A new Finance and Economics Discussion Series paper from the Federal Reserve Board, "United States of U-Star" by Hie Joo Ahn and Jeremy Rudd, estimates state-level trend unemployment rates, or trend U-star. The authors derive these estimates from the trend components of unemployment inflows and outflows across unemployment durations. The paper finds that the estimated trend U-stars, particularly the portions attributable to long-term unemployment, have levels and dynamics that vary substantially across states.
The research shows long-term unemployment makes a relatively small contribution to trend U-star in agricultural states, but takes a larger share for Rust Belt states and several states in the Sun Belt. The authors also report that higher educational attainment and population aging put downward pressure on U-star, while the effects of industrial structure and labor-market rigidities are mixed. The paper uses a nonlinear state space model, an extended Kalman filter, and a Nelson-Siegel model as part of its methodology.
Using the state-level trend U-star estimates, the authors revisit previous work on state-level Phillips curves. They find that the estimated slopes of forward-looking inflation equations are driven by movements in trend U-star rather than by cyclical variation in the unemployment rate. This suggests that state-level inflation-unemployment relationships may reflect structural labor-market differences rather than short-run cyclical dynamics.
The paper is preliminary and circulated for discussion and critical comment. The authors note that the views expressed do not indicate concurrence by other Board staff or by the Board of Governors. What remains unknown is how robust these state-level trend U-star estimates are to alternative model specifications, and whether the finding on Phillips curve slopes holds across different time periods or inflation measures.