A dynamic contracting model with endogenous prices, wages and investment is presented. It accounts for capital and labour underutilization at steady state and displays various patterns of unemployment rate and capacity utilization rate co-movements as a function of demand and supply shocks. The main determining elements of unemployment are the unit cost of capital, the intensity of competition in the product market, the magnitude of uncertainty and union characteristics.

In this paper, we intend to offer a critical review of the different concepts of capital and the different methods which are used to measure it. At the same time, we propose an endogenous approach which tries to measure the capital stock in terms of value, explicitly referring to depreciation as an economic variable decided by the firm. Applying this procedure to the Spanish macroeconomic data, we observe a big capital destruction since the mid-seventies which is not mentioned in official statistics.

This paper deals with the question whether uniform provision of a purely private good should be implemented at the social optimum, in a second best economy where personalised lump-sum transfers are not feasible, but no market failure exists. The answer is that it depends on the balance of private and social gains and losses from public provision: necessary conditions for the optimality of uniform provision are derived and discussed. The sensitivity of these conditions to changing rules for the choice of the user charge is also investigated, and it is suggested that public provision is most likely to be optimal when it is free of charge. Finally, it is emphasized that the assumption that a perfect substitute for the publicly provided good is not available plays a crucial role in obtaining the above results.

This paper considers a special non-linear time series problem, that of testing for co-integration in a Bayesian framework when there is a break in the co-integrating relationship. It is shown that a partial linearization of the likelihood function solves many puzzling questions, in particular identification and common factor restrictions which are originally imbedded in the model. A generalization of the Jeffreys’ prior is derived for the dynamic parameter which monitors co-integration. The procedure is applied to a one time much debated question in France which concerns the wage regulation policy implemented at the beginning of the eighties.

Cet article traite de l’effet d’intervalle sur le marché à terme de la Bourse de Bruxelles. L’analyse empirique est menée sur les actions belges cotées sur ce marché pour une période allant de l’introduction du Computer Assisted Trading System (C.A.T.S.), début 1989, au 31 décembre 1992. La présence d’un effet d’intervalle est mise en évidence, l’estimation du risque systématique des 19 actions composant l’échantillon dépend de la longueur de l’intervalle de temps choisi pour calculer les rendements. L’estimation des betas à l’aide d’un modèle GARCH ne donne pas de meilleurs résultats que la régression linéaire simple et les valeurs des betas dépendent toujours de la longueur de l’intervalle de temps. L’effet d’intervalle semble donc ne pas être dû à la présence d’hétéroscédasticité conditionnelle. Il ressort également que, sur base de données journalières, les modèles proposés par Scholes et Williams et par Dimson qui ajustent le beta pour tenir compte de la faible fréquence des échanges, donnent aux coefficients une valeur plus proche de leurs valeurs asymptotiques que le fait la régression linéaire simple.