Performativity, self-interest and pay

A while back I burbled about how 'self-interest' might in part/for some people be a norm, and Charlie commented that this was getting close to performativity. Well in Michael Callon's piece on the performativity of economics in the book I mentioned below he makes the same point (though he says this isn't performativity):

Neoclassical theory is based on the idea that agents are self-interested. If I believe this statement and if this belief is shared by other agents, and I believe that they believe it, then what was simply an assumption turns into a reality. Everyone ends up aligning himself or herself to the model and everyone's expectations are fulfilled by everyone else's behaviours. To predict economic agents' behaviours an economic theory does not have to be true; it simply needs to be believed be believed by everyone.

The other thing I have toyed with is whether performance-related pay works a bit like this. And wouldn't you know it Callon has covered this too in the same piece:

Enforcing incentives inspired by economic theories and their assumptions about human or organisational behaviours causes these behaviours to fit the theory's predictions. When workers are paid on the basis of performance, they end up complying with the anthropological models that fit the incentives imposed on them. If we consider the firm is a nexus of contracts, and we set up procedures to make these contracts explicit and to ensure their enforcement, the firm does become a nexus of contracts.