Decomposing Beveridge Curve Dynamics By Correlated Unobserved Components

Abstract

Between 1979 and 2009, the German labour market moved along a Beveridge curve with changing slope that used to shift outwards but shifted inwards after severe labour market reforms had come into force. We analyse these dynamics and focus on the macroeconomic outcome of the reforms. For that purpose, we construct a new empirical model that links equilibrium unemployment theory to a flexible unobserved components model: we disentangle permanent and transitory components of matching efficiency and separation rate as well as unemployment and vacancies. Cointegration and identification are addressed. We find that matching efficiency and separation rate each account for about half of the inward shift. Thereby, the increase in trend matching efficiency is extraordinary and testifies to a permanent improvement on the labour market. Its visibility, however, was retarded by an overlay with a structural increase in tightness.

On the identification of multivariate correlated unobserved components models

“…In general, our results can provide a useful basis for growing strands of literature that apply and develop correlated UC models, e.g. Morley (2007), Sinclair (2009), Startz & Tsang (2010, Weber (2011), Klinger & Weber (2014).…”

Identifying asymmetric effects of labor market reforms

“…Their main aim was to accelerate labor market flows and reduce unemployment duration. See among others Krause and Uhlig (2012), Launov and Wälde (2016), and Klinger and Weber (2016a) for a quantitative analysis of the labor market effects of these reforms. Dustmann et al (2014) are more skeptical that the Hartz reforms alone explain the beneficial development of the German labor market after 2005.…”

“…In addition to this theoretical anchoring, we take two further steps in order two obtain an eco nomically interpretable measure of reform components. First, while the dynamics of our structural reform components are modeled as permanent, our unobserved components approach allows to control for transitory components potentially arising from business cy cle influences, compare Davis et al (2013), Fujita and Ramey (2009) or Klinger and Weber (2016a). Second, we explicitly filter out potential effects from a changing structural compo sition of the pool of unemployed, e.g., with regard to qualification, age, or the length of the unemployment spell.…”

Patterns of Unemployment Dynamics in Germany

“…Although the transition rates have been adjusted for the German reunification, the striking movement in the early 1990s suggests an investigation of their trend behavior. A closer look on the development of the German Beveridge curve reveals a substantial right shift in 1991 since many workers became unemployed when Eastern Germany was transformed towards a market economy (see Klinger/Weber, 2012). If a significant number of those workers moved to the Western part to enhance their re-employment probability, the registration at Western German employment agencies would indeed trigger a downward shift in the Western German job finding rate.…”

“…In particular, Fernald (2007) demonstrates that VARs with longrun restrictions are sensitive to low-frequencies. Even if low-frequency movements do not 13 Further evidence may come from Klinger/Weber (2012).…”