Quick note on inflation, exchange rates, and wages

August 18th, 2017 at 11:28 am

A few weeks back, I did some simple modelling of the impact of the weakening dollar on US inflation, suggesting it would boost price growth a bit, but was nothing to get wound up about. And, of course, a little more inflation would be welcomed.

Researchers at Goldman Sachs took an interesting, deeper dive into this question (no link). They came up with the same qualitative answer—a modest bump, not a big increase. In fact, they show that the actual impact on inflation of the dollar’s appreciation, 2014-16, was about half of what would be predicted by conventional models.

What’s changed to dampen the elasticity of the overall price index to import prices?

First, though this just labels the question posed above, the pass-through of dollar/exchange-rate changes to import prices has fallen from 60 to 25 percent. There’s a hint in the figure as to why this occurred. That big drop in the exchange rate pass-through in the latter 1990s coincides with the Asian financial crisis, which led to a sharp downshift in pricing by Asian exporters, notably China.

Source: GS Research

The moral of this story is that one reason for the flattening of the price Phillips Curve is the increase in global supply chains and changes in export pricing. Prices are less responsive to exchange rates. Research has also found the wage-price pass-through to be considerably dampened. Put it all together, and one concludes that the dominant model of inflation is in ill-repair at best.

Whether that’s for now or for eternity is another question, but I’m pretty much with Larry Summers on this: “The Phillips curve is at most barely present in data for the past 25 years.”  Thus, the instincts of some on the FOMC to be patient with rate hikes right now looks theoretically and empirically warranted.

Print Friendly, PDF & Email

One comment in reply to "Quick note on inflation, exchange rates, and wages"

  1. Serene says:

    It isn’t very hard to fix these problems with a little original thinking. I think most economists are tied into a framework larger than they truly understand, and so they have difficulty envisioning novel ways to fix it because they have fixed their ideas to dogma rather than insight.

    Why not use country-specific capital throttles rather than tariffs and quotas? It is so much simpler and it would work better too. The main barrier is to form a set of rules in which to operate, whereby the throttles are tied to an agreed upon set of metrics. Obviously wage imbalances would be one important metric.