Paul Krugman writes:

I’ve written before about the CIBC study (pdf — second item) suggesting that high oil prices, by making shipping much more expensive, may reverse a significant amount of globalization.

A new NBER working paper by Koopman et al provides some numbers for a thought I’ve had about this: that the effects of higher transport costs may be especially large because so much of world trade these days involves vertical specialization.

What the authors find is that on average, half the value of Chinese manufactured exports consists of imported inputs; in the higher-tech areas, the number is often more than 80%.

Here’s my thought: in such cases, you want to think of transport costs in terms of a sort of effective rate of protection calculation, similar to the ones often used to show that third-world industries are much more protected than the nominal tariff rates might suggest.

Suppose that China engages in some export activity, but half the value is imported inputs. And suppose for the sake of argument that the price China can get for the final good in the export market is fixed by price of import-competing producers. Let the initial price be 100.

Now add transport costs that raise the price of anything shipped back and forth by 10%. This means that China has to cut its fob price to 91, so as to keep the cif price at 100. It also means that those imported inputs now cost 55, instead of 50.

And the result is that the value added China is able to collect falls from 50 — 100 minus 50 — to 36 — 91 minus 55. That 10 percent rise in transport costs in effect reduces the payoff to China from producing the good by almost 30 percent.

OK, you can question some of this: a lot of the inputs into Chinese production come from other Asian countries, so the transport cost won’t be as big a deal as shipping to the US. But the double squeeze, both on export prices and input costs, is real: CIBC mentions Chinese steel, which is suffering from higher cost of acquiring Australian iron ore as well as higher costs of shipping to the US.

If high oil prices persist, we could be seeing a large drop in world trade.

The mode of transportation most affected by oil prices will be air. Trucking would probably be hit as well. How badly would rail and ocean shipping be affected though? In 2005, Burlington Northern Santa Fe reported that its fuel costs were 19.5% of its total operating expenses. In 2007, the percentage increased to 26.0%. If I am doing the math correctly (and, of course, assuming ceteris paribus), it’s a 9% increase in operating expenses (cost of crude oil to refiners, meanwhile, was $50.24 a barrel in 2005 vs. $67.93 in 2007, a 35% increase). That’s railroads; need to look into shipping…