My own introduction to Coase's answer was memorable, though rather painful. When I was in graduate school at Washington University, Douglass North was on my dissertation committee. At my defence, he asked a question. It seemed like a complicated question, and I went to the board and wrote some equations. Finally (and mercifully), Doug interrupted me. Waving his hand slowly, addressing a not-very-bright child, he said, 'Michael, the answer is two words ... transaction costs!'And North went on to develop that insight in ways that hugely enhanced our understanding of both economic history and the new institutional economics.
And I should have known. For North, it didn't really matter what the question was, the answer, or at least the start of the answer, had to do with transaction costs. He had fully appreciated the Coasian insight that economic (and many political and social) institutions had as their primary function the optimisation of transaction costs.
Of course in some cases institutions develop to minimise transactions costs, firms, for example, occur when they can carryout an activity at a lower cost than the market but in other cases institutions develop to increase transaction costs. Take as an example the secret ballot which makes it impossible to tell if someone has voted in the way he was bribed to vote or not. Such an increase in the transaction costs of voting makes vote buying that much harder.