I think most of the desirable aspects of government can be framed in terms of such coordination problems; i.e. as a mechanism for groups to enforce collective d

I think most of the desirable aspects of government can be framed in terms of such coordination problems; i.e. as a mechanism for groups to enforce collective decisions on members, where deviations benefit that individual but make everyone worse off on average. For example, I can benefit from polluting (e.g. saving money by using cheap, poorly refined fuels). My pollution costs other people, but it's only small, since I'm just one person. Hence it's rational for me to pollute. However, I don't want other people polluting, since that will cost me. Since there are lots of other people, it may even add up to more than I would gain by being the only polluter (saving money versus, say, getting cancer). From the perspective of individuals, this is a coordination problem. The categorical imperative (what we would like everyone to do) is to not pollute; but this is unstable, since there is incentive to deviate (e.g. save money). In a competitive environment, polluters have an advantage, so the polluting strategy will dominate. Collective government is a mechanism to solve such problems. The group decides (e.g. by vote) that freedom from pollution is worth more than freedom to pollute, and agrees to punish polluters, disincentivising that activity, and thus changing the payoff calculation. Of course, collective government itself introduces all sorts of new coordination problems, like corruption, revolving-doors, watching-the-watchmen, a disconnected political class, etc. (many of which are particularly evident right now!)