Today, I'm thinking about Herbert Simon on labor. In Simon's view, an important and often overlooked characteristic of the employment relationship is the degree of authority granted the employer over the behavior of the employee. In the extreme, a sales contract - the standard economic approach - would have employers paying merely for the product of labor. However, in practice, the employment relationship is one in which employees would likely prefer to do less work while employers would prefer they do more.
Simon used game theoretic notions of bounded rationality to explain the various situations in the labor market that would result in either sale contracts or employment contracts would arise. According to Simon's model, employment contracts will be preferred to wage contracts as information asymmetry about effort expenditure increases and as labor disutility decreases.
I find his approach interesting, mostly for its relative abandonment of supply and demand curve in favor of satisfaction functions. Rather than simply proving the existing of a point of intersection, Simon proposes that a bounded set of wage-behavior pairs could exist that would be acceptable to both the worker and the employer.
Simon proposes an extension to explain the existence and persistence of unions as well. Unlike mainstream economic models, it does not rely on supply and demand, and hence on so-called "market imperfections." Rather, Simon shows that union administration functions as a shift in authority, reducing uncertainty for all parties involved. I wonder what extensions can be made for public assistance or a reserve army.