Is the government's intertemporal budget constraint fulfilled in Sweden? An application of the Kalman filter

This paper deals with the issue of whether the Government complies with its budget constraint for the case of Sweden during the period 1963-2000 using quarterly data. It is found that government spending and government revenue are nonstationary (integrated) but cointegrated. A random coefficient model against a fixed coefficient model is also tested for. The results show that the random coefficient model, which is nonlinear, is preferred to the linear fixed parameter model. This model, which takes into account the Lucas critique, is estimated by the Kalman filter and it provides further empirical evidence that the government follows its intertemporal budget restriction.