This paper revisits the growth-rate version of Okun’s law by explicitly distinguishing between its demand-side and supply-side formulations and assessing their implications for the measurement of potential output and output gaps in OECD countries.
Using annual data for 14 OECD economies over 1960–2023, the paper estimates a demand-side formulation linking changes in unemployment to output growth and a supply-side formulation linking output growth to unemployment changes, allowing for multiple structural breaks identified with the Bai–Perron methodology.
The results reveal significant heterogeneity and structural instability in the output–unemployment relationship, with regime changes dated endogenously in each direction. The supply-side formulation is relatively stable and closely replicates conventional (AMECO) estimates of potential growth, whereas the demand-side formulation yields systematically higher potential output, suggesting that persistent demand constraints may have reduced both actual and potential growth since the late 1970s.
The measurement of potential output is highly sensitive to the theoretical framework adopted; NAIRU-based approaches may underestimate productive capacity under prolonged demand weakness.
The results are consistent with the view that macroeconomic policies focused primarily on inflation control involved persistent output losses. While the research design does not identify this effect causally, it establishes the quantitative relevance of the question and indicates that demand management may have lasting effects on both employment and growth.
By showing that conventional measures of potential output may understate productive capacity, the results bear on how much room for growth and employment policymakers believe an economy has, with consequences for the design of fiscal and monetary policy.
The paper integrates two interpretations of Okun’s law within a unified empirical framework, formalises why the two regressions are distinct statistical objects whose estimated slopes cannot be reciprocal and links them explicitly to the measurement of potential output, showing that the formulation adopted drives the economic and policy conclusions.
