there are only two things you can do with the world's oil production: consume it, or store it.
If the price is above the level at which the demand from end-users is equal to production, there's an excess supply -- and that supply has to be going into inventories. End of story.
First, let me reiterate that I do not believe that the oil price today reflects a bubble. So in that respect, Krugman and I are not on opposite sides. Nonetheless, I do think that his model of the oil market has some strange properties.
Krugman ignores two elements of the oil market. Explictly, he ignores forward prices. Implicitly, he ignores the decision by producers either to pump oil or keep it in the ground. As Tyler Cowen put it,
Isn't it easy enough to argue that the relevant hoarding is of oil in the ground rather than oil in strategic reserves or panic stockpiles?
Tyler goes on to cite a number of reasons why oil suppliers might not be rational. I can come up with others, such as suppliers doing a naive comparison of cost of production compared with the current price of oil to determine production levels, rather than taking into account the opportunity cost of leaving more oil in the ground and selling it next year. If producers ignore opportunity cost in that sense, then the market really does not depend on future conditions, and forward prices may be irrelevant. But, again, in that case the market is not rational.
Perhaps Krugman believes that oil producers are irrational in some specific way. If so, then I do not think that his simple exposition is sufficient to articulate his model.
If Krugman believes that producers are rational, then the question I have for his model is this: what does it predict would happen if everyone's estimate for oil demand in the year 2010 were to jump suddenly?
In my model, the estimate of high oil demand in 2010 would lead to a rise in futures prices. At existing spot prices, this would make it rational for producers to leave oil in the ground, rather than sell it today. This in turn would cause the spot price today to rise. Then, at today's (slightly) lower production level and (much) higher spot price, the market would operate as Krugman postulates--oil would be either consumed or stored in inventory.
This model does not rule out the possibility that the futures market reflects a speculative bubble. It could be that estimates of future demand for oil are exaggerated. As it happens, I do not believe that estimates of future market conditions are erroneous. However, one cannot rule it out simply because we do not observe bulging inventories at oil refineries.
If there is a bubble, then at some point market expectations for future oil demand will decrease. That would cause futures prices to fall. This would lead producers to want to pump more oil now, before prices drop. That in turn would cause the spot price to fall. In a bubble scenario, the price drop would be dramatic. Once again, let me say that my guess is that this will not happen. I would assume that the market has it about right.