In 1981, President Reagan’s decision to fire striking air traffic controllers seemed to mark the death knell for labor actions in the United States. Union membership has plummeted by over 60% since 1970, and worker participation in strikes has seen an even more dramatic 90% drop. For decades, strikes have seemed to be like vinyl records—relics of the past. Perhaps they evoke a sense of nostalgia for a bygone era, romanticized and somewhat symbolic for those who remember them fondly, but they have been perceived as ineffective tools in the modern landscape.
Yet, like the vinyl record, recent years have seen a surprising resurgence in teacher strikes. From Wisconsin in 2011 to the nationwide “RedForEd” movement in 2018, educators have taken to the picket lines with renewed vigor, seeming to garner public support and win significant concessions. This unexpected revival prompted my colleagues Matthew Kraft (Brown University), Matthew Steinberg (Accelerate), and me to investigate the prevalence, causes, and impacts of teacher strikes in the 21st century.
Uncovering data
Our first challenge was a lack of comprehensive data. Since 1982, the Bureau of Labor Statistics has only collected data on strikes involving 1,000 or more employees. This means they are missing a lot of teacher strikes, as 97% of school districts employ fewer than 1,000 teachers. To fill this gap, we created a novel dataset by systematically searching for and reviewing approximately 90,000 news articles, a process that we validated with administrative data from two states. This painstaking process allowed us to document 772 U.S. teacher strikes from 2007 to 2023, providing the most comprehensive dataset on teacher strikes in recent history.
The landscape of modern teacher strikes
Our research reveals a complex landscape of teacher strikes across the United States. Figure 1 visualizes the distribution of strikes geographically. We count any public school district that had to cancel at least one day of school as a separate striking district. This means we include legal, illegal, coordinated, and individual strikes, as well as “wildcat” strikes, “walk-outs,” or “sick-outs” that lead to school closures. When strikes are coordinated across districts, we conceptualize those as multiple strikes based on the number of districts cancelling classes.
The impact of these actions has been substantial, affecting roughly 11.5 million students and leading to the cancellation of 3,403 school days (or 48 million student days) over 16 years. Nationally, the median number of strikes per year is 12.5, resulting in an average of 89 days of canceled school annually. Most strikes are brief, with 65% ending in five days or less and a median duration of two days. The longest strike lasted 34.5 days.
Surprisingly, 75% of the strikes we identified occurred in states where teacher strikes are illegal. This reflects the old labor maxim, “there is no such thing as an illegal strike, only an unsuccessful one.”
Why do strikes occur?
Understanding why teachers strike is crucial to grasping the broader implications of these actions. From our search process, we tracked the reported reasons for strikes from 2007 to 2023. Teacher compensation emerged as the primary driver, cited in 89% of strikes. This encompasses both salaries and benefits, reflecting ongoing concerns about educator pay across the country. Working conditions were another significant factor, mentioned in 59% of strikes. This category includes issues such as class sizes and general school expenditures (48%), non-instructional staff (16%), and labor rights (10%). Interestingly, 10% of strikes, particularly those occurring in 2018 or later, included “common good” demands, such as addressing housing or immigration issues. This trend suggests a broadening of teacher unions’ advocacy beyond traditional workplace concerns, positioning educators as advocates for wider community issues.
The political and economic impacts of teacher strikes
With teachers, and likely other public sector workers, strikes are more than just a negotiation tool in a dual-actor collective bargaining process. In the first peer-reviewed paper to come from this research, Matthew Kraft and I conceptualize teacher strikes as something fundamentally different from private sector strikes. In addition to their use in collective bargaining negotiations, strikes also serve as public signals in a contested policy arena, communicating to both the public and political leaders the need for educational change.
To test this idea, we look at both the political and economic effects of teacher strikes. Using a “difference-in-differences” approach that accounts for the non-random timing and location of strikes, we estimate the causal effect of teacher strikes by comparing trends in striking districts to what we would have expected based on trends in non-striking districts.
What we find is striking (pardon the pun): Teacher strikes lead to a 130% increase in education-focused political advertising, resulting in education ads representing 11% of all campaign ads in the area where the strike occurred. Effects are largest in the period closest to an election, when ads matter the most. Effects are also larger for the first strike in a given area—subsequent strikes in the same district seem to provide weaker signals to political leaders.
Further, in a separate working paper with collaborators Leslie Finger (University of North Texas) and Hyesang Noh (University at Albany), we find that strikes don’t just affect campaigns, they reverberate into election outcomes. Teacher strikes reduce voter turnout among Republicans and decrease the Republican vote share in subsequent general elections in surrounding areas.
These political effects also translate into tangible economic outcomes. In a new NBER working paper, we show that teacher strikes have real impacts on educators and classrooms. As illustrated in Figure 2, strikes increase annual teacher compensation by about $10,000 (inflation-adjusted to 2018 dollars, or roughly eight percent of the average teacher’s earnings), an effect that is sustained over time. Working conditions see improvements too, with class sizes decreasing by half a student on average, and expenditures on non-instructional staff compensation increasing by about seven percent. Importantly, they do this via new revenues (e.g. via increased taxes) and expenditures, not reallocations of existing funds. Strikes lead to ~$1,000 per pupil increases, increases that are sustained at least eight years post-strike.
We also look at how strikes affect student achievement. Notably, research on teacher strikes in other countries has found negative effects, but that research focuses on strikes that lasted much longer than the typical U.S. strike. We find that U.S. strikes do not affect student achievement in the five years following a strike, though long-term effects remain uncertain.
Shorter vs. longer strikes
There are some notable differences between shorter and longer teacher strikes. Prior literature from outside the U.S. has found that very long strikes negatively affect students, with research from Canada finding that strikes lasting 10 days or longer lead to small, negative effects on student test scores. Nine out of 10 strikes in the U.S. last fewer than 10 days.
When we examine differences in effects between these longer strikes and strikes that last fewer than 10 days, we find that they lead to similar increases in teacher compensation and working conditions. However, longer strikes cause small (three percent to five percent of a standard deviation) declines in student achievement in the year of the strike and the subsequent year. Two years after the strike, the negative effects dissipate. These results are consistent with prior research and align with realities on the ground where strikes of less than two weeks might reasonably be made up at the end of the year or during school breaks.
The broader political effects also vary by strike length in important ways. We find that shorter strikes (five days or less) are the most efficacious at capturing the attention of political leaders. In contrast, strikes lasting more than five days have the largest negative effects on Republican turnout and vote share.
Taken together, this suggests that both short and long strikes are effective at achieving wins for teachers. The short strikes seem to accomplish this by providing powerful signals to political leaders, whereas the longer strikes accomplish this in a more traditional industrial relations process of holding out until management concedes (potentially at the expense of students).
Implications for education, labor, and upcoming elections
For teachers and their unions, this work comes at a critical moment for the teaching profession, which is facing historically low levels of prestige, interest, preparation, and satisfaction. Strikes and other labor actions are also on the rise, partly driven by teachers. Our research emphasizes the importance of the credibility of the signaling mechanism of strikes. Teacher strikes have tended to happen only when the situation is dire, making them a trustworthy signal of serious problems. Organizing a strike, like producing and listening to vinyl records, requires more time and effort compared to quicker, more convenient alternatives. This is part of what makes them both rare and powerful.
The implications of this body of research extend beyond education policy. For the labor movement as a whole, this research emphasizes that collective action remains a potent force for change, even in an era of union decline. Specifically, our work highlights the effectiveness of coordinating actions across multiple workplaces or regions; for example, coordinated teacher strikes took place in more conservative areas where strikes were more likely to be illegal but nevertheless were able to make similar gains as individual district strikes. We also emphasize the importance of leveraging public opinion and political pressure alongside traditional bargaining tactics. Importantly, teachers have achieved wins through relatively short strikes that minimize harm to students. Workers may benefit from similar strategies, especially if they frame demands in terms that resonate with other workers and the wider community.
As we make sense of the recent elections, this research can also help consider the impacts of recent protests over wages, the use of automation and AI, and international conflicts. Some may not think of strikes as protests, but like all forms of protest, strikes are “a mode of political action oriented toward objection to one or more policies or conditions, characterized by showmanship or display of an unconventional nature, and undertaken to obtain rewards from political or economic systems while working within the systems.” This research, together with previous studies, suggest protest effects on election outcomes likely depend on: the level of disruption, presence of violence, duration of the action, and crucially, its location. In a tight race, the impact of protests in battleground areas (like North Carolina or Pennsylvania) could tip the electoral balance.
Just as vinyl records have again become relevant, strikes have reemerged in the spotlight. Both possess a unique resonance that modern technological alternatives struggle to replicate. However, the delicate balance is key—overplay a record, and it loses its magic; overuse strikes, and they risk becoming background static. In the end, the trick will be figuring out how workers can judiciously use strikes as a tool to resist oppression in an unequal American political economy.
Related Content
Author
