Pierre Villa

Consistent Conjectural Equilibria and Economic Policies Coordination

In the recent literature on optimal policy coordination, the essential part is played by the comparison between Nash equilibrium, obtained when each country uses its policy instruments in order to minimise its loss function by taking as fixed its partners’ policy, and the Pareto equilibria, obtained when the countries cooperate efficiently. But the notion of Nash equilibrium is unsatisfactory for two related reasons. In a Nash equilibrium, economic authorities are assumed to consider that their partners do not react to their actions even though in reality they do. The Nash equilibrium depends upon the way each country imagines its partners’ policies, because each country is supposed to reason as if these policies were unchanging. But what do we call an unchanging economic policy? Is it the stability of instruments or the stability of the intermediate targets of economic policy?

A joint estimation of demand for investment and labor* *Originally published as “Une estimation conjointe des demandes d'investissement et de travail,” Annales de l'INSEE 38–39:237–258, 1980.

Publisher Summary The joint estimates of factor demands on annual and quarterly data bear out the results obtained for investment: (1) returns to scale increase, (2) the influence of user cost corresponds to a low elasticity of substitution when the model is putty-putty, and (3) adjustment lags are shorter for relative factor costs than for demand. Although they comply with this ranking, the adjustments for employment are consistently shorter than for investment. The quarterly-data estimates furnish a sizably smaller assessment of the elasticity of substitution than the annual-data estimates. This finding can be easily interpreted if the true model to be of the putty-clay type is considered. In this case, the putty-putty specification measures the apparent elasticity that comes into play in the ex ante choice of production technique. For a same lifetime and a same long-term elasticity of substitution, this apparent elasticity is four times smaller at the quarterly level than at the annual level.