No blood for oil

Over at EconLog Bryan Caplan takes on the issue of "blood for oil", he writes;

Stephen Smith makes an argument that seems popular across a wide swath of the political spectrum:

Yeah, and how many billion dollars per year does the United States need to spend even on just the military to make this oil available? How much does it cost just to perpetuate the House of Saud?

The left-wing take on this argument is that it's bad to spend blood for oil; the right-wing take is that it's good (or at least necessary) to spend blood for oil, and we should just face facts. In a recent piece in Public Choice, however, I argue that - whatever else you think about U.S. foreign policy in the Middle East - it's an economically illiterate way to get oil.

And he's right, it is an economically illiterate way to get oil. Think about it this way, its just the theory of the firm writ large. Invasion amounts to a form of vertical integration, a very hostile takeover. That is, one "firm", the US, wants to vertically integrate with another "firm", Iraq, who is a supplier of an input for US production, oil. The question is does integration make sense? In a world of incomplete contracts we know integration makes economic sense when there is a possibility of a hold-up problem due to the relationship specific nature of investments. Given that oil is an more or less homogeneous good and there are a number of different supplies it is not clear what relationship specific investments have to be made and thus its not clear what the danger of hold-up is. Or to put it another way, our two firms should merge if they have highly complementary assets and not merge if they have independent asserts. Given there are a number of other supplies of oil its not obvious that oil from Iraq is highly complementary to US assets. If you think of this as a "make" - invade Iraq - or "buy" - purchase on the open market - decision, its not clear why a "make" decision is optimal.