, hence transitory income is potentially important in overcoming the downpayment constraint. Using a novel approach to the measurement of permanent and transitory income, we estimate a two-stage model in which households decide whether to purchase housing in the first stage and the quantity to be purchased in the second. The results indicate a significant role for transitory income in both decision stages.

Environmental regulations set maximum allowable rates for sulfur dioxide emissions from electric utilities. By ignoring firm differences in marginal abatement costs and preventing emissions trading, these standards do not minimize the cost of reducing einissions. This paper estimates marginal abatement cost functions for 56 utilities for 1973-79. Marginal costs vary substantially across firms due to differences in the price of low and high sulfur fuels and the intensity of regulation. The potential savings from a cost minimizing reallocation of abatement resources are estimated for five regions. Current expenditures are found to be 47% higher than cost minimizing levels. I MPLEMENTATION of the 1970 Clean Air Act Amendments relies extensively on technology-based emission standards. Economists long have recognized that these standards can result in an inefficient allocation of pollution control resources.' Engineering-based restrictions ignore or oversimplify differences in abatement costs among polluters. Limited empirical research suggests that current standards result in substantially higher costs, as much as ten times higher, than a cost-minimizing regulatory scheme.2 These studies, however, are based on engineering estimates of marginal abatement costs. The models often assume that all polluters adopt control technologies that are required only for new plants and that have been largely unadopted by polluting firms.3 The impact of changing factor prices on marginal abatement costs typically is ignored. The difficulty is that the resulting estimates of the cost of regulation neither adequately reflect the range of control options available to polluters nor take into account how polluters actually have responded to environmen-

The purpose of this paper is to present new evidence on employer search to fill a position. The study is based on data for recent hires collected in the 1980 Employer Opportunity Pilot Project (EOPP) survey of employers. The paper investigates the effect of factors such as training, employer size, and labor market conditions on employer search. Employer search is measured by the number of applicants interviewed prior to an employment offer and the average number of hours spent by an employer recruiting, screening, and interviewing per applicant interviewed. The paper also documents the relationship between employer search and wages.

This study examines institutional production of higher education. An interesting aspect of this production process is that two of the more important inputs, students and faculty, enter upon considerable self-selection. To address this interdependence, the production relationship is specified by a three-equation simultaneous model in which the quality of college output, faculty, and students are treated endogenously. The significance of simultaneity is demonstrated in an empirical model estimated via three-stage least-squares for a sample of 174 private undergraduate institutions. The results offer clear implications regarding the allocation of institutional resources across the basic factors in educational production. This paper examines the input-output relationship for private undergraduate education. The study falls generally within the economic literature that has analyzed the educational process via a production function specification (Astin, 1968; Bowles, 1970; Summers and Wolfe, 1977; Hanushek, 1979; McGuckin and Winkler, 1979). However, we extend the argument that research assessments of the educational process are not dealing with a production function in the classic sense. For example, the purchaser of the product-the student-is also one of the more important inputs. Further, the non-profit orientation of most universities reduces incentives for cost minimization.' The implication is that the educational process is far more complicated than a simple, production-functional rendering indicates. To demonstrate this point, we estimate a three-equation simultaneous model in which the quality of students, faculty, and college output are treated endogenously.2 This study's broad objective is to identify more clearly the relative contribution of the many human and nonhuman resources combining to produce quality under-

This paper examines the extent to which information obtained from past geographic mobility affects both post-move job-search and earnings in subsequent migration [in the United States]. The study considers this linkage between past and present mobility by estimating earnings frontiers for various categories of interstate migrants partitioned by prior mobility history....[The] results demonstrate that migrant groups exhibiting high relative levels of human capital stock do not necessarily possess superior pre-move labor market information. It is also demonstrated that the incentive to acquire this pre-move information is tied to psychic cost and variation in this cost among migrant types. (EXCERPT)