Explaining the Recent Behavior of Inflation in the United States

Standard models relating price inflation to measures of slack in the economy suggest that the United States should have experienced an episode of deflation (falling prices) during the Great recession and the subsequent sluggish recovery. But although inflation reached very low levels, prices continued to rise rather than fall. More recently, these standard models suggest inflation should have increased as the unemployment rate declined and labor markets tightened, but inflation has remained well below the Federal Reserve’s policy target. These modern Phillips curve models, which incorporate expectations about future inflation, have in the past performed reasonably well in forecasting inflation. The failure of these models during the last several years presents a troubling finding both for economists’ understanding of inflation and for policymakers’ ability to ensure steady growth and low (but positive) inflation. My paper uses econometric methods to estimate modified versions of the modern Phillips curve model of inflation. The paper assesses three hypotheses as potential explanations for why traditional Phillips curve models have performed poorly in recent years. One hypothesis argues that the flexibility of prices varies with the level and variability of inflation, so a stable, low-inflation environment may cause inflation to be less responsive to slack in the economy. A second hypothesis points to the public’s expectations becoming anchored at a positive level of inflation because of confidence in the Federal Reserve’s ability to keep inflation above zero. A third hypothesis emphasizes increased regional synchronization of the business cycle as a factor reducing the responsiveness of inflation to slack. The paper considers the extent to which each of these hypotheses may be important for explaining the recent behavior of inflation. My preliminary results indicate that core inflation has become less responsive to measures of slack in the economy due primarily to increased synchronization of regional business cycles. This represents a possible reason why traditional Phillips curve forecasts showing core inflation rising to the Fed's target have been incorrect over the past few years even as the labor market tightened. I explore explanations for why the increased synchronization of regional business cycles might cause the slope of the Phillips curve to vary over time, focusing on implications of the sticky-price and sticky-information approaches to price adjustment. These implications suggest that the inflation environment and uncertainty about regional economic conditions should influence the slope of the Phillips curve. I introduce proxies to account for these effects and find that a Phillips curve modified to allow its slope to vary with uncertainty about regional economic conditions can best explain the recent path of inflation.