Test d’une hypothèse d’investissement à écarts multiples

According to the multiple-gap hypothesis, investment decisions, determined by the joint action of multiple determinants, result from the dominance at the time of the decision of the most constraining of these factors. This hypothesis is tested on Dutch investment data for the period 1951-1974 by means of three types of models: a "rigid" model in which a single regime is selected at any time, and two "flexible" models allowing for a joint action of the various determinants, the last model incorporating a switching mechanism with respect to the dominant factor. Estimation problems are discussed and the results obtained with the different models are compared.