Okun’s law: evidence of 13 selected developed countries

Abstract

This paper re-examines the validity of Okun’s Law for 13 selected developed countries over 1970–2013. λ _maxand λ _trace tests reveal cointegrating relationship between unemployment rate and real GDP growth in all countries except Germany. However, dynamic OLS estimates portray a somewhat different picture. The estimates of bivariate error-correction model (ECM) unveil that Okun’s Law is quite valid only for the USA and South Korea. At the same time, evidence is relatively weaker for Canada, Finland, France, Japan, Italy, The Netherlands, New Zealand, Sweden, UK and Australia. However, it is invalid for Germany.

Appendix

(Derivation of Growth Rate Form of Okun’s Law).

In its original form, Okun’s Law is:

$$ \uptheta \left(\mathrm{U}*-\mathrm{U}\right)=\left(\mathrm{Y}\hbox{--} \mathrm{Y}*\right)/\mathrm{Y}* $$

(4)

Where, U is the unemployment rate, Y is real GDP, U* is natural rate of unemployment, Y* is potential real GDP and θ is the Okun’s coefficient.

To derive the growth rate revision, equation (4) is expanded as follows:

$$ \theta U*-\theta U=\mathrm{Y}/\mathrm{Y}*-\mathrm{Y}*/\mathrm{Y}* $$

(5)

Totally differentiating equation (5) and assuming Y* constant at full employment, with respect to all variables,

$$ \theta dU*-\theta dU=\mathrm{dY}/\mathrm{Y}*-\mathrm{dY}*/\mathrm{Y}* $$

(6)

For simplification, if the natural rate of unemployment is unchanged, dU* = 0. Now, rearranging the terms,

$$ \mathrm{dY}/\mathrm{Y}*=-\theta dU+\mathrm{dY}*/\mathrm{Y}* $$

(7)

Usually, Y will be close to Y* in the long run – within a few percent, at least. Therefore, dY/Y* can be approximated by dY/Y, the growth rate of GDP. Finally, in growth rate version,

$$ \mathrm{dY}/\mathrm{Y}=-\theta dU+\mathrm{dY}*/\mathrm{Y}* $$

(8)