Abstract
Debate over the effect of technology and automation on job creation or job destruction has
been an ongoing debate in economics for some time. The recent developments in automation
and the speed with which machine is replacing labor in some industries has worried many
economists (Krugman, 2013, Graetz, G & G Michaels, 2015). While a number of recent
studies present evidence on the negative effect of automation on employment by occupations
(Oschinski M & R Wyonchi, 2017), none presents empirical evidence on the effect of
automation on jobs at the macroeconomics level. This study utilizes the traditional model of
the relationship between the real GDP growth and unemployment rate, estimated for US
economy in 1962 and publicized as Okun’s Law. The relationship implies that a one percent
increase in GDP growth above the normal growth of GDP results in .4 percent decrease in
unemployment rate. Although the relationship between GDP growth and unemployment rate
may be affected by other economic variables in the short-run, a variable that may result in
structural change in this relationship in the long-run is technology and its effect on
unemployment. Technological advancement may lead to substitution of capital for labor,
resulting in less response of GDP growth to unemployment rate than what Okun’s law
proposed. The main objective of this paper is to test whether such a structural change in the
relationship has occurred or not. Using data for the last sixty years for eight industrialized
countries, this paper compares the average response of unemployment rate to real GDP
growth in three decades of 1955-1985, with the recent three decades of 1986-2015.
