ABSTRACT
This article examines the relationship between the unemployment rate and the labour force participation rate in Spain. Cointegration analysis is performed for aggregate, male and female time-series. Results suggest that there is no a long-run relationship between the two variables for the aggregate and male cases. However, the findings support a long-run relationship between the two variables for the female time-series. Thus, the unemployment invariance hypothesis is supported in the two former cases but not in the latter.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 Although limited time-series, such as the UR and LFPR, cannot be integrated in the usual sense, see the discussion in Granger (Citation2010), in many theoretical and applied studies they are modelled as I(1) processes. Cavaliere (Citation2005) and Cavaliere and Xu. (Citation2014) show that in the presence of bounds, unit-root tests based on standard asymptotic critical values become oversized. Albulescu and Tiwari. (Citation2017) is one of the scarce studies that applies this methodology using unemployment rates. As far as our analysis is concerned, we believe that in light of Cavaliere (Citation2005) and Cavaliere and Xu. (Citation2014)’s findings (i.e. traditional unit-root tests are oversized), not rejecting the null hypothesis further strengths our findings.
2 The cointegration test used is the Johansen implemented in the Stata 14.0 package using the vecrank and vec commands.