Inflation in the United States: Bargaining power, labor market institutions, and the Phillips curve

This dissertation tests the hypothesis introduced in Setterfield and Lovejoy (2006), which finds roots in Cornwall (1990, 1994), that the short-run tradeoff between inflation and unemployment can be improved through two routes: rising employment insecurity or a stronger social bargain between firms and their employees. Specifically, the sensitivity of employment insecurity and the social bargain will be tested in two mainstream Phillips curve models---one that econometrically implies a time-invariant NAIRU and one that explicitly measures a time-varying NAIRU. I improve upon the empirical methodology of Setterfield and Lovejoy by correcting for nonstationary data, which has non-trivial implications for their results; by developing better proxies for employment insecurity; and by creating an original and better proxy for the social bargain. To test for the robustness of my results, a wide range of sensitivity tests are performed by incorporating improved measures of labor market tightness, economic activity, inflation expectations, and supply shocks. In addition to the unemployment rate, four measures of the cost of job loss serve as alternative barometers of labor market pressure. Unemployment and output gaps---each calculated relative to a time-varying trend---are also used as measures of economic activity. Inflation expectations are proxied either by a one-year lag of inflation or by forecasts from the Livingston Survey. I also control for three supply shocks---the productivity-wage gap, food and energy prices, and non-petroleum import inflation---that have been identified as integral to the inflationary dynamic. An additional robustness check is performed by using both revised and real-time data. All models are tested for stability and inflation is forecasted for key out-of-sample periods. It is shown that my proxies for employment insecurity (the part-time share of employment) and the social bargain (the share of private employees covered by an employer-provided pension) are highly significant in mitigating inflation in the short run, can affect the value of the NAIRU in the long run, and are insensitive to all of the above specifications.