Testing Okun’s law with Swiss industry data

Abstract

Okun’s law postulates a stable relationship between quarterly output growth and changes in (un)employment. This proposition has so far been tested with macroeconomic data at the highest level of aggregation. The article goes beyond that in extending the analysis to industry data from Switzerland, applying a method suggested by the International Monetary Fund. Another focus is on whether expansions in production have become more ‘jobless’ over the most recent business cycle compared to earlier ones. This does not seem to be the case in Switzerland, except in the construction industry.

Keywords:

• Okun’s law
• jobless growth
• industry data
• Switzerland

JEL Classification:

• C22
• E12
• E20

Acknowledgements

I would like to thank my colleagues at KOF Yngve Abrahamsen and Michael Siegenthaler for their generous input to this article and Dario Florey and Philippe Stauffer from the SFSO for suggesting an explanation for the strong productivity growth in the Swiss insurance industry. Two referees for this journal also provided valuable insights and suggestions. All remaining errors are mine.

Notes

¹ In fact, Okun (1962) presents a second version of the law, which postulates a stable relationship between the output gap and the deviation of the unemployment rate from its ‘natural’ level, taking a rate of 4% as a proxy for the latter.

² The short peak-to-peak cycle between 2001Q2 and 2002Q2 is ignored; instead the whole period from 2001Q1 to 2008Q3 is considered one peak-to-peak cycle.

³ The Economist (2014), comparing the German and British educational systems, pointed to the trade-off between vocational training and flexibility.

⁴ The 95% confidence intervals around the estimated dynamic betas were calculated assuming that the lagged variables are pairwise uncorrelated. In this case, the variance of the sum of the variables equals the sum of their variances.