Unemployment Hysteresis in G7 Countries: Linear and Nonlinear Evidence

The unemployment hysteresis hypothesis suggests that macroeconomic shocks may generate persistent effects on unemployment rates. This study investigates the validity of unemployment hysteresis in G7 countries by accounting for linearity, structural breaks, and nonlinear adjustment dynamics. Monthly, seasonally adjusted unemployment rates covering January 2000–December 2024 are analyzed. Linearity is first examined using the Harvey et al. (2008) test, which identifies nonlinear dynamics in Canada and Italy, while unemployment series in the remaining G7 countries display linear behavior. For linear cases, stationarity is assessed using the Augmented Dickey–Fuller (ADF), Phillips–Perron (PP), and Kwiatkowski–Phillips–Schmidt–Shin (KPSS) tests, together with Perron (1989) and Zivot–Andrews (1992) tests allowing for structural break. Nonlinear series are analyzed using the Kapetanios–Shin–Snell (KSS), Kruse (2011), and Hepsağ (2021) tests. The results indicate persistent unemployment in linear economies, nonlinear mean reversion in Canada, and strong hysteresis in Italy, highlighting heterogeneous unemployment dynamics across G7 countries.