There's been an interesting debate happening over the weekend on Twitter between Ryan Avent and Noah Smith on the question of what happened to Japan. Smith's view is nothing much all that interesting. Japan closed the per capita income gap with Europe and then, like Europe, stayed poorer than the USA because the US is the richest country around. Avent sees Japan basically closing the gap with Europe but then falling steadily behind over the past 20 years.
Who's right? And how can they disagree so strongly about the basic facts?
It all goes back to the mysteries of Purchasing Power Parities. A PPP adjustment is supposed to take into account the fact that stuff costs different amounts in different countries. Suppose your company needs to send one guy to Buenos Aires and another guy to Oslo. You're going to want to remember that because Norway has higher taxes and much higher wages than Buenos Aires, that paying the same flat $100,000 salary to both people is going to result in wildly different living standards. Now add on to this the fact that international economic comparisons are made more difficult by the fact that exchange rates tend to fluctuate. Common sense says that the relative living standards of Canadians and French people doesn't change that much from one year to the next. But the price of a Canadian dollar in euros can swing pretty wildly from year to year or even day to day as modern financial markets tend to have a lot of instability built into them.
There's a theoretical proposition which says that real exchange rates ought to converge over the long run according to Purchase Power Parity. In other words, the price of the Norwegian krone, the Argentine peso, and the US dollar ought to align themselves over the long term such that $100,000 in Oslo buys you what $100,000 in Buenos Aires buys you.
If that actually happened, then Avent would clearly have the right view of the argument. The correct way to assess Japanese economic performance would be to rely on the PPP-adjusted figures, and they clearly show decline. But PPP convergence doesn't happen in the real world. That's for a variety of reasons, but the most important of them is that in modern economies most of what people do isn't tradable—it's local services. There's no "steakhouse arbitrage" whereby you can exploit the fact that restaurants are cheaper in Argentina than Norway. Given the fact that PPP convergence fails, I'm always uncertain of in which contexts it's best to try to apply it. If PPP convergence held, then market exchange rates would give us a real world test of whether PPP algorithms are getting it right. But since it doesn't hold, we have no test of PPP. There are relatively few contexts in which economists would say we ignore market data and instead favor the output of an opaque international bureaucracy. But most economists side with Avent and prefer the use of PPP-adjusted data. In part that's for convenience—it strips out the instability—but I also think part of the reason is that use of PPP data tends to make the US look better.