A structural analysis of expectation formation based on business surveys of French manufacturing industry

Marc Ivaldi, University of Pennsylvania


Using panel data of individual firms drawn from French Surveys, a structural analysis is developed to study the formation of production plans and the rationality of expectations. These two subjects are central issues for the study of the business cycle and for the design of macroeconomic policy. The production decision of a firm is defined as the optimal solution of a dynamic stochastic optimization problem. The empirical work amounts to recovering the structural parameters characterizing the model of the firm from estimates of the reduced-form model which is here the derived decision (feedback) rule. To perform this task using survey data, which consist only of qualitative information, requires a specific econometric analysis. First, the variables entering the theoretical model are interpreted as latent variables which, in turn, reveal the categorical survey responses when they cross certain thresholds. Second, the statistical specification of the economic model is obtained by means of an errors-in-latent-variable model. Then, to estimate such an econometric model, a two-step procedure is applied. First, the correlation matrix of the latent variables is estimated using the theory of the polychoric correlation coefficient. Second, this correlation matrix is the input for a method of moments: the estimates of the model parameters are derived either by maximum likelihood or by weighted least squares when a correct asymptotic covariance matrix of the correlation coefficients may be obtained. The estimation results support the idea that French firms smooth production in the short run, because adjustment costs of production have higher relative weight than those of other components of the cost function. Production is less variable than demand, in part because cost shocks are much less important determinants than changes of demand. Consequently, macroeconomic fluctuations should be related to the smoothing of supply shocks. A by-product of this study is the use of the estimated model for forecasting the change of industrial production directly from the survey data. The preceding analysis of production plans is based on the assumption that firms are rational. In order to give some justification for this assumption, direct tests, i.e., free of behavioral assumptions, are performed using an original testing procedure in the context of latent variable models. They offer evidence that the Rational Expectations Hypothesis may not be rejected for quantity variables such as changes in demand or production. Tests are not conclusive for price but tend to support rejection. Finally, the study points out that a careful treatment of the missing data problem inherent to survey data is needed to advance towards their systematic use in econometric analysis of the business cycle.

Subject Area


Recommended Citation

Ivaldi, Marc, "A structural analysis of expectation formation based on business surveys of French manufacturing industry" (1990). Dissertations available from ProQuest. AAI9026583.