The dual approach to the public capital hypothesis: the case of The Netherlands

This paper uses a flexible functional form approach to account for the direct and indirect effects of public capital on the private sector. Furthermore we explicitly incorporate energy in the production function. We use the duality between production and cost functions and base our investigation on the estimation of a translog cost function for five manufacturing sectors of the Dutch economy. Our primary objective is to see whether public capital affects the production cost and factor-demand equations of the private sector and estimate several public capital elasticities. Our preliminary results presented in this paper should be interpreted with great care. In the first place there is severe autocorrelation in the model which means that the model is not specified correctly. In the second place the second-order conditions for cost minimization are violated in a number of cases. The problem of autocorrelation is solved by introducing a dynamic version of the model without public capital. The conclusions from the dynamic model are twofold. Autocorrelation is reduced, but the first and second-order conditions for minimizing costs are violated. Our results indicate that assessing the importance of public capital using a flexible cost function approach is difficult. The results found in the literature are not unambiguous either. Furthermore, autocorrelation might possibly be solved by rewriting the model as a truly dynamic model. However, in order to model the adjustment process a proper identification of the optimal shares is needed. Our impression is that the cost function approach does not provide us with reliable estimates for the optimal shares.