Globalization Didn't Bring Down US Prices After All

Can cheap Chinese imports bring down the US inflation rate? It seems intuitively obvious that they can. After all, if widgets come down in price, and they make up a certain part of the price index, then the price index must come down too.

Larry Ball, however, in a recent paper, demolishes the concept, and the high-profile economists such as Ken Rogoff who seem to believe in it.

The problem is that the prices affected by trade are relative prices. Imports of Chinese shirts make shirts cheaper compared to other goods and services. Inflation is the aggregate change in nominal prices. There is no “natural” or “obvious” connection between inflation and relative prices, as any pattern of relative-price changes is consistent with any inflation rate.
One way to appreciate this point is to remember that, for every relative-price decline, there is by definition a relative-price increase. Instead of focusing on declines in the relative prices of imports, we could note the rising relative prices of domestically-produced goods. Should we worry that these price changes put upward pressure on inflation?

Ball comes with heavyweight backing from Brad DeLong, so I'm half inclined to believe him. But my economics isn't good enough for me to have complete confidence in the argument's being watertight. Why are the effects of trade always relative and not nominal?

I do understand that if inflation is "always and everywhere a monetary phenomenon," then, as Ball puts it, "the accounting theory of inflation is always and everywhere a fallacy." But at the same time I can't help but think that the addition of hundreds of millions of low-cost Chinese workers to the global labor force has had profound effects on the global economy in general – effects which must have spilled over somehow into the US inflation rate.

Still, this is what economics is good at: generating counterintuitive results. Unless and until I can poke a hole in Ball's argument, I guess I'll have to believe it. But if Ball is right, then, as "Dave" points out in DeLong's comments, there are some interesting implications:

Is it obvious that the 1970s inflation was really due to oil shocks, or was that a convenient scapegoat to avoid having to admit the need to run the presses overtime once the bills for vietnam started coming due?

After all, if cheaper widgets can't reduce inflation, then in what way can more expensive oil increase it?