Paper short abstract:

Within the development studies framework, this proposal addresses the relevance of considering that the effects of economic growth on unemployment may be asymmetric and unevenly distributed across the population (i.e. gender or age), as well as tricky to measure using econometric methods.

Paper long abstract:

Okun’s law is one of the best-known empirical regularities in macroeconomics. It addresses the issue of how much a country’s output is ‘lost’ when unemployment exceeds its natural or trend rate. In the existing literature, most of the studies assume a linear long-run relationship between changes in output and unemployment, which implies that economic expansions and recessions have the same effect on unemployment. Nevertheless, this negative relationship may take a nonlinear form, in the sense that changes in output may cause asymmetric changes in unemployment. Consequently, the objective of this paper is to test the linear assumption of Okun's law by deploying a time series analysis of asymmetric adjustments using seasonally adjusted quarterly data from Chile for the period 1996-2019 disaggregated by sex. The analysis is based on an asymmetric error-correction model, which provides a different approach to describe the short-term dynamics between the variables, but at the same time presents different challenges for its implementation. Thus, beyond the analysis of the data, the study explores advantages and disadvantages of procedures that require cointegrated non-stationary time series to establish a long-run equilibrium. Finally, the relevance of this research lies in the consequences of a misinterpretation of the effects of business cycles on unemployment if asymmetry is ignored. For instance, the existence of asymmetric effects shows different speeds of adjustment towards long-run equilibrium. Therefore, the effectiveness and required ‘size’ of stabilization policy on the real economy will depend on the ‘regime’ in which the Okun’s relationship is found.