Has Globalization Changed Inflation?
Globalization has had little effect on the rate of inflation in the United States.
Many observers have suggested that the behavior of U.S. inflation has been changed by the "globalization" of the economy. In 2005, for example, The Economist declared that recent experience "makes a mockery of traditional economic models of inflation, which ignore globalization." According to such commentators, globalization has helped to reduce inflation in the recent past and will help it to remain low in the future.
In Has Globalization Changed Inflation? (NBER Working Paper No. 12687), NBER Research Associate Laurence Ball questions this view. He reviews theory and evidence on the behavior of U.S. inflation, and concludes that globalization has had little effect on the rate of inflation in the United States.
Ball first questions the extent of globalization. Commentators suggest that inflation has been influenced by increasing trade between the United States and other countries. While trade has increased, however, this has occurred slowly over many decades. The last quarter century, when U.S. inflation has been tamed, is not noteworthy for particularly rapid increases in trade.
Ball then turns to arguments about why increased trade might influence inflation, and finds them flawed. One argument, suggested by Kenneth Rogoff of Harvard, is based on the idea that globalization has changed the Phillips curve, the short-run tradeoff between output and inflation. In this story, the tradeoff has become less favorable, with a given increase in output causing a larger rise in inflation. In theoretical models of inflation, such a change reduces the incentive for the Federal Reserve to pursue expansionary policies, leading to lower inflation.
This argument is questionable on theoretical grounds, but the biggest problem is empirical. The literature on the Phillips curve suggests that this relation has changed, but in the wrong direction. A given change in output has a smaller effect on inflation today than it did in the 1970s or early 1980s. Therefore, if the slope of the Phillips curve (the output-inflation tradeoff) were a key determinant of inflation, we should have seen rising inflation in recent decades.
Ball then examines another claim about the effects of globalization, which also relates to the output-inflation tradeoff. This claim is that globalization has weakened the link between U.S. inflation and the level of output in the U.S. economy, with economic booms causing less upward pressure on inflation. According to this view, what matters for inflation is output in the entire global economy.
Once again, Ball raises questions about the alleged effects of globalization. Empirically, there continues to be a close association between the level of U.S. output and changes in U.S. inflation, with at most a secondary role for output in other countries. A well-known study from the Bank for International Settlements has reported large effects of foreign output, but the statistical claims in that study do not withstand careful scrutiny.
Finally, Ball examines the role of falling prices for imported goods. Many policymakers and journalists cite increases in imports of low-cost goods from countries such as China and India. To many observers, it seems obvious that lower prices for imports contribute to lower inflation, since inflation is an average of the economy's price changes.
However, this idea rests on a confusion between relative prices and the aggregate price level. As Milton Friedman pointed out long ago, changes in the price level - that is, the inflation rate - depend on monetary factors. Trade with China and India reduces the relative prices of certain goods, which increases U.S. living standards, but there is no obvious effect on inflation.
There appear to be some historical episodes, such as the oil-price increases of the 1970s, when changes in relative prices did affect inflation. However, these involved large, sudden shocks to the economy. The steady rise in foreign trade has caused downward trends in some relative prices, but such smooth changes are unlikely to affect inflation significantly.