Incoherent Thinking About an Iranian Oil Embargo

In the not so distant future – indeed perhaps only months from now – the United States and Europe may enact a mix of sanctions against the Central Bank of Iran (CBI) and Iranian oil exports as part of an effort to stall the Iranian nuclear program. Sanctions against the CBI would leave many current Iranian oil customers without any way to pay for their crude, effectively triggering a partial boycott on Iranian oil exports. Explicit sanctions would, of course, do the same.

This prospect has elicited a flood of market impact analyses. The ones I’ve seen are proprietary, so I won’t post or quote them here, but they appear to have one important thing in common. Parts of each analysis look country-by-coutnry to see who is currently most dependent on Iranian oil. They then suggest that those countries are at greatest risk in the event a cutoff. Other parts of each analysis estimate the rise in world prices that sanctions would provoke as a way of explaining how bad the consequences would be.

This is, quite literally, incoherent. To argue that each country will be affected differently depending on its level of imports from Iran, one must conceive of the world oil market as a place where geography matters deeply. This is inconsistent with thinking about a single world price for oil.

It’s easiest to understand this with an example. Spain imports roughly fifteen percent of its oil from Iran, while France imports only three percent. The first type of analysis I’ve cited would lead one to conclude that Spain would suffer more than France from an oil embargo on Iran. Since both countries import essentially all their oil, that would have to mean that they would experience substantially different oil prices. Think of it this way: if Spain needs to cut consumption by fifteen percent in order to balance its internal market, but France needs to cut its consumption by only three percent, Spain will need to experience a much larger rise in prices. This is, of course completely inconsistent with any calculation that treats the world as an integrated market with a single price.

If, on the other hand, Spain imports oil from France to balance its market (or buys oil that France would have purchased otherwise), leading prices to converge, then the fact that Spain was initially more dependent on Iran than France was will be irrelevant.

One can take this a step further. There is actually some reason to believe, at least in the short term, that a disruption in Iran could affect different consumers in different ways. (It can take time to rearrange shipping logistics and the like.) But for every country that is hurt more than a simple global calculation would suggest (because it can’t simply replace its supplies at the prevailing world price), there is another that doesn’t feel the full brunt of the disruption.

To see how this works, take another caricatured example. Let’s imagine that there are two suppliers in the world, A and B, each of which export one million barrels of oil each day, and two consumers, C and D, each of which import a million barrels a day. C buys its oil from A; D buys its oil from B. Now the world cuts A off entirely. There are two extremes at which one can analyze the outcome. In the first, markets are ultra-rigid, so C is toast. But in that case, D is sitting pretty, since it keeps importing the same amount of oil as before from B, and the market balances at the same price as before. In the second, markets are fully flexible. Both C and D reduce their consumption to half a million barrels a day each, splitting the total output from B. The world experiences a uniform price rise in order to trigger these cuts, and both countries suffer equally. The fact that C was importing from A before the cutoff is irrelevant.

Of course, one can imagine cases that lie in between these two extremes. Here, there is not one global price; C isn’t as badly damaged as in the first case, but it’s worse off than a simple global price estimate would suggest. D, meanwhile, is worse off than in the first case, but it’s better off than simple global price estimates would indicate.

What matters here is that simultaneously applying the two modes of analysis – one that makes geography central, and one that assumes it away – leads to incoherent results. If we want to really understand the likely impacts of an embargo on Iran, we need to square these two ways of thinking about the world, rather than pretending that they can coexist.