A Comparative Study on Hysteresis in US and Turkish Labour Markets and the New Keynesian Phillips Curve

This study aims to analyze the structural traits of the US (1990-2011) and Turkish economy (2000-2012) with regards to hysteresis (persistent unemployment) in labor markets. Under the New Keynesian framework two major lines of thought explain unemployment patterns: Non Accelerating Inflation Rate of Unemploymet - NAIRU (or the natural rate of unemployment) and the alternative theory of hysteresis. Following pioneering work of Blanchard and Summers (1986) empirical research has emerged on hysteresis for OECD countries. Following, Jaeger and Parkinson (1991), Roed (1996), Mitchell (1993), Arestis&Mariscal (2000) have conducted studies on unemployment series for various OECD countries. Although with different methodologies, studies diagnose the presence of hysteresis in many countries. Ball in 1997 has revealed significant relation between change in NAIRU, fall in inflation and length of disinflation policies for OECD countries for 1980’s where the only relevant labor market explanatory variable magnifying hysteresis isunemployment benefits. In a similar work dated 1999 Ball detected monetary policy variable as explanatory for NAIRU (and alternatively hysteresis) dependent variable . There are even fewer studies on Turkish labor market hysteresis. While Ener and Arıca (2011) refuted (only refuting work) hysteresis with panel unit root stationarity tests, Gözgör (2012) has detected hysteresis by the same tool of analysis. Bildirici, Ersin, Türkmen &Yalçınkaya (2012), Küçükkale (2002) have also detected hysteresis by statistical analysis and econometric estimation respectively. Nonstationarity of the unemployment series provide us preliminary evidence due to possible presence of hysteresis in both economies. Following, separate wage inflation equations are estimated in the form of New Keynesian Wage Phillips Curve (NKWPC) and New Keynesian Phillips Curve (NKPC) by OLS and GMM methodologies (for Turkey and USA respectively) with employment (and unemployment) vs rate of change of employment (and unemployment) as explanatory variables Alternatively wage and price equations are further estimated against output gap and growth of excess demand variables. Analysis results did reveal hysteresis patterns for the rate of change of excess output estimations for both economies, whereas employment, unemployment variables turned economically and statistically insignificant/unmeaningful which hint that hysteresis patterns are not due to Insider-Outsider mechanism of labor markets but rather macroeconomic environment and policy which fit better into Long Term Unemployment (human capital) hypothesis framework. Low turnover and shooting long term unemployment are also detected during times of recession in US, whereas in Turkey turnover (total separations) is high during busts with high unemployment along with long run unemployment hike. As opposed to US, pattern in Turkey fits more into human capital framework where rising long term unemployment can be reversed with job creation by appropriate policy measures. Unfavorable shocks and economic policy response or labor market structure may trigger hysteresis according to which economic policy choice should be made: demand management policies vs structural reforms in labor markets. Although there is capturing picture around recessions (and/or depressions) in the form of stylized facts, there is stil need for further research to reach solid empirical background to build up the theoretical dynamics.