Ryan Avent writes

Second, we also need to note that rising oil prices represent both demand shocks and supply shocks to the American economy. Dear oil can impact demand directly, by reducing real household income, and indirectly, by influencing consumer confidence. If rising oil prices were purely a problem of demand, then the only thing to fear would indeed be fear itself—by households or by overactive central banks. They are not, however. Soaring oil prices can also dent an economy’s productive capacity. America relies on petroleum as an input to production in lots of different ways—directly, in the case of things like chemicals and plastics, indirectly, in the role oil plays in supply chains and labour markets (as in commuting). When oil prices spike some American production becomes uneconomic. Were the central bank to treat this disruption as a purely demand-oriented phenomenon, it would generate lots of inflation without returning the economy to its previous output peak

Much of the confusion on the blogosphere comes from conflating consumption and output.

Imagine that instead of pumping oil out of the ground in an incredibly capital intensive process primarily dependent on having access to rare geological deposits that we made oil this way: unskilled laborers squeeze dirt with their hands until it forms mush balls which were then turned into oil by mixing them with limestone, one of the most abundant minerals on earth.

Now imagine there is a huge spike in the global demand for oil. Do we think this will raise unemployment or lower it?

Alternatively, imagine that the United States produced three times as much oil as it consumed and that by law an oil dividend check was paid out to all US households from any profits from oil and gas production.

Now there is a huge spike in the global demand for oil. Is this contractionary or expansionary for the the US economy.

The point is that it matters what happens to the flow of funds. The fact that consumers in other countries bid up the price of scarce resources is a reason why consumption of those resources by US consumer might fall but it is not a reason why US unemployment should rise.

Even if it makes a whole host of US production technologies less effective, that is a strong reason why US consumption should fall but it is not itself a reason why unemployment should rise.

This is essentially the same mistake people are making on housing. That the bubble goes away is a reason for consumption to fall, but it is not – absent nominal effects – a reason for unemployment to rise.