The optimization process behind most of the theoretical and empirical models previously outlined takes place somewhat myopically in that the employer unilaterally determines employment and utilization levels and, moreover, bases the employment decision on influences that are not contingent on outside events. Some justification for the seeming narrowness of approach in this respect is provided by Hall and Lilien (1979) (see also Hall and Lazear, 1984) who discuss efficient wage bargains under significant uncertainty with respect both to product demand (on the firm-side) and employment opportunity costs (in households). Potential problems concerned with moral hazard, costly legal disputes and the reliability of outside statistical indicators limit the practicability of enforcing wage contracts that are contingent on outside events. The authors argue that, under these conditions, such relatively fixed mechanisms as overtime and shift premium payments schedules can generate efficient layoffs in response to short-term fluctuations and that any cumulation of inefficiencies may be dealt with through periodic contract renegotiation. From casual observation, there seems to be some justification for adopting this stance. None the less, the demand models so far discussed clearly do omit any explanation of a number of important labour market phenomena that are linked with the issues of hours of work and employment. Perhaps the most important of these, especially in the context of the US economy, is that many layoffs are not once-for-all separations between workers and firms but consist of short spells of unemployment followed by re-engagement with the original employer. The quantitative significance of such temporary layoff unemployment in the USA is well documented (see Feldstein, 1973, 1978; Topel and Welch, 1980; Topel, 1982). The usual means of attempting to explain this type of short-term unemployment involves an extension of the conventional type of labour demand model to embrace implicit wage contracts that provide some degree of wage and employment insurance to workers. Specifically, the employment and wage decisions of an optimizing firm that provides such insurance would involve consideration of workers’ utility. As a consequence, the effects of state subsidies to wage income within the total employment/hours of work decision are directly integrated into the model (see, especially, the seminal contributions of Feldstein, 1976, and Baily, 1977). It is this latter aspect of the implicit contract literature that is of most concern to the developments here. (For excellent general critiques of the literature, see Stiglitz, 1984, and Rosen, 1985.)

The main implications stemming from the implicit contract approach to the effects of changes in unemployment subsidies on the firm’s allocation among workers, hours per worker and layoffs form the subject of Section 8.2. Some mention is also made of modifications with respect to unionized firms. A specific implicit contract model is outlined in Section 8.3 which not only illustrates the way in which the earlier myopic models are

generalized but also provides a number of fiscal extensions to the earlier literature. The section emphasizes, however, a number of shortcomings of implicit contract theory in this area of research.