Employed Americans often spend more time at their job than at any other location outside their home (US Bureau of Labor Statistics 2023). While at work, individuals form bonds with their coworkers and managers, forge social and political identities, gain skills and resources that they can apply in other areas of their lives, and receive political appeals and communications (see, e.g., Estlund Reference Estlund2003; Frymer and Grumbach Reference Frymer and Grumbach2021; Mutz and Mondak Reference Mutz and Mondak2006; Verba, Schlozman, and Brady Reference Verba, Schlozman and Brady1995). More broadly, as political theorists have argued, the workplace is itself a private government where managers govern their employees on and off the job (Anderson Reference Anderson2017; Dahl Reference Dahl1986). Yet for all that the workplace serves as a crucial site in Americans’ political lives, US-focused political scientists have been slow to consider it as such and slow to consider the political consequences of labor actions like strikes—even though there is a richer tradition of studying these topics in comparative politics (on the absence of these topics in American politics, see, e.g., Ahlquist Reference Ahlquist2017; Frymer Reference Frymer2010; Hertel-Fernandez Reference Hertel-Fernandez2020a; Mazumder and Yan Reference Mazumder and Yan2024; on comparative treatments see, e.g., Hacker et al. Reference Hacker, Hertel-Fernandez, Pierson, Thelen, Hacker, Hertel-Fernandez, Pierson and Thelen2021; Hall and Soskice Reference Hall, Soskice, Hall and Soskice2001; Iversen, Pontusson, and Soskice Reference Iversen, Pontusson and Soskice2000; Lindvall Reference Lindvall2013; Thelen Reference Thelen2015; Reference Thelen2019). This paper studies how individuals come to form preferences and opinions about relationships between managers and workers in this private government and the possibilities of labor action to improve the standing of workers in their jobs.
More specifically, this paper examines the effect of firsthand exposure to large private sector strikes. This is an especially salient question given the state of the contemporary American labor movement, which saw a resurgence of mass strikes before, during, and after the COVID-19 pandemic (US Bureau of Labor Statistics 2024). With union membership at historic lows, the question is whether these strikes might inspire additional interest in labor action and unionization. Recent work examining widespread teachers’ strikes during 2018 found that exposure to those strikes among parents increased support for the teachers and the labor movement, as well as parents’ own interest in labor action, though not necessarily through traditional unions (Hertel-Fernandez, Naidu, and Reich Reference Hertel-Fernandez, Naidu and Reich2021). But do these findings extend beyond the 2018 wave of teacher strikes to other large strikes? In particular, does exposure to large private sector strikes shape mass opinion about those strikes, interest in labor action, and support for the labor movement given the distinctive legal, economic, and political context of the private sector labor movement compared to public sector unions (e.g., Rosenfeld Reference Rosenfeld2014; Walker Reference Walker2020)? And what might a research agenda around strike effects look like?
I address both questions in this paper. I tackle the first question in the context of a recent large private sector strike. In April 2019, roughly 31,000 employees of the grocery store chain Stop & Shop went on strike across over two hundred stores in Massachusetts, Connecticut, and Rhode Island to protest benefit cuts and scheduling and pay issues (Campbell Reference Campbell2019). The strike is one of the largest in recent private sector history. After over a week of work stoppage, the union representing the grocery store workers reached an agreement with management that was widely seen as a victory for the workers (Garcia Reference Garcia2019). Throughout the strike, Stop & Shop workers made a concerted effort to publicize their grievances and demands with community members and encouraged customers to boycott the stores, join picket lines, and donate food and money (Manzhos Reference Manzhos2019).
In this way, the striking Stop & Shop workers, like the striking “Red4Ed” teachers, attempted to mobilize goodwill and broader engagement from the members of the public they served. The Stop & Shop strikes offer an important opportunity to study the mass effects of labor actions involving a different type of worker from teachers—private retail workers—with a different occupational identity and relationship to their local communities. There are several differences between the workers that may be relevant for understanding the mass effects of the Stop & Shop strikes.
First, teachers generally have postsecondary degrees and are viewed as being in a higher-status profession than retail workers. For example, in occupational prestige-scoring included in the General Social Survey, respondents rated public school teachers as being 70% higher than supermarket cashiers and grocery counter clerks on the prestige index (Smith and Son Reference Smith and Son2014). In addition, teachers likely have a closer relationship with the families of the children they teach than grocery store workers have with customers. The consequences of work stoppages are also likely to be different between the two strikes. The teacher strikes meant that parents suddenly needed to pull their children out of school and in many cases find alternative childcare, which contrasts with more modest disruptions in grocery shopping for the Stop & Shop strikes. Last, strike tactics varied as well. Although both strikes sought to mobilize public support and participation, the teacher strikes often had explicit public programming to reach and educate community members, especially parents, compared to the Stop & Shop strike.
With these differences in mind, I explore in this paper whether the Stop & Shop strikes translated into greater public support for, and interest in, the labor movement and workplace collective action. To answer this research question, I fielded an original online survey of 1,505 adults in the three states affected by the strike. While not a probability sample, the final set of respondents closely matched the demographic and political makeup of these states. The survey asked respondents for the ZIP code of their residence, which permitted me to calculate the distance between each respondent and their nearest Stop & Shop store. I use this distance as an instrument to measure their exposure to the strike in their local community. Building on past work on the effect of large teacher strikes, my survey includes broader measures of labor interest and attitudes, including more comprehensive measures of interest in labor actions in and out of unions and support for the labor movement as a whole.
Reduced form and two-stage least squares (2SLS) estimates reveal that geographic proximity to the strikes increased support for the strikes overall, and increased the likelihood that individuals took actions supporting the striking workers. I do not find consistent evidence that exposure increased the likelihood that individuals took online action to support the working conditions of retail workers at another large grocery store chain outside the Northeast or increased employed respondents’ self-reported likelihood of taking labor actions at their own jobs, with the possible exception of the strikes fostering greater interest in workers talking with coworkers about workplace issues. I do not find effects of strike exposure on individuals’ views about the labor movement as a whole and do not find effects on how individuals would vote for a union at their own job if they had the opportunity to do so.
By conceptually replicating a strike-effects study in a different context to those identified by previous research, this study suggests both differences and similarities to large public sector strikes involving teachers (Hertel-Fernandez, Naidu, and Reich Reference Hertel-Fernandez, Naidu and Reich2021). That past work suggests that striking teachers can build solidarity with the parents whose children they teach. This study suggests that large-scale private sector strikes can shape mass views in other relationships as well—in this case, between grocery store workers and their customers and communities. Yet my findings also suggest limits to private sector strikes as mechanisms for fostering greater support for, and interest in, labor action, particularly through formal labor organizations. The finding that the grocery store strikes did not increase support for other workers’ labor actions through worker organizations, the labor movement as a whole, or how a worker might vote for a union at their own job suggests that individuals may be drawing distinctions between support for striking workers in a particular labor dispute, other workers, their own personal interest in unions, and support for the labor movement as a whole. It also indicates that there may be differences between public and private sector strikes. Exploratory analysis that I report in this paper indicates that the Stop & Shop strikes may have failed to inspire perceptions of linked fate with local communities in the same way that the teachers’ strikes did, which may explain the lack of an effect on interest in labor action. The failure to foster perceptions of linked fate between the striking grocery store workers and their local communities as compared to the 2018 teacher strikes may reflect differences in the nature of the workers and their social identities, the relationships between the workers and local communities, and the demands and tactics of the strikes in question.
Beyond documenting the effects of the recent strike wave, my results contribute to literatures on collective action and social movements. First, they add to research on the political consequences of collective action, indicating that strikes can have similar effects to other social movements, like civil rights protests, on mass attitudes and other downstream political outcomes (e.g., Amenta et al. Reference Amenta, Caren, Chiarello and Su2010; Anderson et al. Reference Anderson, Charles, Karbownik, Rees and Steffens2023; Enos, Kaufman, and Sands Reference Enos, Kaufman and Sands2019; Lyon and Kraft Reference Lyon and Kraft2024; Madestam et al. Reference Madestam, Shoag, Veuger and Yanagizawa-Drott2013). These findings also shed light on how individuals learn about repertoires of activism from collective action (e.g., Wang and Soule Reference Wang and Soule2016). In particular, they underscore the organizational bases of individual activism (Han Reference Han2014; Skocpol Reference Skocpol2003). Unions, like other civic organizations, play a crucial role in teaching and disseminating the skills and tactics necessary to engage in activism in and around the workplace (Macdonald Reference Macdonald2019; McAlevey Reference McAlevey2016; Schlozman, Verba, and Brady Reference Schlozman, Verba and Brady2012)—but there may also be important limits to the spillover effects of strikes, reflecting those workers’ identities, organizations, and tactics. Last, these results engage with separate but related research on the effects of strikes on economic outcomes. Recent research has documented that while strikes in the 1970s were associated with wage gains for affected workers, after the high-profile firing and replacement of striking air traffic controllers by the Reagan administration in 1981, strikes had null effects on wages (Massenkoff and Wilmers Reference Massenkoff and Wilmers2024; see also Rosenfeld Reference Rosenfeld2006; Reference Rosenfeld2014). This paper affirms that even independent of their economic effects, scholars should be considering the impact that strikes have on the views and actions of participating workers and the communities in which strikes occur.
I conclude by laying out a broader research agenda on the effects of strikes on views, attitudes, and actions regarding labor and the workplace, drawing from both the positive and null results I identify in this paper set against past work on the politics of strikes. More specifically, I call out the need for more US-focused political scientists to study how the occupational and social identities of workers, their organizations, and their demands affect the impact of strikes on the mass public; how and when strikes backfire due to the costs they impose on the public; how political leaders can shape public perceptions of strikes and their impacts on the public; the downstream electoral and legislative consequences of strikes; and how strikes affect internal union structures and new organizing.
Data and Methods: The 2019 Stop & Shop Strike Survey
To gauge public attitudes about, and interest in, labor action in the wake of the Stop & Shop strikes, I contracted with Qualtrics, an online survey firm, to assemble a sample of 1,505 adult respondents that met the demographic characteristics of the three striking states—Connecticut, Rhode Island, and Massachusetts—along the following dimensions within each state: race and ethnicity, education, age, and gender (using estimates from the Current Population Survey).Footnote 1 The survey was fielded during May and early June 2019, about a month after the strikes. The final set of respondents closely resembled the demographic makeup of the affected region (see appendix A). Moreover, I found that the survey respondents exhibited a similar distribution of partisanship and ideology to recent Cooperative Election Study samples.Footnote 2 Nevertheless, despite these strengths to the survey, it is not a probability sample and therefore the survey respondents may well be different from the overall population in these states on other, unobserved characteristics. Importantly, the survey recruitment material did not mention unions, strikes, or worker issues, but instead mentioned that it covered current political events and issues.Footnote 3
The survey asked questions about individuals’ demographic characteristics, experiences with Stop & Shop, and attitudes toward and interest in the labor movement. In all, about 94% of respondents said that they remembered the Stop & Shop strikes. I focus on six outcomes, which I describe in table 1. They include items used by past studies on the attitudinal effects of mass strikes (support for the strike and actions taken in support of the striking workers) as well as additional items, including a behavioral measure of support for other similar workers (with the Publix workers online petition), a broader measure of interest in workplace labor actions in and out of unions, and an item asking about support for the labor movement as a whole. This permits me to conceptually replicate previous work on teachers’ strikes in the Stop & Shop strike context, as well as to test further measures of strike effects (Hertel-Fernandez, Naidu, and Reich Reference Hertel-Fernandez, Naidu and Reich2021). See appendix B for the complete survey text. My primary explanatory variable is the following: “As best as you can recall, did any Stop & Shop workers in your local community go on strike?” I code “yes” as 1 (82% of respondents indicated this response) and “no,” “not sure,” or “do not remember” as 0.
Table 1 Outcome Measures of Labor Attitudes and Actions
Note: † For the original petition, see Nolan and PL+US (Reference Nolan2018).
My research question is whether exposure to the strikes had an effect on each of these outcomes. One approach would simply regress each outcome on exposure to the strikes. Figure 1 provides these regressions, showing results from six separate ordinary least squares (OLS) models where I use as an explanatory variable self-reported measures of strike exposure and the six outcomes I described above. I find that self-reported strike exposure was related to support for the strike, actions taken in support of the striking workers, and union vote, but was not related to interest in supporting the Publix workers, interest in workplace collective action, or support for the labor movement. However, these regressions may well be biased—for instance, because individuals who encountered the strikes might have underlying characteristics that explain their exposure to, attitudes toward, and interest in labor action. For example, individuals who were more predisposed to support unions might have been more likely to follow news about the strikes, to remember the strikes happening in their local communities, and to support the striking workers—all independent of firsthand exposure to the strikes.
Figure 1 Correlations between Strike Exposure and Labor Interest and Attitudes
Notes: OLS regression coefficients on a binary indicator for exposure to the Stop & Shop strikes. State fixed effects and robust standard errors included. 95% confidence intervals shown. Note that the outcomes are all standardized to have a mean of 0 and a standard deviation of 1 to ease comparisons across models.
For a more compelling causal identification strategy, I use an instrumental variables approach and leverage respondents’ physical distance from Stop & Shop stores. Survey respondents reported their ZIP codes, and I matched respondents with the geographic coordinates of the centroid of their ZIP code using the ZIP code distance database from the National Bureau of Economic Research.Footnote 4 I then calculated the geodetic distance in miles between each respondent and each of the 253 Stop & Shop locations in Massachusetts, Rhode Island, and Connecticut.Footnote 5 The key quantity of interest is distance to the nearest Stop & Shop store for each respondent, which I was able to calculate for 1,493 respondents; the average respondent was 3.8 miles away from a Stop & Shop store (minimum: 0.16, maximum: 39.1, standard deviation: 4.0).Footnote 6
The intuition is that respondents living closer to stores will be more likely to shop at those stores and therefore to have come into contact with the strikes, yet distance to Stop & Shop stores is unlikely to shape attitudes toward unions and labor action outside the magnified exposure to the April 2019 strikes. In addition, it is theoretically unlikely that people further away from Stop & Shop stores would be more likely than those closer to stores to report experiencing the Stop & Shop strike in their local community. Last, I assume (with empirical checks I describe below) that there are no other factors that explain both Stop & Shop store location and labor attitudes and actions. Figure 2 reviews the first-stage results for this instrument, showing that distance to the nearest Stop & Shop store is strongly related to respondents’ frequency of shopping at Stop & Shop (left panel) as well as exposure to the strikes in their local community (middle panel). As an additional check, I also consider how individuals learned about the strike in their local community, and look at the proportion of respondents who reported that they had encountered it firsthand or through their family, as opposed to through the media or other channels.Footnote 7 As the right panel in figure 2 shows, individuals who were closer to a Stop & Shop store reported that they were more likely to learn about the strikes in their local community by seeing the strikes firsthand or through a family member, as opposed to through other sources; 54% of respondents reported learning about the strikes in this way.Footnote 8
Figure 2 Distance to Nearest Stop & Shop, Frequency of Shopping at Stop & Shop, and Self-Reported Strike Exposure
Notes: Binned scatterplot with frequency of shopping at Stop & Shop (left), reports of strikes in local community (middle), or reports of encountering the strike firsthand or through family members (right) plotted against respondents’ nearest Stop & Shop store, measured in miles.
Additionally, as I verify in figure 3, distance to Stop & Shop stores is unrelated to other respondent characteristics that might plausibly shape union views or labor action (e.g., Kochan et al. Reference Kochan, Yang, Kimball and Kelly2019), including being a union member, knowing a union member as a friend or a family member, views on economic policy (averaging support for raising the minimum wage, a perception that inequality is too high, and support for government redistribution on a 1–5 scale),Footnote 9 political ideology, and partisanship, which provides indirect support for the independence of the instrument from unmeasured causes of labor attitudes. These results suggest that Stop & Shop stores were not more likely to be located in areas that were more predisposed to support labor or workers’ issues.Footnote 10
Figure 3 Distance to Nearest Stop & Shop and Predictors of Union Support
Notes: Coefficient plot from regression with distance to the nearest Stop & Shop in miles as the outcome, and key demographic characteristics that might predict labor support as predictors. Regression adjusts for state fixed effects, robust standard errors applied. 95% confidence intervals shown.
Last, as a placebo check I verify that it is distance to Stop & Shop stores—and not other retail or grocery stores—that leads to strike exposure by calculating the distance of each respondent to their nearest big-box retail store (Target or Walmart), as well as to other major grocery chains in the Northeast (Market Basket, Trader Joe’s, and Whole Foods).Footnote 11 In appendix C, I show that distance to these other stores does not predict either frequency of shopping at Stop & Shop or reporting a strike in one’s local community.
The Effects of the Stop & Shop Strikes on the Public
Tables 2 and 3 present the main results, documenting reduced form and 2SLS estimates of the effect of strike exposure on union attitudes and labor action, respectively. In the reduced form regressions, I show the relationship between the various outcomes and my exogenous variable, distance in miles between the respondent and their nearest Stop & Shop store. In the 2SLS regressions, I show the results of the second-stage regression between strike exposure and the outcomes, having used distance in miles to the nearest Stop & Shop store as the instrument for strike exposure to produce predicted values of the endogenous variable (i.e., strike exposure) that I use as an explanatory variable in this second stage (Angrist and Pischke Reference Angrist and Pischke2008, chap. 4).
Table 2 Reduced Form Estimates of Stop & Shop Strike Exposure
Notes: Outcomes standardized to have a mean of 0 and a standard deviation of 1. + p < 0.10; * p < 0.05; ** p < 0.01; *** p < 0.001. Robust standard errors. Estimates show reduced form regressions. Models 4 and 10 include only employed respondents; models 5 and 11 include only nonunionized workers in the labor force.
Table 3 2SLS Estimates of Stop & Shop Strike Exposure
Notes: Outcomes standardized to have a mean of 0 and a standard deviation of 1. + p < 0.10; * p < 0.05; ** p < 0.01; *** p < 0.001. Robust standard errors. Estimates show second-stage regression results from 2SLS regressions, using distance to the nearest Stop & Shop store as an instrument for whether a respondent reported strikes in their local community. Models 4 and 10 include only employed respondents; models 5 and 11 include only nonunionized workers in the labor force.
Causal interpretation of the 2SLS regressions should be read as the effect of strike exposure as caused by geographic proximity to a Stop & Shop store. Causal interpretation of these regressions depends on the assumptions that (1) distance to Stop & Shop stores only affects labor attitudes and interests through strike exposure; (2) there are no other unmeasured factors related to both respondents’ distance to Stop & Shop stores and labor attitudes and interests, net of the controls I include; (3) distance to Stop & Shop stores has a monotonic effect on strike exposure (that is, being closer to Stop & Shop stores leads to more strike exposure); and (4) labor attitudes and interests for one respondent do not depend on whether other respondents are exposed to strikes (Sovey and Green Reference Sovey and Green2011).
I include models with and without demographic controls, which include age (in years), race (in five categories), Hispanic ethnicity (binary), gender (binary plus other dummy), employment status (binary), education (in four categories), total family income (in quartiles), political ideology (in seven categories), partisanship (in seven categories), current union membership (binary), and having a close friend or family member in a union (binary; see appendix D for variable definitions). I include these controls in both stages of the 2SLS regressions. All models include state fixed effects. I use robust standard errors and I standardize the outcome variables to have a mean of 0 and a standard deviation of 1.
Table 2 shows that across the reduced form models, distance to respondents’ nearest Stop & Shop store is negatively correlated with support for the strikes and participation in the strikes, indicating that respondents who were geographically closer to Stop & Shop stores were more favorable to the strikes and more likely to report taking actions supporting the strikes. In models without controls, distance to Stop & Shop stores is also correlated with supporting Publix workers and expressing interest in workplace action, but these correlations weaken considerably with the inclusion of controls.
In table 3, I examine the 2SLS regressions that use distance to the nearest Stop & Shop store as an instrument for respondents reporting strikes in their local community. Mirroring findings in table 2, I find that strike exposure is correlated with greater support for the strikes and a greater likelihood of having personally taken actions supporting the strikes. The effect size of strike exposure, as instrumented by distance to the nearest Stop & Shop store, on support for the strikes is meaningful: in the models with controls, moving from no exposure to exposure to the strike is predicted to increase support by 0.81 units and increase the number of actions taken to support the strikes by 1.1 units. By comparison, the differences in support for the strike between Democrats and Republicans is about 0.64 points, and the same partisan difference in the likelihood of participating in the strikes is about 0.34 points.
Without controls, strike exposure is correlated with having clicked to read the Publix Coworker.org petition, but these effects are substantially reduced with controls. Across all models, I find no meaningful correlation between strike exposure and union vote (among nonmembers in the labor force) and support for a stronger labor movement overall; if anything, the coefficient on strike exposure for union voting is negatively signed, indicating that strike exposure may lower interest in unionization.
In appendix E, I explore the specific items in the “actions taken supporting the striking workers” index most correlated with strike exposure. I find that strike exposure is most strongly correlated with individuals reporting that they boycotted Stop & Shop or encouraged others to boycott Stop & Shop, as well as with bringing food to striking workers. I do not find any correlation with participating in the protests by holding signs or marching, or by donating money to the striking workers.
While I do not find any consistent relationship between exposure to strikes and the aggregate workplace action index, in appendix F I explore the individual items within the index.Footnote 12 I find that strike exposure, as instrumented by distance to the nearest Stop & Shop store, was strongly related to interest in one workplace action: getting advice from coworkers. The other workplace actions were not consistently related to strike exposure. This provides suggestive evidence that strikes may increase interest in engaging with coworkers—but not necessarily through formal labor organizations.
Reconciling Differences across Strikes: A Suggestive Role for Linked Fate
Why did the Stop & Shop strikes fail to inspire personal interest in, and support for, labor unions, labor action, and the labor movement while earlier large-scale teachers’ strikes did? One explanation offered by earlier research on the teachers’ strikes was that those strikes may have increased perceptions of linked fate between teachers and parents—and created a sense of group identity that helped to inspire greater interest in the labor tactics deployed by the teachers by showing how teachers were advancing issues key to parents (Hertel-Fernandez, Naidu, and Reich Reference Hertel-Fernandez, Naidu and Reich2021; on the linked fate concept see Dawson Reference Dawson1994; see also Gay, Hochschild, and White Reference Gay, Hochschild and White2016).
The Stop & Shop survey included a survey item probing about linked fate between the striking workers and members of the public, asking: “How much do you think that the strikes at Stop & Shop will affect what happens in your life?” The survey provided three options: “very strongly affect,” “somewhat affect,” and “not at all affect.” Most respondents (60%) said that the strikes would not affect their lives at all, 32% said they would somewhat affect their lives, and 7% reported that the strikes would very strongly affect their lives.
Figure 4 shows the results from four separate OLS regressions, examining the correlation between respondents’ reports of linked fate with the strikes (coded 1–3, with higher values indicating a stronger sense of linked fate with the strikes), and the four outcomes involving interest in, and support for, unions and labor action. Across all regressions, respondents who reported a stronger degree of linked fate between themselves and the striking workers were more interested in supporting other grocery store workers (by signing the online petition in support of the Publix workers), expressing interest in workplace action (if employed), voting for a union at their job (if in the labor force and currently not in a union), and supporting more union influence in society in general. Note that these regressions standardize outcomes to have a mean of 0 and a standard deviation of 1 to ease comparisons, and that these regressions include the demographic controls described above as well as state fixed effects, and use robust standard errors.
Figure 4 Perceptions of Linked Fate and Interest in Broader Labor Action and Support
Notes: Outcomes standardized to have a mean of 0 and a standard deviation of 1. Robust standard errors; 95% confidence intervals shown. Figure plots OLS coefficients on linked fate variable (1–3). Demographic controls and state fixed effects included. “Interest in workplace action” includes only employed respondents; and “vote for union” includes only nonunionized workers in the labor force.
Figure 4 suggests that a sense of linked fate between members of the public and the striking workers may play a moderating role between exposure to the strikes and individuals’ broader interest in unions and labor action. Did firsthand exposure to the strikes change these perceptions of linked fate? Using the same reduced form and 2SLS approaches described above, I study linked fate as an outcome for strike exposure as instrumented by distance from the nearest Stop & Shop store, and report the results in table 4. I find no statistically or substantively significant effect of firsthand strike exposure on respondents’ perceptions of linked fate with the strikers, providing suggestive evidence that one reason why the Stop & Shop strikes may have failed to engender broader labor support—in contrast to the teachers’ strikes—is the lack of a sense of linked fate the strikers created with members of their local communities.
Table 4 Exploring Reduced Form and 2SLS Effects of Strikes on Perceptions of Linked Fate
Notes: + p < 0.10; * p < 0.05; ** p < 0.01; *** p < 0.001. Robust standard errors. Estimates show reduced form models (top) and second-stage regression results from 2SLS regressions (bottom), using distance to the nearest Stop & Shop store as an instrument for whether a respondent reported strikes in their local community.
The Uneven Effects of Mass Labor Strikes and the Future of the US Labor Movement: A Research Agenda
In the years leading up to the COVID-19 pandemic, the United States saw a resurgence of labor activism, including a number of large-scale labor mobilizations in the public sector with a wave of national teacher strikes (Dernbach Reference Dernbach2020). Surprisingly, that activism did not ebb significantly during the pandemic. Instead, many frontline essential workers engaged in labor actions to demand better health and safety protections given the risks they faced while continuing to work during the pandemic (Abrams Reference Abrams2021). And, as the economy rebounded and the virus subsided, many workers took advantage of a tight labor market to organize and go on strike, including significant, long-lasting labor actions by automobile manufacturing workers, actors, and Hollywood writers, among others (Escobar and Zhang Reference Escobar and Zhang2023). The critical question arising from this resurgence of labor activism since 2018 is whether it will translate into higher levels of union membership, reversing the decades-long decline in membership since mid-century and the corresponding decline in labor’s economic and political clout in the American political economy (Hacker et al. Reference Hacker, Hertel-Fernandez, Pierson, Thelen, Hacker, Hertel-Fernandez, Pierson and Thelen2021; Hacker and Pierson Reference Hacker and Pierson2010; Lichtenstein Reference Lichtenstein2002; Rosenfeld Reference Rosenfeld2014).
This paper has examined whether, by exposing more members of the public to strikes, large-scale labor action can affect mass attitudes about and support for the striking workers, interest in labor action, and support for the labor movement as a whole. Replicating previous work on large public school teacher strikes held in 2018, this study finds that firsthand exposure to private sector strikes increased support for the striking workers and actions taken to support striking workers (for instance, through boycotts of striking stores). But departing from the past study of teacher strikes, this study finds that the grocery store strike did not necessarily inspire interest in other forms of collective labor action—with the possible exception of inspiring greater interest in working together with one’s coworkers. Moreover, exposure to the grocery store strikes in this study did not increase support for the labor movement as a whole.
One implication of this finding is that individual workers are drawing a distinction between specific labor actions and unions and the labor movement in general—not necessarily extrapolating experiences with the former and applying them to the latter. This raises a potential limitation of strikes for educating the public about the labor movement as a whole—and, read together with past research on the teachers’ strikes, suggests that the nature of the union and the striking workers may matter in the broader lessons that workers take away from a labor action. While only suggestive, the analysis presented in this paper points to a potential role for perceptions of linked fate between striking workers and members of the public, and points out that strikes that fail to engender such connections and shared identities may fail to translate into greater interest in the labor movement and labor action.
A future research agenda on the dynamics of the recent wave of mass strikes and its impacts on the public ought to further explore these issues, and could be part of a better understanding of the role of labor in the American political economy (Hacker et al. Reference Hacker, Hertel-Fernandez, Pierson, Thelen, Hacker, Hertel-Fernandez, Pierson and Thelen2021). Below, I sketch out six lines of much-needed additional inquiry:
-
• How do the occupational, class, and demographic identities of workers and their demands matter for the impact of strikes on the mass public? In particular, are there certain occupations that—because of the nature of their work or the interactions they entail with customers, clients, or other members of the public—are more likely to engender support and interest in further labor action? And do the demographics of these occupations matter, especially given the identities of potentially affected members of the mass public, as well as the occupational stereotypes or class identifications that members of the mass public may hold? For example, are strikes more likely to be persuasive to members of the public when they occur in occupations that are closer to the work that members of the public do, or among workers who share a greater degree of demographic identity? This is especially important given that there have been different waves of labor organizing and collective action in recent years that span occupational and class identities (Kochan et al. Reference Kochan, Fine, Bronfenbrenner, Naidu, Barnes, Diaz-Linhart and Kallas2022), including a wave of labor actions involving low-paid service sector workers (such as the recent Starbucks organizing drive and the earlier “Fight for $15”) alongside mobilizations among more white-collar, professional occupations (such as teachers, actors, Hollywood writers, media workers, and tech sector workers).
-
• When do strikes backfire in terms of public opinion? Research on the consequences of strikes for production and service delivery suggests real economic costs to labor actions—with healthcare strikes negatively affecting patient care, manufacturing strikes negatively affecting product quality, and teachers’ strikes negatively affecting student performance (e.g., Baker Reference Baker2013; Gruber and Kleiner Reference Gruber and Kleiner2012; Krueger and Mas Reference Krueger and Mas2004; Mas Reference Mas2008). At what point do these costs to consumers and clients become too high and result in backlash against the striking workers, their unions, and their supporters, with members of the public becoming less supportive of the strikes and workers on strike? And are there groups of the public that are more likely to be sensitive to these costs than others in terms of their support for workers and strikes?
-
• How can actions by political leaders shape public perceptions of strikes and their impacts? During many of the larger strikes in recent years, politicians—including President Joe Biden—delivered high-profile gestures of support, including making speeches, walking picket lines, and calling on employers to bargain fairly with unions (Klein and Carvajal Reference Klein and Carvajal2023). In an increasingly polarized era, how are these gestures of support perceived by striking workers themselves and by members of the public? And do these statements help or hinder unions’ bargaining strategies?
-
• Are there downstream political consequences of strikes, and if so, under what conditions and through what channels? Anecdotally, union leaders describe how the activation, mobilization, and networking of workers that occurs during strikes can feed later political activity—especially if union members connect their experiences during the strike to elected officials. For instance, after the wave of Red4Ed teacher strikes in 2018, many teachers were inspired to volunteer for political campaigns and even run for office for the first time. Many of these teacher candidates won, and even in the races where they lost they succeeded in mobilizing fellow educators into politics and ensuring that educational issues were part of campaigns (Blanc Reference Blanc2019; Reilly Reference Reilly2018; see also Lyon and Kraft Reference Lyon and Kraft2024). Indeed, research suggests that educational spending tends to increase following teacher strikes (Lyon and Kraft Reference Lyon and Kraft2024). But strikes could also be demobilizing for workers and unions, especially if workers lose or inspire significant backlash from powerful opponents. More systematic research is needed to understand when and how strikes translate into downstream political activity—including volunteering; organizational building; and inspiring workers, especially blue-collar workers, to run for office themselves (Carnes Reference Carnes2013; Reference Carnes2018).
-
• How do strikes affect new union organizing? This paper and earlier work examine individual-level views and interest in union organizing, but more work is needed, especially over a longer historical period, to examine the impact of strikes on new union organizing drives, tracing whether they spill over into new occupations, regions, and sectors, and whether those drives translate into recognized unions with first contracts.
-
• How do strikes change internal union organizations? Last, journalists and labor leaders anecdotally note that recent strikes in the automobile industry have helped to revitalize previously stagnant union organizations (such as the United Auto Workers), bringing in new leaders with a stronger connection to the grassroots rank-and-file and building interest in bold new organizing strategies (Lichtenstein Reference Lichtenstein2023). Older historical research confirms that strikes can indeed be transformational moments for unions (Voss and Sherman Reference Voss and Sherman2000), as does qualitative work studying individual strikes, such as the 2012 Chicago Teachers Union strike (Ashby and Bruno Reference Ashby and Bruno2016). But more work is needed to understand how the recent strike wave is—and is not—transforming the internal organization and leadership of unions, with attendant consequences for members’ engagement in unions, internal democracy, and choice of strategies, including new organizing.
More broadly, the upsurge in strikes and other labor actions since 2018 creates important new opportunities for studying how workers think about workplace democracy (Mazumder and Yan Reference Mazumder and Yan2024). As the esteemed political theorist Robert Dahl (Reference Dahl1977, 8) warned decades ago, “[B]ecause the internal government of the corporation [is] not itself democratic but hierarchical and often despotic, the rapid expansion of this revolutionary form of economic enterprise meant that an increasing proportion of the demos would live out their working lives, and most of their daily existence, not within a democratic system, but instead within a hierarchical structure of subordination.” Strikes can be one important way that workers develop and express their rights to economic democracy in the workplace against such subordination and despotic control (see especially Gourevitch Reference Gourevitch2018). As a field, there continues to be much more Political Science could do to study the internal organization of the workplace, when and how workers think about potential labor action, and the mobilizing role of labor unions and other worker organizations in an era of labor resurgence (Frymer Reference Frymer2010; Hertel-Fernandez Reference Hertel-Fernandez2020b).
Supplementary material
To view supplementary material for this article, please visit http://doi.org/10.1017/S1537592724000902.
Data Replication
Data replication sets (Hertel-Fernandez Reference Hertel-Fernandez2024) are available in Harvard Dataverse at: https://doi.org/10.7910/DVN/ZBSXGO.
Acknowledgments
This study received human-subjects research approval from the author’s institutional review board.
Open access
