Recently several writers have examined job search behavior by an unemployed worker.' In the analysis, the period of time between successive wage offers to an individual has commonly been taken as constant. John McCall, for instance, assumed that an individual invariably obtained one job offer per period. Dale Mortensen (1970b) equivalently assumed a constant probability of receiving a wage offer each period. In this context, job search has been characterized as search for an acceptable wage.2 The implication is that the expected length of unemployment depends solely on the individual's acceptance (or reservation) wage, optimally chosen, and the existing wage offer distribution. One aspect of job search heretofore not emphasized is the individual locating a vacancy (wage offer). This element of search is captured below by introducing a random visiting of firms by the job searcher to locate an employment opening.3 This extension permits presentation of a more complete job search theory which, as Charles Holt (1975) suggests, makes allowance for the time it takes to search firms for vacancies as well as the time it takes to search vacancies for suitable wages. The approach allows explicit consideration of an involuntary aspect of search not present in the existing literature. Section I develops a model of job search by an individual that incorporates the above aspects of search. This model then allows in Section II a fruitful discussion of the elements affecting an individual's acceptance wage and the duration of unemployment. It becomes clear that the expected duration of unemployment entails more than a comparison of an acceptance wage and the wage offer distribution. Indeed, the model makes possible computation of a proxy for the probability that the typical individual engaged in job search will accent a wape offer. It is found that * Assistant professor, Purdue University. This research was partially supported by a Baker-Weeks Fellowship provided through the Brookings Institution and a doctoral dissertation grant (No. 91-44-74-39) from the U.S. Department of Labor, Manpower Administration. I am indebted to Herschel Grossman and John Kennan for helpful comments on earlier drafts of this paper. In addition, I benefitted from comments by an anonymous referee. However, the usual disclaimers hold. I For example, see articles by McCall, Mortensen (1970a,b, 1974), Reuben Gronau, Michael Rothschild (1974), and Robert Lucas, Jr. and Edward Prescott. Kenneth Burdett offers an excellent review of several results of such efforts. In addition, Rothschild (1973), focusing on market equilibrium, has provided a survey of . . recent theoretical work characterizing markets whose participants act on the basis of sketchy and incomplete information (p. 1283). 2 The analysis considers expected income maximizing behavior by the unemployed job searcher. Some recent models of job search have replaced expected income maximizing behavior with expected utility maximizing behavior in a nontrivial manner (see John Seater or John Danforth). This alternative approach suggests inclusion of such factors as the nonpecuniary aspects of job offers, leisure, consumption, and risk in the individual's optimal search policy. I This job search method of going directly to an employer is the predominant method of job search according to recent surveys (see Employment and Earnings).

This paper extends the literature on executive compensation by developing and testing a principal-agent model in the context of project selection. The model's focus on executive project selection decisions highlights the multidimensional nature of executive choices that affect the value of the firm. An executive not only makes an effort choice that determines the quality of information on which to base a decision but also sets the decision criteria for selecting projects. A project selection framework is also shown to introduce endogenous uncertainty into compensation that can influence the executive's effort choice. Using an extensive data set, our empirical work supports the main hypotheses of the model, including the significance of executive rank in determining the extent of use of incentive pay in general and equity-based incentive pay in particular.

Recently, several papers examined the quit decision. Donald Parsons considered a quit rate model based on the expected return to employed job search. Parsons relied on existing empirical evidence to justify the simplifying assumption in the model that workers quit only when a preferable job has been located. J. Peter Mattila provided further evidence that the majority of quits have lined up new jobs before quitting and suggested that quits into unemployment be viewed as i.. a fairly small, constant exogenous flow .. . (p. 239). In Section I, we provide a more complete theory of quit behavior within the context of an information and search approach. By identifying the cost of search as the utility value of time spent searching, new choice variables in optimal search strategy, the intensity of search and labor supply during search, are added.' This permits our model to encompass the three options facing an employed individual: employed job search, unemployed job search, or no job search. One result is that, contrary to the hypothesis of Mattila, the second option may be viewed as utility maximizing rather than 6exogenous behavior. To test the theoretical predictions gained in Section I, a new economy wide measure of the quits entering unemployment, synonymous with the number choosing unemployed job search, is computed in Section II. One result, consistent with our model's prediction, is that quits entering unemployment are procyclical with the demand for labor. Section III contains concluding remarks.

The existence of counteroffers can lead to a variety of important labor‐market features. This article develops a model of the selective use of counteroffers in which a firm decides whether to extend counteroffers after a worker informs the firm of an alternative offer. We outline factors that can influence the employer’s net value of making a counteroffer and, thus, affect the likelihood of a counteroffer. We provide a new empirical analysis that examines whether proxies for these factors do, in fact, influence the likelihood that a firm would consider a counteroffer to an employee with a competing offer.