Does a mother's welfare receipt increase the future dependency of her children? Does the welfare system, thus, stimulate the dependency of future generations? Parameter estimates reported here suggest such intergenerational effects. The sample comprises young girls and their mothers. After control for observed and unobserved heterogeneity, a mother's welfare participation is found to increase her daughter's later welfare dependency.

This paper demonstrates how the circumstances of quitting a previous job affect the probability of a later voluntary job change. The theoretical section describes a model of expected-utility-maximizing job search and mobility. Although the inability to observe utilities associated with specific jobs precludes a direct test of the utility maximizing mobility hypothesis, a testable implication is derived in the context of a simple job search model. Utility maximization implies that workers who rely exclusively on employed search to find a new job are less likely to quit again. This theoretical implication is confirmed by the parameter estimates.

Ahstract-Analyses of voluntary labor mobility suggest that job search facilitates job change while specific training inhibits mobility. Further, given that specific skills cannot be transferred between jobs, and since both search and training are costly, it is reasonable for workers to specialize in search or specific training on a given job. Training or search specialization, however, implies that estimation methods which treat the incidence of a quit as exogenous underestimate mobility effects on wages. The larger wage effects reported here result from simultaneous estimation but also reflect more accurate measurement of wage growth between jobs.

The distinction of quits and permanent layoffs remains a contentious issue in the labor economics literature. Basically, there appear two contrasting views of such turnover. While both interpretations of turnover are derived from job-matching models that assume wealth-maximizing behavior of workers and firms, each involves different assumptions about the feasibility of individual labor contract negotiations following hire. On the one hand, application of the Coase theorem to employment contracts implies that turnover will occur only when the value of the worker's product in some alternative employment exceeds current job productivity. Since the actual division of the job match rent is assumed costlessly renegotiable, only the presence of a truly more productive job assignment would prevent the renegotiation of the current rent-sharing contract to the mutual benefit of both worker and firm. The likelihood of a job change is thus directly related to the level of total match productivity irregardless of whether the worker, in response to some later survey question, characterizes the separation as a quit or permanent layoff. According to this formulation of worker and firm attachments, quits and permanent layoffs are, at least in any sense relevant to actual resource allocation, indistinguishable. An implication of this model is of course that mobility enhances productivity (see Gary Becker, Elisabeth Landes, and Robert Michael, 1977). Counter to this view stand models of job matching that suggest that the presence of transactions costs in the form of costly contract renegotiations preclude any easy application of the Coase theorem to employment contracts. Most forcefully advocated by Masanori Hashimoto (1981), these models of job matching imply that quits and permanent layoffs are different. The difference results from the fact that, with costly contract renegotiation, mobility is to a large extent determined by the actual rent-sharing agreement formulated at the beginning of employment. According to this view then, the total value of the match does not exclusively determine job change, and thus not all mobility is productivity augmenting. This paper offers a model of job matching with costly post-hire negotiations in some aspects similar to that of Hashimoto, but with particular relevance to the population of young workers only beginning their labor force participation. While Hashimoto's discussion concentrates on longer-term firmspecific training decisions, the current model focuses on the period immediately following hire when worker productivity is to a large extent governed by endowed capabilities rather than determined by learning on the job. The theoretical focus on inexperienced workers as well as the corresponding youthful composition of the empirical sample are motivated to control for unobserved factors such as firm-specific human capital stocks as much as possible. In further contrast to Hashimoto's emphasis, the model yields empirical implications concerning the role of wages in the determination of mobility that contrast to the implications of an otherwise similar zero negotiations cost model of job matching. Derivation of the empirical implications concerning the relationship of wages to quits and layoffs represents the primary contribution of this paper. Finally, empirical results are presented that, at least for a sample of very young men over a short period of time, *Department of Economics, University of Houston, Houston, TX 77004. I thank Finis Welch, Dan Mitchell, and the referee for comments on an earlier version of this paper.

The effects of various regulations on hospital costs are estimated using a two decade long panel data set which spans the initiation, and in some instances the repeal, of various forms of hospital regulation. The long panel fosters two improvements over previous research. First, as state hospital cost levels may affect states' incentive to regulate, fixed effect estimators alleviate omitted variable bias derived from the states' regulatory discretion. Second, the long panel permits the estimation of many different regulatory program effects, but also facilitates the analysis of potential regulatory program interaction. The empirical results suggest that previous studies have exaggerated regulatory cost savings: although some interaction effects are indicated, hospital costs appear unresponsive to most regulatory programs.