A lot of people have trouble wrapping their heads around the idea that the basic “Econ 101” model – the undifferentiated, single-market supply-and-demand model – doesn’t work for labor markets. To some people involved in debates over labor policy, the theory is almost axiomatic – the labor market must be describable in terms of a “labor supply curve” and a “labor demand curve”. If you tell them it can’t, it just sort of breaks their brain. How could there not be a labor demand curve? How could there not be a relationship between the price of something and how much of it people want to buy?
Funny thing is this is pretty similar to what Sandwichman is saying in Boundless Thirst for Surplus-Labor. The “lump of labor” is a partial equilibrium model and rebuttals to the “fallacy” also invariably rely on partial equilibrium models. They are both wrong.