Strikes in the United States, 1881-1972

Abstract

This thesis examines strike activity in the United States from 1881 to 1972. Although part of its aim is simply to describe and analyse trends of activity, its main concern is sociological: the explication and testing of the theory of the institutionalisation of industrial conflict. Since official statistics on strikes are the basic data employed, chapter one considers the problems with the use of such statistics. Phenomenologists criticise the use of all official data, on the grounds that there is no 'true' distribution of social phenomena independent from the reality of everyday life, and that official figures are merely records of bureaucratic procedures. But, unlike crime statistics, strike figures measure a true distribution of activity. They simply record how often labour-management disputes develop into collective and complete stoppages of work; the strike is a category used in everyday life and its meaning as a tactical weapon in disputes is clear. Although the strike, as a 'stoppage of work to express a grievance or enforce a demand', cannot be defined independently from actors' meanings and intentions, this intentionality aspect is not a crucial problem: given the strike's role in the employment relationship, the inferring of intentions is not difficult. The main problems are practical; for example, it is hard to know where one strike ends and another begins. Such problems can be by-passed, however, if we concentrate on trends in activity. Since the different American recording agencies have used very similar definitions and procedures, their data are statistically reliable, especially for workers involved and 'days lost' in strikes. Their validity is more questionable, but, since we are concerned with trends and not with strike statistics as indices of the absolute amount of conflict at one time, we may proceed to use the official data, although considerable caution will be required regarding information on such things as the 'causes' of strikes. Chapter two examines the theory of the institutionalisation of industrial conflict. Its 'conventional' version, developed to its fullest extent by industrial relations writers of the 1950s, argues that institutions will grow up to channel and thus to control conflicts of these institutions, collective bargaining is the most important. As unions and management learn to co-operate, strikes become tactical weapons in the bargaining process and lose any wider political significance; not only are they institutionally controlled, but their frequency and intensity fall as institutional means are found for the resolution of conflict. One strand of the conventional view stresses the importance of industrialisation for the amount of conflict; in the 'early' period activity will be high because of the disruption caused by rapid industrial change, but subsequently workers will become accustomed to industry, and conflict will become less intense. This view is taken up by Shorter and Tilly in their book Strikes in France 1830-1968. They set out to chart the course of strikes from 'early' to 'modern' forms of protest and, although they criticise previous writings on the effect of industrialisation, can be taken as giving a distinct variant of the institutional isat ion theory. They stress the political role of the strike and thus widen the scope of the theory; and, in arguing for a transformation of the strike into a political demonstration instead of an economic battle, they follow the broad lines of the theory. Finally, a 'radical' version of the theory is identified. This grew up in response to the conventional line and argues that institutionalisation has not taken place on equal terms; there is a fundamental inequality of power in capitalist societies which has meant that the trade unions have been incorporated into the system on terms set by employers and the government. Workers have become enmeshed in a system over which they have little real control. This set of theories, being concerned with long-term trends in labour relations, is clearly suitable for testing against strike data. The central prediction is that the frequenty and duration of activity will decline as instituionalisation progresses, and strikes will also become more 'predictable' and 'orderly'. A weaker argument is that the effect of institutionalisation on industrial conflict will be contingent on economic circumstancesj we must therefore examine economic influences on strikes and whether any change in their impact can be attributed to instituional developments. It is also necessary to examine trends within industries and regions, to see if trends at a more detailed level than that revealed by the aggregate statistics support the instituionalisation theory. Chapter three tests the theory against the data on strike trends for 1881-1972. The notable feature of American strikes has been their failure to change 'shape': their frequency, size and duration have remained remarkably constant. This basic finding sets the framework within which all other results must be examined. Thus the observed rise in union involvement in strikes fits the theory's predictions, but does not counteract the fundamental weakness of the view that the amount of conflict will decline. Trends in the issues in strikes show the expected increase in strikes for union recognition during the 1930s, when collective bargaining was being established nationally, but these recognition strikes have not disappeared subsequently. Within industries, one would expect activity to peak when bargaining was being established, but to decline thereafter; but an examination of long-term trends in several sectors showed that such a pattern was unusual. Similarly, regional trends in activity do not fit the institutional isat ion model. Thus, changes in activity cannot be explained by either of the processes identified by the 'conventional' theory, namely collective bargaining and industrialisation. Neither is it possible to suggest that trends towards institutionalisation will be limited to particular industries or regions. Chapter four investigates the more subtle argument that the nature of economic and other influences on strike activity will be affected by the degree of institutionalisation. Thus the period 1946-72, during which institutionalised relations have been established, is compared with two earlier periods, representing the years of rapid industrialisation (1881-1910) and the period when bargaining was not institutionalised (1900-39). Using a multiple regression framework, the overall explanatory power of the economic determinants of strike activity does not vary markedly between these periods; but there is some evidence to suggest that the direction of the influence of particular variables has changed. The fact that a change occurred fits the institutionalisation model, but the precise direction of the change cannot be explained by it. Strike activity has been broadly influenced by economic conditions, but this link has not been direct; strikes are associated not with the current unemployment rate but with measures of the height and depth of previous business cycles. A general, and not an immediate, influence is at work. Shorter and Tilly expect that the degree of working class organisation will be the crucial variable intervening between economic conditions and strikes, and that strikes up to the New Deal should reflect political and not economic influences. Detailed examination of several models of strike activity suggests that these expectations are unfounded; the influence of union density on strikes has been relatively weak, and political and economic variables have not operated in the predicted directions. Thus any shift in the 'institutional parameters' of conflict has not altered the importance of political and economic influences on strikes; rather, activity has continuously reflected economic conditions, although the precise direction of the effect of these conditions has changed. Chapters five to seven consider three key periods for industrial conflict. The first, relating to the period 1881-1905, concentrates on the effects of industrialisation on strike activity. Trends over time in the amount of activity are not explained by industrialisation, and nor can the detailed characteristics of this activity be explained by the impact of industry. In broad terms, this can be explained as the result of the simultaneous operation of two factors. On the one hand, strikes were becoming increasingly 'sophisticated' as workers learnt to use them as instruments in planned campaigns: for example, stoppages tended to grow larger and more successful. On the other, strikes were used to resist attempts by employers to destroy the day-to-day control of operations enjoyed by craftsmen; this meant that strikes tended to be small, since they were limited to specific plants and groups oi workers, and unsuccessful. Hence, no overall trends on the dimensions of 'sophistication' can be seen. As one would expect from the argument about job control, strikes increasingly came to reflect the struggles of skilled men. Craft workers had the most to lose from the introduction of mass production methods, and were in the strongest position to put up resistance. Hence, strikes were concentrated in the least technically advanced industries. But, again, no simple pattern emerged; for example, within industries there was no systematic tendency for increasing mechanisation to be associated with distinctive patterns of strike activity. The struggle for job control was often a lengthy one which bore no relation to immediate conditions. Similarly, a comparison of skilled and unskilled occupations reveals that no simple pattern of response (such as the 'successful' and 'unsuccessful' craft responses suggested by Shorter and Tilly) existed. Thus, although industrialisation had broad effects on strike patterns, there was considerable diversity

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 69 citations exist upstream.