Much economic analysis is based on stylized relations between the nature of economic processes and the functions economists use as their tools of analysis. For example, if production processes are duplicable, constant returns to scale production functions are obtained. The student of elementary economics is told that if goods are difficult to substitute, their demand is inelastic; if easy to substitute their demand is elastic. In similar fashion it is the purpose of this paper to categorize the nature of jobs and to show that it has implications for the elasticity of demand for labour with respect to the wage. We see a job as like a dam site. A dam which underutilizes a dam site, even though productive in the sense that water is usefully stored or electricity is usefully produced, will nevertheless be costly in the sense that the valuable dam site is wasted. Even at zero cost (and hence a benefit/cost ratio of infinity) it may not pay to use a dam which underutilizes the site. We picture jobs and workers in the same way. Jobs are pictured as being like dam sites and workers of different skills as being like the potential dams on the dam site. Workers of sufficiently low skills will not be able to get jobs even at zero wages, not because their output on those jobs is negative, but because they underutilize the jobs themselves. Nor may it matter that the firm will have to pay significantly positive wages to hire skilled workers on its jobs. This image of the job gives reason for pessimism about the wage elasticity of demand for labour of a given skill, since it says that unskilled workers, no matter how low they bid their wages, may still be unable to bid jobs away from skilled workers. This has at least five consequences. First, it shows that in a demand downturn, in which prices of final goods and services are below the full employment level, wage flexibility alone will not be sufficient to restore full employment. Skilled workers will receive jobs and positive wages; unskilled workers will not be able to capture jobs even at zero wages. Second, the minimum wage is often considered a major cause of unemployment in general and among youth in particular (see Feldstein (1973) for one example). The effects of minimum wages on employment depends critically on the wage elasticity of demand, which is low for unskilled workers if, as in our argument, they cannot capture jobs even at zero wages. Third, Feldstein among others (1973, pp. 19-26) has argued that wage subsidies should be paid to encourage employees to hire youth in ladder jobs. Feldstein's recommendation rests implicitly on the belief that the elasticity of demand for such youths in such jobs is fairly high-at least in the sense that employers would more than willingly hire such workers with sufficient reduction of the minimum wage. Wage costs thus serve as an upper bound to his estimates of the costs of such programmes. If, on the other hand, jobs are as pictured in this paper, like dam sites, even at a zero wage firms may be quite unwilling to hire unskilled workers in ladder jobs because they may underutilize the ladder jobs themselves. The other side to the subsidy issue is manpower training programmes. Such programmes have been strongly criticized in the US for their high cost per worker. Implicit in much of this criticism has been the comparison with on-the-job training on the supposition