Activité économique, inflation et activité de grève

Après avoir discuté les bases théoriques sur lesquelles la majorité des auteurs ont appuyé leurs études des relations entre l'activité économique, l'inflation et l'activité de grève et après avoir spécifié leur modèle, les auteurs résument les résultats empiriques des études antérieures et présentent leurs propres résultats.The theoretical explanation of a possible relationship between economic activity, inflation and strike activity must deal with the impact of economic cycle on the behaviour of firms and employees. This behaviour is related to the costs of a wage agreement; since these costs are related to economic activity and inflation, we can expect to find a relationship between the last two and strike activity.However, our examination of the problem led us to conclude that forces motivating the parties involved, tend to be contradictory for any given level of economic activity and/or inflation. The agressiveness of one of the parties tends to be cancelled by the other's flexibility. Given the present state of knowledge, the final outcome therefore appears a priori, to be undetermined. The impact of economic activity and inflation on strike activity thus becomes an empirical matter.Our first task was then to define what is meant by strike activity. In so doing, we found that the different statistical data describing this phenomenon could be related into a simple logic and consistant model.Four equations are needed to integrate, and serve as an analytical basis for the different aspects of strike activity. Two behavioural (functional) equations related to decision-making are based on the number and duration of strikes. The two other equations are identities. The first identity exogenises the average size of strikes. The second, the overall variable of man-days lost, is the simple product of the number, length and average size of strikes.We used the ordinary least squares for estimating the functional equations. The period studied ranges from the first quarter of 1967 to the second quarter of 1974.The explanatory factors used in the number of strike equation are the deviations of the GNP (in constant dollars) from its long term trend, inflation, lagged nominal wage variations, and the number of negociations. To the best of our knowledge, this last variable appears for the first time in this type of equation.The average duration of strike equation is formed from the same variables (economic activity, inflation and lagged wage variations) with the exception that the number of negociations is replaced by the number of employees involved in strikes. This variable should reflect the financial and organizational burden of strike activity upon unions.The statistical results obtained were quite satisfactory. A negative relationship was found between economic activity and strike activity, which contradicted the findings of previous research. This tends to confirm, in our opinion, the empirical nature of the relationship. On the other hand, we found a positive relationship between the economic cycle and the length of strikes. Since this variable has a considerably greater effect on man-days lost, we finally found a positive relationship between economic activity and strike activity. This reconciles us with the works of Vanderkamp 1.Furthermore, it was found that strike duration was sensitive to inflation but that was not true for the number of strikes.Finally, the variable for wage variations was found to have a significant positive influence, both on the number and average duration of strikes. Also, the number of negociations and number of employees involved in strikes both played a significant role. On the whole, 90% of the variance in man-days lost could be explained by the model