A further investigation of unemployment persistence in European transition economies

Abstract

This paper analyses the dynamics of the unemployment rate in the eight countries from Central and Eastern Europe which joined the EU in 2004. Unit root tests allowing for nonlinearities and structural changes suggest that the unemployment rate is not stationary in most of the sample countries. Tests allowing for fractional integration, however, reveal that shocks are highly persistent, implying a slow rate of convergence to the natural rate of unemployment. The unemployment rate is least persistent in Hungary and Slovenia, more persistent in the Czech Republic, Slovakia and the Baltic States and extremely persistent in Poland. The degree of persistence appears to reflect the different levels of economic and institutional development in the countries and possibly also the role of the government. J.E.L. Classification: C32, E24

Econometric testing of unemployment hysteresis in selected CEE countries: Lessons for the Serbian economy

“…The estimation of the long-memory parameter supports the conclusion. Our evidence differs from findings previously reported in Cuestas, Gil-Alana & Staehr [7] that covers the same economies, but over the sample ending in 2007. As we analyze the dynamics during and after the Great Recession, it could be argued that the dynamic properties of the unemployment rates in Hungary and Slovenia have changed significantly as a consequence of the 2008-2009 crisis.…”

“…This implies that results will be derived from the explicit modelling of the unemployment rate dynamics. Depending on the value of d, we can assess how a time series reacts to the impact of unexpected random shocks which has relevant implications for the testing of hysteresis in unemployment (see, for example, [7] and [3]). …”

“…In five countries, unemployment was found to be a stationary process with highly persistent structural changes. Identical data set is considered in Cuestas, Gil-Alana & Staehr [7], along with the unemployment rate for the EU15. The level of persistence is estimated to be high, although it differs substantially between countries in the sample.…”

“…Apart from unit root testing, explicit modelling of unemployment rate dynamics was performed in some papers. Linear autoregressive and autoregressive fractionally integrated moving average (ARFIMA) models were used in Cuestas, Gil-Alana & Staehr [7]. The Markovswitching autoregressive model was estimated in LeonLedesma & McAdam [17] and Mladenović & Anić [18].…”

Structural Break, Nonlinearity and the Hysteresis hypothesis: Evidence from new unit root tests.

“…There are a large number of studies considering the structural breaks. The plenty number of these studies succeeded to reject the unit root null (Camarero and Ordonez;Lee and Chang;2008;Furuoka, 2014a), however the majority of the studies were unable to reject the hysteresis effect in the unemployment rates (Chang et al, 2007;Chang;2011;Cuestas et al, 2011;Cheng et al, 2014;Furuoka, 2014b). Moreover, having low power in small samples, particularly when the unemployment series are near-unit root processes, the conventional univariate unit root tests are less reliable and accurate.…”