A spatial econometric multivariate model of Okun's law

A system of two dynamic spatial panel data model equations is developed in which output growth and the change in the unemployment rate are interdependent. The parameters of the model are estimated by recently developed maximum likelihood techniques for multivariate spatial econometric models, using data of twelve provinces in the Netherlands over the period 1974-2018, covering four major economic downturns of the Dutch economy. By using time-cumulative marginal effects derived from the impulse response function of this model, it is found that Okun's law is dominated by the relationship that runs from output growth to unemployment. The amount of growth that is needed to reduce unemployment by one percentage point is shown to depend on the extent to which spillover effects to neighboring regions and output multiplier effects are accounted for.