The Relative Power of Organized Labor: Difference in Arguments and Findings

Kim et al.’s argument is that organized labor strikes require concurrence with riots and demonstrations to effectively press democratic governments to make social policy concessions. Thus, strikes alone cannot exert any influence on governments’ social spending choices, and only when strikes coincide with riots or demonstrations do democratic governments respond to organized labor’s social policy interests.

Kim et al. analyze the effect of four different instances of (non)concurrence of protest events on social security and welfare spending (SSWS). They find that strikes alone do not affect SSWS, strikes concurrent only with riots positively affect SSWS, strikes concurrent only with demonstrations do not influence SSWS, and strikes concurrent with both riots and demonstrations have a positive effect on SSWS. However, this latter effect is likely to be masking the effect of strikes and riots, since concurrence with demonstrations has no significant effect on SSWS.Footnote ¹ Thus, compared to my analysis, Kim et al.’s finding of a positive effect of strikes concurrent with riots is the most differentiated, given that I did not focus on concurrence of events. However, I did find that strikes have a positive effect on social security and welfare spending. I develop more on this below.

I find that the explicit focus on the concurrence of protest events is a novel way to analyze the impact of mobilization on social policy in particular, and on public policy in general. Theoretically, however, several questions remain unclear and could be further addressed. Kim et al. agree with my argument that, compared to the masses, organized labor is in a relatively better position to effectively press the government to obtain concessions, it organizes around specific claims, and its main mobilization strategy—strikes—carries significant economic and political costs. In contrast, mobilization by the masses is highly heterogeneous, usually covers different issues, and is often subject to more effective demobilization tactics from governments.

My argument speaks about the capacity of organized labor protest to expand social policy concessions that correspond to its social policy preferences (i.e., SSWS), in relation to collective protest from less organized social groups that would benefit from more progressive social spending (i.e., human capital spending). In this regard, I highlight that very often, large-scale mass protests have seen the participation of different social groups, including critical participation or even leadership of organized labor (2014, 1951). The emphasis here is that protest by the masses involves diffuse demands and is more easily demobilized by authorities, while organized labor usually articulates specific policy demands, and strikes can have high costs for governments due to their disruptive nature.

Accordingly, it is less likely that as a result of their protest, the masses will obtain social policy concessions that benefit them in the form of human capital spending. In my analysis, I found that the number of strikes has a long-term effect on social security and welfare spending, while protest by the masses (i.e., the sum of demonstrations and riots) does not lead to increases in any type of social spending. Importantly, however, demonstrations and riots prevent cutbacks in human capital spending.Footnote ²

Kim et al. instead argue that even if mobilization by organized labor has intrinsic advantages, it is at a numerical disadvantage that can only be overcome by the concurrence of strikes with protest by the masses. As in my article, Kim et al. consider riots and demonstrations to be protest by the masses. In their view, concurrence takes place because organized labor grievances resonate with the general public, so other groups join their claims using other mobilization tactics, such as demonstrations and riots.

The second possibility is that concurrence takes place regardless of shared grievances; but even in this case, the higher level and intensity of social conflict leads governments to respond to organized labor’s interests by increasing the type of spending that benefits organized labor (i.e., social security and welfare spending). A third possibility, mentioned only marginally by Kim et al., is that organized labor actors participate in and lead both strikes and riots. From a strategic standpoint, organized labor can use simultaneously different tactics to pursue its demands, or strikes can evolve and turn violent, giving place to riots, as a response to government actions. This is actually not uncommon in Latin America, where unions are known to be highly militant, and their interactions with police during protest activities often involve violent clashes with public security officers. I acknowledge that I did not consider this possibility in my paper, and it was inaccurate to categorically conceive riots as protest by the masses. Below I consider this third possibility in a more systematic way.

Linking Theory to Data: The Predominant Role of Organized Labor

If strikes and riots are not independent events, and those events where Kim et al. observe strikes concurring with riots follow the possibilities mentioned above (i.e., organized workers using strikes and riots simultaneously to press their demands, or strikes evolving and ending with some level of peace disturbance and violence), it is plausible to argue that organized labor actors are the key drivers of strikes and riots. If this is the case, however, Kim et al.’s claim that organized labor requires the masses to effectively press the government is not entirely accurate, as it does not recognize that organized labor actors take the leadership of contentious events. An alternative interpretation would suggest that governments are more pressured to placate labor and make concessions when strikes turn violent or show some degree of public peace disturbance, either provoked by challengers or as a response to state repression. This is critical to explaining the finding that the concurrence of these events only affects social security and welfare spending (the type of spending that benefits organized labor), but does not influence other types of spending that tend to benefit the masses (i.e., education and health).Footnote ³

This does not neglect the significance of concurrent events, which might be relevant for certain social and public policy outcomes. However, the question here is whether nonlabor actors lead riots that concur with strikes. If this is the case, then one of the mechanisms noted by Kim et al. is at work: regardless of the specific grievances in the rioting of nonlabor actors, the higher level of conflict as a consequence of the concurrence of strikes and riots amplifies the demands of labor.Footnote ⁴ If the answer is negative, then it is likely that strikes evolve and turn violent, giving place to riots, or that workers use both tactics, strikes and riots, in their repertoire of contention. Under these scenarios, strikes and riots are predominantly led by organized labor.

To shed light on this question, I dug into the news articles behind the cases coded as “strikes concurrent only with riots.” Interestingly, I found in the narratives strong evidence suggesting that strikes are accompanied by violent clashes with the police or violent tactics (e.g., burning of vehicles). In these cases, strikes and riots tend to be led by organized labor actors. Kim et al. report 19 cases of strikes concurrent only with riots. Of these, 15 cases occurred in countries coded as democratic.Footnote ⁵ I listed the countries and years when these cases took place and searched for the news articles in LexisNexis Academic using as the news source the New York Times. Footnote ⁶ I briefly mention these cases below.

1. In 1988, about 85,000 rail workers went on general strike in Brazil demanding a pay increase of 60 to 80 percent (New York Times 1988a). In the same year, there was another strike in the oil industry in which workers clashed with troops, violence surged, and three workers died in the confrontations (New York Times 1988b).

2. On February 28, 1971, the Colombian Workers’ Union threatened to go on a general strike. The threats were accompanied by farm invasions by peasants and rioting with a high level of violence, in which 8 confrontations ended in death and more than 50 people reported being hurt. A week later, the general strike took place, and the news media reported that the strike was not successful, except for urban transport (New York Times 1971a, b). Years later, in 1977, another general and highly violent strike involved about 18 deaths and hundreds of arrests. In this case, the news media reported that the strike “marked the first time in about 20 years that organized labor has formed a united front across political and ideological barriers…. Unions had called the strike to protest runaway inflation that has doubled food prices and may reach 35 percent in ’77” (Vidal Reference Vidal1977). This report suggests the participation of other groups beyond labor, but it was organized and orchestrated by labor unions. The two other cases in Colombia also involved a general strike with a high degree of violence and hundreds of people arrested (New York Times 1981; Neilan Reference Neilan1999). During 1999’s indefinite general strike, the severe levels of violence caused workers to call the strike off when, after the second day, more than 70 people had died.

3. Similar events were reported in the Dominican Republic, where organized labor strikes involved violence and deaths (New York Times 1989a, b).

4. In Ecuador, general strikes against economic measures involved workers’ blockades of roads and led to student demonstrations, which ended with violent clashes between students and the police and the closing of schools and universities (New York Times 1982, 1983a).

5. In Guatemala, the news media reported a week of strikes in the transportation sector, which ended violently with the death of nine persons and hundreds arrested and injured in rioting (New York Times 1978).

6. The Nicaraguan case also illustrates the finding that strikes often involve some level of violence and rioting. In July 1990, Sandinista-led government employees went on strike and, along with students, blocked major roads in the capital, using violence and burning vehicles (Uhlig Reference Uhlig1990a). The strike was part of a series that had occurred earlier in the year, after which the government had proceeded to grant important concessions and to accommodate workers’ demands.

Despite the agreement, workers increased their demands and continued striking (Christian Reference Christian1991).

7. In Peru, workers called for a 24-hour strike against economic policies that involved violent clashes with the riot police forces, the killing of one demonstrator, and dozens injured (New York Times 1983b). During 2003 and 2005, Peru experienced demonstrations and discontent over economic policies and corruption scandals against President Alejandro Toledo. Among the most militant groups were teachers’ unions, which went on strike; similar to the cases above, their strikes also ended with violent clashes between protestors and police across the country (Reuters 2003).

These cases reveal that indeed, labor strikes involve a certain degree of violence and that labor actors often resort to riot tactics in pursuing their demands, in contrast with Kim et al.’s claim that labor actors need mass mobilization and support from nonlabor actors to effectively press democratic governments and gain concessions in the form of social security and welfare spending.