ImplicationsForCAPTheory

The reason I'm particularly interested in this model is that it sets the limit case of what rational individualist agents can achieve without interacting. (At least in terms of providing discrete public goods, though I think the results can be generalised further.)

This isn't - nothing under any circumstances.

It is possible to get productive equilibria in the model, though in restricted circumstances. (See SimulationResults.)

Key factors are:

General Cost Level - If people want it enough, it will happen, because even a very small probability of being the swinger is worrying enough to make you contribute if you want it badly enough.

Group Size - The smaller the group, the higher the probability of being the swinger. So although the ParadoxOfVoting may not be resolved by this model for large scale elections, it may be enough to explain some small group situations.

Uncertainty - 'Variability' of the distribution of costs across the group can effect the outcome in various ways. One possibility could relate to agents' 'certainty' about their guesstimates of the contribution rate. It is one thing to feel very sure that the good will be provided without me with 80% contributing. It is another if I only have a vague hunch there could be 80% contribution. An interesting thought - in revolutionary periods when 'all that is solid melts into air', anything becomes possible, and it may be that your action can make a difference.