There is a widely-held view that top union leaders in Argentina often call strikes not only in response to bread-and-butter issues like wages, benefits, working conditions, and job security, but also in response to political and organizational factors like electoral considerations, attempts to put the government on the defensive, efforts to influence public policy, conflict between union leaders and base-level militants, and struggles among factions of the national union leadership. The proposition examined here is that these noneconomic causes make themselves felt primarily in strikes called by top union leaders, whereas bread-and-butter issues tend to prevail in strikes called by local union leaders. This central thesis concerning the “scope” of strikes is connected, in turn, to a methodological point. It will be argued that the size of strikes is measured more validly by the level of the union leadership at which the strike is called than by the number of workers involved in the strike — at least in countries, like Argentina, where the trade union movement is organized primarily along industrial rather than craft or enterprise lines.
To flesh out and test these propositions, I compiled a database of 3,116 strikes in Argentina between January 1984 and June 1991. Each strike was coded according to month, site, economic sector (metalworking, public administration, sugar milling, teaching, etc.), scope (national, provincial, municipal, enterprise, workplace, etc.), province, stated cause (if available), number of workers involved, and duration. The data were then aggregated by month and by quarter, and the monthly and quarterly aggregates were regressed on independent variables measuring economic and political factors plausibly related to strike rates. The results of these analyses confirm that economic factors predict monthly and quarterly variation in the number of small strikes better than such variation in the number of big strikes. Political variables, however, appeared to be related only weakly to strike frequencies, perhaps because the instruments used to measure them were too blunt for the task.