Union-endorsed candidates win over 60 percent of school board races and are more likely to support teachers’ unions’ favored policies. During the COVID-19 pandemic, areas with stronger unions kept schools closed longer, in alignment with union policy positions. At the state-level, fewer policy reforms that run contrary to teachers’ union interests are enacted in areas where teachers’ unions spend more on candidate donations than other interest groups.
Expenditures in newly unionized school districts increase by 12.3 percent, on average. When CBAs marginally increase in strength by one standard deviation, school districts spend 1.2 percent more, which equates to about 111 dollars more per pupil. Alternatively, laws like Right-to-Work and restrictions on collective bargaining led to a reduction in school expenditures of around 3.8 percent.
Teacher salaries increase by between 4 percent and 12 percent with the onset of unionization or collective bargaining and decrease by between 3 percent and 9 percent when teachers’ unions and collective bargaining are substantially restricted by state laws. Strong teachers’ unions are also successful at directing new revenue for schools towards teacher salaries. In states with strong teachers’ unions, teacher salaries increase by 84 cents for every 1 dollar increase in state revenue to school districts, compared to just 32 cents per every dollar in new state revenue in states with average teachers’ union power.
For every 10 percent increase in union strength, experienced and novice teachers are 4 to 7 percent less likely to turnover. Laws drastically reducing teacher collective bargaining rights increased teacher turnover by about 1 percent per year, and increased retirements by 18 percent in the year after reforms were implemented.
Marginal increases in the strength of CBAs have no impact on student achievement. When teachers’ unions are substantially weakened by state policies, the research suggests short-run student achievement declines or does not change. Right-to-Work policies that reduced unions’ abilities to collect fees for services resulted in no change in student outcomes. In the case of Wisconsin Act 10’s reduction of the scope of collective bargaining to just wage increases at or below the rate of inflation, short-term achievement decreased by 0.20 standard deviations (SD), primarily at low-performing schools, and potentially due to the increase in teacher turnover. However, in the long-run, school districts experienced an influx of higher quality teachers due to more flexible teacher pay arrangements which was associated with increased math and reading performance.
Teachers’ unions in the United States are membership-based, non-profit organizations largely funded by dues-paying members, namely classroom teachers, but also other education support professionals, higher education faculty and staff, and retired educators. The two largest teachers' unions, the National Education Association (NEA) and the American Federation of Teachers (AFT), maintain extensive national organizations. In 2023, the NEA had over 2.8 million members and the AFT had over 1.7 million members.1 While the National Labor Relations Act (NLRA) (1935) governs the formation of private sector trade unions, teachers’ unions operate in the public sector and thus the NEA’s and AFT’s state and local affiliates are subject to state laws. The diversity of state laws governing the formation and operation of teachers' unions means that the size, scope, and activity of state and local teachers’ unions vary geographically.
The primary function of local teachers’ unions has been the negotiation of collective bargaining agreements (CBAs) over wages and working conditions on behalf of their membership. Initially embraced by the AFT, the earliest CBA for public school teachers was inked by the AFT-affiliate, the United Federation of Teachers in New York City in 1962, following a one-day strike of over 20,000 educators.2 Today, state laws allow collective bargaining between teachers and their employers through unions in 44 states and prohibit it in 6 states.3
Even as teachers’ unions still negotiate CBAs in most states, state laws governing the formation of and operation of teachers’ unions have changed in a number of places over the last 15 years.4 Act 10 (2011) signed into law by Wisconsin governor Scott Walker made headlines for reducing teachers’ (and other public sector union members’) collective bargaining rights to just wage increases at or below inflation, and similar laws followed in Ohio, Idaho, Indiana, and Tennessee. Also states like Wisconsin, Michigan, Indiana, West Virginia, Kentucky, and Missouri removed the ability of teachers’ unions to charge non-union members for collective bargaining representation (i.e., Right-to-Work) prior to the supreme court making this the law of the land in the landmark Janus v. AFSCME (2018) ruling. Virginia (2021) stands as the lone state in recent years to actually return teachers’ union collective bargaining rights after prohibiting them for nearly 50 years.
The key arguments for teachers’ unions and collective bargaining, sometimes called the voice perspective, primarily center on providing teachers with a voice in their wages and working conditions. Hirschman’s (1970) exit-voice hypothesis argues that workers have two options when it comes to their wages and working conditions. First, they can exit and reward a new employer with their expertise and labor, thereby signaling to their old employer that wages and working conditions were insufficient. The second response, voice, involves the use of collective pressure to bring the actual wages and working conditions closer to the desired outcomes. In the case of education, some argue that absent an organized voice, teachers would be subject to the arbitrary and capricious exercise of power by school administrators, which could result in unfair and/or inequitable pay and benefits, larger class sizes, heavier workloads, less safe schools, and, ultimately, higher teacher turnover. Further, the working conditions of teachers are often cited as the learning conditions of students, and teachers, through their day-to-day work in America’s classrooms, are best positioned to advocate on behalf of students through their unions.
The key arguments against teachers’ unions and collective bargaining, sometimes called the rent-seeking perspective, center on the routinization and standardization of teaching and schools through the labyrinth of rules found in CBAs. These rules could bind schools to practices that are less innovative and less effective. Further, teachers’ unions may drive up educational inputs (i.e., teacher salaries) to reward their membership with less regard for the productivity of these inputs in yielding positive outcomes for students (such as student achievement and graduation rates), thus resulting in inefficient practices.5 For example, teachers’ unions may advocate for job protections like lower class sizes, longer planning periods, and shorter work days than would otherwise be optimal or necessary to deliver a high-quality education to students. Furthermore, to deliver these working conditions, school districts may have to spend more money and hire more teachers than would be ideal. The difference between the optimal allocation of resources and the allocation achieved due to the advocacy of teachers’ unions represents a rent for unions and an inefficiency in the school system.6
The tension between the two dominant conceptual perspectives on teachers’ unions, rent-seeking and voice, has been frequently discussed in the research literature.7 These perspectives have their advantages. First, they allow a clear framework for understanding what teachers’ unions do within schools to affect education inputs, like expenditures, and education outputs, like student performance. Second, they help forecast what teachers’ unions effects might be with new policy changes, like changes to the scope of collective bargaining, changes to school finance, and changes to school choice. Third, they help highlight the context-dependent nature of teachers’ unions—that in some contexts they may be rent-seeking, to the detriment of productivity, and in other contexts they may amplify voice, to productivity’s enhancement. In what follows, I overview the extent to which the research literature aligns with these theoretical perspectives, emphasizing what we currently know and do not know about the impact of teachers’ unions and collective bargaining in public education.
Prior to overviewing the research evidence in more detail, I first discuss how researchers frequently measure the impact of teachers’ unions and collective bargaining in education and why the method of measurement matters. I then describe the research on teachers’ unions’ impact on inputs into the education system like school expenditures, teacher working conditions, the teacher labor market, and education policy. I continue with a discussion of the impact of teachers’ unions on education system outputs like student achievement. I conclude with a discussion of what we still need to know about teachers’ unions and collective bargaining in education.
Measuring Teachers’ Unions and Collective Bargaining in Research
One of the challenges with obtaining valid estimates of the impact of teachers’ unions and collective bargaining is that unions and bargaining negotiations are not randomly assigned to school districts. Instead, teachers’ unions tend to organize and negotiate in locations where working conditions are more challenging and where student achievement is lower (because these are the locations where teachers may believe unions add the most value).8 Consequently, simple comparisons between unionized and non-unionized districts likely reflect these demographic or contextual differences rather than unionization or union strength. For instance, simple analyses that compare unionized to non-unionized districts often find that students in unionized school districts perform worse than students in non-unionized districts. The difference in student performance may have little to do with actual union activity but may simply be spuriously related to the conditions that make unionization more or less likely.
Researchers have used various approaches to overcome this challenge, including:
Timing of Union Certification. The first approach is to add a time dimension to the analysis. Although whether a school district is unionized is correlated with school district characteristics, this approach assumes the timing of unionization is less or not at all correlated with key outcome variables.9 One challenge with this approach is that most unions in the United States were certified decades ago, limiting the ability to explore unionization’s impact on more recent outcomes. However, some researchers have begun to examine charter school unionization, as the certification of charter school teachers’ unions has been more recent.10
Exploiting Policy Shocks to Union Strength. The second approach is to use the timing of external policy shocks that amplify or diminish the power of teachers’ unions. Under this approach, individual policy changes may be correlated with union strength and key outcomes, but the timing of policy changes are plausibly random. For example, researchers have explored how important outcomes, like student achievement, change when teachers’ union power was substantially diminished by the enactment of Right-to-Work policies, which prohibited teachers’ unions from collecting fees as a condition of collective bargaining representation.11 Researchers have also used the timing of collective bargaining negotiations relative to school finance reforms or tax referendums to explore questions about union power and school district expenditures.12 Furthermore, studies have examined how student outcomes and teacher labor markets change in states that have greatly reduced collective bargaining rights for teachers.13
Using Marginal Increases/Decreases in Union Strength. The third approach is to examine how outcomes change relative to marginal increases in teachers’ union strength. This method is particularly useful for analyses within states where there is little variation in the presence or absence of unions and collective bargaining (i.e., most school districts have teachers’ unions and collective bargaining). Teachers’ union strength has been measured in several ways. Stronger unions negotiate better contracts for their members, so one way to measure union strength is to assess the strength of the collective bargaining agreement (CBA).14 The biggest development in capturing CBA strength is Strunk and Reardon’s (2010) Partial Independence Item Response (PIIR) model for estimating latent CBA restrictiveness.15 Using information from hand-coding CBAs for over 600 different provisions in California, the researchers generated a valid and reliable measure of CBA restrictiveness (or the degree to which a given CBA limits school and school district leaders in administering schools). This approach has been used in other states, on charter school CBAs, and on COVID-19 pandemic MOUs with teachers’ unions.16 Beyond CBAs, marginal changes in teachers’ union strength have been captured in other ways. Teachers’ unions spend considerable amounts of money on federal, state, and local election campaigns, so one way to capture their strength is to look at the proportion of campaign donations in a given jurisdiction from teachers’ unions. 17 Another way is to use information from local or state teachers’ unions’ tax forms to capture the amount of teachers’ union revenue and expenditures per teacher, and the size of teachers’ union membership relative to the number of teachers in a jurisdiction.18 Some researchers have used a combination of all these measures, along with surveys of policymakers, to create an index of teachers’ union strength.19
Teachers’ Unions, Elections, and Policymaking
While discussions of teachers’ unions often focus on their impact on students and teachers, most of these impacts are because of their influence on education policy at the local, state, and federal levels. Scholars have noted that local and state control over education policy favors teachers’ unions.20 School board elections typically have low voter turnout, which allows highly organized local unions the opportunity to shape who governs school districts and who ratifies their CBAs.21 At the state level, there are few education-focused interest groups as organized as teachers’ unions. State-level teachers’ unions lobby in an effort to encourage the enactment of policies they believe will make schools better (and also grow union power and influence) and discourage the enactment of policies they believe will make schools worse (and also weaken union power and influence).22
At the local level, school boards made up of teachers’ unions’ favored candidates and former educators are more likely to adopt union-friendly CBAs.23 Thus, it is not terribly surprising that teachers’ unions exert considerable effort to influence school board elections, and, as a result of that support, union-endorsed candidates do very well. From 1995 to 2020, union-endorsed candidates in California won 71 percent of their races. In Florida, 63 percent of union-endorsed candidates between 2010 and 2020 won their races. Union endorsement significantly increases the likelihood of a school board member winning an election, even after controlling for candidate quality.24
The COVID-19 pandemic provided an opportunity to test how influential teachers’ unions were in local education policymaking. Initial decisions to close schools for in-person learning during the spring 2020 semester were largely left to state governors.25 However, decisions to keep schools closed for in-person learning during fall 2020 and subsequent semesters were largely left to local school boards. School districts with stronger unions were much less likely to open for in-person instruction during the fall 2020 semester.26
A number of studies explore how teachers’ unions are shaped by state education policies. For example, the growth in laws during the 1970s and 1980s that gave teachers’ unions the right to bargain on behalf of teachers drastically increased teachers’ union membership.27 By contrast, relatively recent enactments of Right-to-Work laws, which bar the requirement that teachers join unions as a condition of employment or pay a fair share fee, decreased teachers’ union membership by about 24,000 members, or about 40 percent of the average membership.28
Relatively few studies focus on how teachers’ unions shape state education policies. One study examined proposed and enacted teacher policies across five legislative cycles, coding them based on whether they expanded or weakened union rights. The researchers found that fewer favorable laws were proposed in areas where teachers’ unions spent more on elections and lobbying compared to school choice advocacy groups or business groups.29 Another study found similar results when examining specific education policies such as those governing public school choice, teacher evaluation, and teacher pay. States with a higher proportion of campaign contributions from teachers’ unions were less likely to adopt reform-oriented policies in these areas.30
Even fewer studies have explored how teachers’ unions shape federal politics. The limited research available suggests that the largest national teachers’ union, the NEA, spends millions of dollars on federal elections, and their supported candidates win more often than they lose.31 While the flow of NEA campaign contributions to Democratic candidates in federal elections is significant, there is evidence that the NEA also has several allied Republican lawmakers, particularly in states that lean Democratic. During a six-year span (2010 to 2016), the NEA took an official position on 148 pieces of federal legislation, and final legislative votes aligned with their position 79 percent of the time.
Teachers’ Unions and Expenditures on Education
One argument in favor of teachers’ unions and collective bargaining is that they demand more resources for schools and students and direct expenditures in productive ways based on the insights of their membership. However, opponents of teachers’ unions argue that teachers’ unions unnecessarily drive up schooling costs and direct funding in ways that primarily benefit union membership and not students.
According to the best evidence, it is generally the case that school districts with the strongest unions and the most comprehensive CBAs spend more per student and spend differently than school districts with weaker or non-existent unions or no collective bargaining rights. For example, when school districts unionize following the passage of laws granting the right to teacher collective bargaining, their expenditures increase by 12.3 percent.32 School districts also spend more as the restrictiveness or strength of teacher CBAs increases. A one standard deviation increase in CBA strength is associated with a 1.2 percent increase in school district expenditures, which equates to about 111 dollars more per pupil for the typical school district.33 Further, the spending increases associated with CBA strength are not spread evenly across categories in school district budgets. Most of the increases associated with CBA strength are in expenditures on instruction (about 73 dollars more per pupil for each standard deviation increase in CBA strength) and instructional support (about 20 dollars more per pupil for each standard deviation increase in CBA strength).
When school districts receive new revenue, expenditures on education go up more in areas with stronger teachers’ unions compared to areas with weaker teachers’ unions.34 When new state school revenue comes in, school districts may reduce local tax effort (sometimes called crowd-out or supplanting) or may grow their strategic budget reserves by not spending the funds. In terms of crowd-out, prior research finds that districts in states with stronger teachers’ unions maintain their local tax effort to a far greater degree than school districts in states with weak teachers’ unions when receiving new state revenue. In terms of spending, prior research also finds that the average school district spent 50 cents of every dollar of new state revenue. However, school districts in states with teachers’ unions that were one standard deviation stronger spent 69 cents for every dollar increase in new state revenue. Additional research finds that when teachers’ unions and school districts are negotiating new CBAs right after the approval of an increase in new local tax revenue for schools, they spend 50 dollars more per pupil on teacher benefits, and they reduce their reserve funds by an average of 6.8 percentage points compared to school districts that are not in bargaining negotiations. 35
Finally, school expenditures decrease when teachers’ unions are weakened by new legislation. Following the enactment of Right-to-Work laws, expenditures within school districts decreased by 800 dollars per pupil in the decade following the law’s enactment.36 Further, when states enacted laws like Wisconsin Act 10 that diminished the scope of collective bargaining over teacher salaries and benefits, school districts’ expenditures decreased by 3.8 percent.37
Teachers’ Unions, Teacher Salaries, and Teacher Working Conditions
The enhancement of working conditions is frequently cited as the primary advantage of unionization in education.38 According to this perspective, teachers’ unions emerged in public education to address the deficiencies of the centralized, bureaucratic, and management-dominated school systems of the early 20th century. During that period, teachers individually faced inadequate compensation and were vulnerable to managerial and administrative abuses. Opponents, again, argue that teachers’ unions drive up spending on working conditions to inefficient and unproductive levels, leading to higher costs for school districts and lower student achievement.
The preponderance of evidence suggests a positive relationship between teachers’ unions and teacher salaries. One study shows that teacher salaries increase by between 4 percent and 12 percent with the onset of unionization or collective bargaining39 and decrease by between 3 percent and 9 percent when teachers’ unions and collective bargaining are substantially reduced.40 In school districts with strong teachers’ unions, teacher salaries increase by 84 cents for every 1 dollar increase in state revenue, compared to just 32 cents per every dollar increase in state revenue in school districts with average teachers’ union power.41 When new revenue comes to school districts from reforms to school finance systems, teacher salaries increase by $4,730 in states with strong teachers’ unions compared to little to no change to the salaries of teachers in states with weaker teachers’ unions.42 In other words, strong teachers’ unions successfully direct new revenue for schools towards teacher salaries.
Teachers’ unions have a strong preference for allocating higher salary increases to veteran teachers.43 Some studies have found that union activity increases the salaries of experienced teachers by as much as 28 percent with no change in the salaries of novice teachers. 44 This is less the case in areas where teachers’ unions are weakened. In the case of Act 10, after existing CBAs expired, pay schemes began to move away from allocating based on seniority and towards allocating based on flexible and more individualized negotiations with teachers. As a result, higher quality teachers received higher pay. Teachers in Wisconsin received a 0.4 percent higher salary for each standard deviation increase in teaching quality.45
Teachers’ unions also help mitigate gender pay gaps. In school districts with traditional collective bargaining, the gender pay gap is $790 lower than in school districts without collective bargaining.46 The gender pay gap among teachers increased in Wisconsin following the introduction of flexible pay arrangements, mainly because female teachers were between 12 percent and 23 percent less likely than men to have experience negotiating pay.47
While it is likely that CBAs meaningfully affect working conditions beyond salaries, very little research has explored the relationship between collective bargaining and improvements in other working conditions. Much of the research that has done so focuses narrowly on classroom time and class sizes. Prior research finds that unionization/collective bargaining is associated with a decrease in the hours teachers spend in the classroom.48 The effect on class sizes is less clear. Some research finds that unionization and bargaining are associated with a decrease in class sizes.49 However, other studies find that unions are regularly willing to trade higher class sizes for improved teacher salaries, as mentioned above.50 This is consistent with surveys that find teachers are willing to have larger class sizes in exchange for higher salaries.51
Teachers’ Unions and Teacher Turnover
Consistent with the idea that when employees can utilize voice, they are less likely to exit jobs, research suggests that teachers’ unions and collective bargaining reduce teacher turnover. Teachers are less likely to turn over in areas where unions are stronger. For every 10 percent increase in the density of unions, experienced and novice teachers are 4 to 7 percent less likely to quit.52 Multiple researchers have also explored what happens to teacher turnover when teachers’ unions are weakened. Following Act 10 in Wisconsin, teacher turnover increased by approximately 1 percentage point per year.53 Retirements also increased by 18 percentage points among those aged 55 and older in the year after Act 10.54 Additionally, higher-quality teachers sought out school districts using flexible pay schemes that paid more for additional student learning, while lower-quality teachers sought out school districts that still used the more traditional step-and-column salary schedule.55 Further, Act 10 led to an increase of teaching degrees at Wisconsin institutions by 1.5 percentage points, or about 20 additional degrees per institution, with these teaching candidates increasingly coming from more selective institutions with average ACT scores of one or more standard deviations above the mean.56 This may be because higher-quality teacher candidates could make more under the flexible pay reforms.
Teachers’ Unions and Student Performance
Having established that teachers’ union activity is related to increased expenditures on education, the final concern centers on their effect on student performance. The best available research evidence suggests that teachers’ unions have no impact on student performance or a slight negative impact on near-term student performance. Marginal increases in the strength or restrictiveness (the degree to which a given CBA limits school and school district leaders in administering schools) of CBAs have no impact on student achievement.57 However, when paired with evidence on teachers’ unions and CBAs impact on expenditures, school districts become less efficient in improving student performance in the presence of teachers’ unions. Researchers have measured this by creating school efficiency ratios that divide student achievement proficiency rates by school district expenditures. This research finds that a standard deviation increase in contract restrictiveness is associated with a decrease in school district efficiency of 1.3% to 2.4% (or 0.03–0.05 SD).58 In other words, as the strength of school districts’ CBAs increase, they must spend 1.3% to 2.4% more (or about 200 dollars per pupil), on average, to achieve the same level of student performance. In another study, researchers found that school districts that allocated new revenue from tax referendums prior to their next bargaining cycle experience an extra 0.002 SD increase in student achievement for every $1,000 in annual per-pupil expenditures. The researchers interpret this as evidence of the rent-seeking impact of teachers’ unions, as school districts not under the pressure of collective bargaining negotiations more efficiently allocate their resources to improve student achievement.59
There may be one context where teachers’ union power improves student performance: Historic state spending increases on education. During the historic court-ordered school finance reforms of the 1970’s and 1980’s, states sent additional funds to school districts, particularly those serving students from economically disadvantaged backgrounds.60 Recall from the discussion on teachers’ unions’ impact on school expenditures that school districts with stronger teachers’ unions maintain their local tax effort to a far greater degree than school districts in states with weak teachers’ unions when receiving new state revenue from school finance reforms. The maintenance of this local tax effort coupled with the new state revenue as a result of union involvement translates into academic benefits for students. After school finance reforms, student achievement in school districts with average state teachers’ union power increased by 0.009 SD per year whereas student achievement increased by 0.013 SD per year in school districts with strong teachers’ unions.61 Furthermore, the effect of strong teachers’ unions on achievement was particularly concentrated in low-income school districts after the reforms. Low-income school districts in states with strong unions saw an increase in achievement by 0.16 SD per year, compared to 0.08 SD per year in low-income districts with weak teachers’ unions, and no significant change in achievement in high-income school districts with weak and strong unions. This can be taken as evidence that strong teachers’ unions, by striving to maximize the welfare of their members, make large school finance reforms more effective at raising student achievement. They do this by ensuring that new state school funding does not supplant but instead supplements local school funding.
The long term impacts of teachers’ unions on students may be negative, though more research needs to be done in this area. One study explored the impact of the passage of laws requiring that school districts negotiate CBAs with teachers’ unions (which mainly occurred between 1960 and 1987) on later life earnings. They found that students exposed to 10 years of collective bargaining law from kindergarten to 12th grade experienced a decrease in annual earnings by $2,134 (3.93 percent) and a decrease in the number of weekly hours worked by 0.42 hours at ages 35 through 49. 62 Additionally, exposure to such laws led to a 1 percentage point decrease in the likelihood of employment, a 0.8 percentage point decrease in the likelihood of labor force participation, and an increase in the likelihood of sorting into a low-skill occupation. These effects were found among men but not women. Exposure to bargaining laws did not significantly decrease years of schooling, which is surprising given the impacts on labor market outcomes. The authors interpreted the effects as evidence that collective bargaining decreased the quality of education – at least for men -- such that students had fewer skills, particularly non-cognitive skills, in adolescence and early adulthood that would make them successful in the labor market.
When teachers’ unions are substantially weakened by state policies, short-run student achievement declines or does not change, but long-run student achievement may improve. In the case of Wisconsin’s Act 10, achievement decreased by 0.20 standard deviations (SD), primarily at low-performing schools.63 Right-to-Work policies, which again removed the ability of teachers’ unions to charge non-union members for collective bargaining representation, resulted in no change in student outcomes.64 In terms of longer run impacts, as mentioned earlier, Act 10 gave school districts the opportunity to adopt flexible pay schedules that compensated teachers for student performance gains. Districts that adopted such schedules experienced an influx of higher quality teachers. This influx was associated with a 0.06 SD increase in math performance and a 0.04 SD increase in reading performance. 65
Finally, research suggests that the effects of CBAs on student achievement are not uniform across all areas of the contract and for all students. Researchers found that, for example, increases in the restrictiveness of CBA class size provisions were associated with decreases in student achievement gaps, while increases in the restrictiveness of CBA teacher evaluation and teacher leave policies were associated with increases in student achievement gaps. These effects were small in magnitude (between 0.02 SD and 0.05 SD), mainly in math, and primarily affected low-income students relative to high-income students.66 Thus, the equity implications of CBAs are not considerably positive or negative; However if those negotiating contracts are looking for ways to improve the small negative impact of contract language on students from economically disadvantaged backgrounds, they may look to renegotiate common CBA evaluation systems, like those built primarily on administrator observations of teachers. These systems tend to strain the capacity of school principals and do little to improve student performance and student achievement gaps.67
Policymakers, practitioners, and scholars continue to debate the role and impact of teachers’ unions in public education in the United States. Even though policies governing teachers’ unions has shift significantly over the last decade—due to changes to collective bargaining rules and a Supreme Court ruling that limits public sector unions’ ability to collect fees for union services—teachers’ unions are still visible in multiple facets of the education system. As highly-organized interest groups, they shape federal, state, and local elections, where their endorsed candidates win more often than they lose. They actively negotiate CBAs with school district administration in 44 states. They organize teachers for collective action, and the number of strikes and walkouts by teachers’ unions has increased in recent years.68
Whether the education system would be better off in the absence of teachers’ unions is difficult to determine. During the historic school finance reforms of the 1970’s and 1980’s, they helped drive educational spending in ways that were productive for students. More recently, however, increases in the strength of CBAs has yielded higher school expenditures but no changes in student performance. Where teachers’ unions have been substantially weakened through restrictions on collective bargaining, school expenditures decreased and student achievement declined in the short-run but increased in the long-run. And in states that have made changes to other union-related laws, like the ability of teachers’ unions to collect membership dues (i.e., Right-to-Work), student achievement did not measurably change, but school expenditures declined.
Ultimately, the current state of research points to some concerns about the impact of teachers’ unions on the efficiency of the education system, consistent with the rent-seeking perspective. The rent-seeking perspective may be correct that teachers’ unions, at least more recently, have directed educational expenditures that, if used in other ways, would be more effective as improving student achievement. It follows then, that further restricting teachers’ union influence, such as narrowing the scope of collective bargaining to fewer issues, may make school spending more efficient. However, doing so will likely come at the cost of greater teacher turnover, at least in the short-term.
Areas for Future Research
While the rent-seeking and voice perspectives provide helpful lenses through which to understand teachers’ unions and their impact on education, some have recently critiqued these viewpoints for their “industry-specific” conception of teachers’ unions—that is, the perspectives are primarily focused on what unions do within the education sector and are less concerned with teachers’ unions’ impact on the advancement of middle- and working- class interests, more broadly.69 For example, more teachers’ unions in the United States are organizing for the “common good” and making demands around matters of broader social justice issues like affordable housing, healthcare and racial justice.70 Are these rent-seeking/voice behaviors or something entirely different? This area of research is currently limited to a few case studies, some of which are outside of the United States. Broadening the conceptualization of teachers’ unions in new research is important for a few reasons. First, doing so will more fully capture the impact of teachers’ unions beyond the education-specific effects currently measured in the research literature. For example, researchers might explore the impact of teachers’ union power on income and racial inequality. Second, doing so will help highlight union activities beyond local-level collective bargaining, where most of the current research literature focuses. For example, researchers might highlight the extent to which teachers’ unions build progressive coalitions in the state legislature to influence policies outside of education.
More work can also be done within education to understand the contexts in which teachers’ unions enhance student achievement versus the contexts in which they harm student achievement. For example, as already mentioned, in the context of state school finance reforms, teachers’ unions appear to facilitate improvements in student achievement by helping maintain local investments in education. However, when school districts and teachers’ unions negotiate over local tax revenue as a result of a passed tax referendum, they experience lower returns to student achievement than when not negotiating. Are there certain thresholds of teachers’ union power, where if fallen short of or exceeded, their influence becomes less efficient for driving educational improvement? Are there specific areas of the CBAs that if removed/added to the scope of bargaining, educational outcomes will improve? This type of research will help policymakers and educational leaders better understand the contexts where teachers’ unions help drive educational improvements versus the contexts where they do not.
Form LM-2 Labor Organization Annual Report, National Education Association and American Federation of Teachers, 2023↩︎
Murphy Marjorie, Blackboard unions: The AFT and the NEA, 1900-1980, (Ithaca: Cornell University Press, 1990), vol.124.↩︎
“Collective bargaining laws,” National Council on Teacher Quality, updated January 2019, https://www.nctq.org/contract-database/collectiveBargaining.↩︎
Bradley D. Marianno. "Teachers’ unions on the defensive?: How recent collective bargaining laws reformed the rights of teachers." Journal of School Choice 9, no. 4 (2015): 551-577. And Marianno, Bradley D. Marianno and Katharine O. Strunk. "After Janus: A new era of teachers’ union activism." Education Next 18, no. 4 (2018).↩︎
Caroline M. Hoxby, “How teachers' unions affect education production,” The Quarterly Journal of Economics 111, no. 3 (1996): 671-718.↩︎
Joshua M. Cowen, and Katharine O. Strunk. "The impact of teachers’ unions on educational outcomes: What we know and what we need to learn." Economics of Education Review 48 (2015): 208-223.↩︎
Joshua M. Cowen and Katharine O. Strunk, "The impact of teachers’ unions on educational outcomes: What we know and what we need to learn." Economics of Education Review 48 (2015): 208-223.↩︎
Bradley D. Marianno and Katharine O. Strunk, “The bad end of the bargain?: Revisiting the relationship between collective bargaining agreements and student achievement,” Economics of Education Review, no.65 (2018): , 93-106.↩︎
Michael. F. Lovenheim, “The effect of teachers’ unions on education production: Evidence from union election certifications in three midwestern states,” Journal of Labor Economics, no.27 (2009): 525-587.↩︎
Jordan D. Matsudaira and Richard W. Patterson, “Teachers’ unions and school performance: Evidence from California charter schools,” Economics of Education Review, no.61 (2017): 35-50.↩︎
Melissa A. Lyon, “The Effect of Right to Work Laws on union membership and school Resources: Evidence from 1942–2017,” Educational Researcher 52, no.6 (2023): 339-347.↩︎
Eric Brunner, Joshua Hyman, and Andrew Ju, “School finance reforms, teachers' unions, and the allocation of school resources,” Review of Economics and Statistics 102, no.3 (2020): 473-489.↩︎
Jason E. Baron, “The effect of teachers’ unions on student achievement in the short run: Evidence from Wisconsin’s Act 10,” Economics of Education Review, no.67 (2018): 40-57.↩︎
Katharine O. Strunk and Jason A. Grissom, “Do strong unions shape district policies? Collective bargaining, teacher contract restrictiveness, and the political power of teachers’ unions,” Educational Evaluation and Policy Analysis 32, no.3 (2010): 389-406.↩︎
Strunk, Katharine O., and Sean F. Reardon. "Measuring the strength of teachers' unions: An empirical application of the partial independence item response approach." Journal of Educational and Behavioral Statistics 35, no. 6 (2010): 629-670.↩︎
Katharine O. Strunk, Joshua Cowen, Dan Goldhaber, Bradley D. Marianno, Roddy Theobald, and Tara Kilbride, “Public school teacher contracts and state-level reforms: Assessing changes to collective bargaining restrictiveness across three states,” American Educational Research Journal 59, no. 3 (2022): 538-573; Bradley D. Marianno, David S. Woo, and Kate Kennedy, “Collective Bargaining Agreement Restrictiveness in Unionized Charter Schools,” Educational Policy (2023): 08959048231178024; Bradley D. Marianno and Annie A. Hemphill, “Pandemic memoranda: how school district and teachers' union administrators modified collective bargaining agreements during the COVID-19 pandemic,” Journal of Educational Administration 61, no. 6 (2023): 604-622.↩︎
Michael Hartney and Patrick Flavin, “From the schoolhouse to the statehouse: Teacher union political activism and US state education reform policy,” State Politics & Policy Quarterly 11, no. 3 (2011): 251-268; Bradley D. Marianno, “Compared to what? Changes in interest group resources and the proposal and adoption of state teacher policy,” Policy Studies Journal 48, no. 4 (2020): 982-1022.↩︎
Bradley D. Marianno, “Compared to what?” 982-1022.↩︎
Amber M. Northern, Janie Scull, and Dara Zeehandelaar, “How strong are US teacher unions? A state-by-state comparison.” Thomas B. Fordham Institute, retrieved February 18 (2012): 2018.↩︎
Michael T. Hartney, How policies make interest groups: Governments, unions, and American education (Chicago, IL: University of Chicago Press, 2022).↩︎
Terry M. Moe, “The union label on the ballot box: How school employees help choose their bosses.” Education Next 6, no. 3 (2006): 58-67.↩︎
Terry M. Moe, Special interest: Teachers unions and America’s public schools (Washington, D.C.: Brookings Institution Press, 2011).↩︎
Katharine O. Strunk and Jason A. Grissom, “Do strong unions shape district policies?” 389-406.↩︎
Michael T. Hartney, “Teachers’ unions and school board elections: A reassessment,” in Interest Groups in US Local Politics, edited by Sarah Anzia (Cham: Springer Nature Switzerland, 2023), 59-84.↩︎
Matt Grossmann, Sarah Reckhow, Katharine O. Strunk, and Meg Turner, “All states close but red districts reopen: The politics of in-person schooling during the COVID-19 pandemic,” Educational Researcher 50, no. 9 (2021): 637-648.↩︎
Bradley D. Marianno, Annie A. Hemphill, Ana Paula S. Loures-Elias, Libna Garcia, Deanna Cooper, and Emily Coombes, “Power in a pandemic: Teachers’ unions and their responses to school reopening,” AERA Open 8 (2022): 23328584221074337; Michael T. Hartney, and Leslie K. Finger, “Politics, markets, and pandemics: Public education’s response to COVID-19,” Perspectives on Politics 20, no.2 (2022): 457-473; Matt Grossmann et al., 637-648.↩︎
Michael T. Hartney, How policies make interest groups.↩︎
Melissa A. Lyon, “Heroes, villains, or something in between? How”Right to Work” policies affect teachers, students, and education policymaking,“ Economics of Education Review 82 (2021): 102105.↩︎
Bradley D. Marianno, “Compared to what?” 982-1022.↩︎
Michael Hartney and Patrick Flavin, “From the schoolhouse to the statehouse,” 251-268.↩︎
Bradley D. Marianno, “Down but not out: The National Education Association in federal politics,” Educational Policy 32, no. 2 (2018): 234-254.↩︎
Caroline M. Hoxby, “How teachers' unions affect education production,” The Quarterly Journal of Economics 111, no. 3 (1996): 671-718.↩︎
Bradley D. Marianno, Paul Bruno, and Kathrine O. Strunk, “The effect of teachers’ union contracts on school district efficiency: Longitudinal evidence from California,” Sage Open 11, no. 1 (2021): 2158244020988684; Katharine O. Strunk, “Are teachers' unions really to blame? Collective bargaining agreements and their relationships with district resource allocation and student performance in California,” Education Finance and Policy 6, no. 3 (2011): 354-398; Randall W. Eberts, “How unions affect management decisions: Evidence from public schools,” Journal of Labor Research 4, no. 3 (1983): 239-247; Randall W. Eberts and Joe A. Stone, Unions and Public Schools: The Effect of Collective Bargaining on American Education (Lexington, MA: Lexington Books, 1984).↩︎
Eric Brunner, Joshua Hyman, and Andrew Ju, “School finance reforms, teachers' unions, and the allocation of school resources,” Review of Economics and Statistics 102, no. 3 (2020): 473-489.↩︎
Jason Cook, Stéphane Lavertu, and Corbin Miller, “Rent-Seeking through collective bargaining: Teachers unions and education production,” Economics of Education Review 85 (2021): 102193.↩︎
Melissa A. Lyon, “The Effect of Right to Work Laws on union membership and school Resources: Evidence from 1942–2017,” Educational Researcher 52, no. 6 (2023): 339-347.↩︎
Eunice S. Han and Emma Garcia, “The Effect of Changes in Legal Institutions Weakening Teachers’ Unions on Districts’ Spending on Teacher Compensation,” American Journal of Education 130, no. 2 (2024): 000-000.↩︎
Nina Bascia, “Teachers as professionals: Salaries, benefits and unions,” International handbook of research on teachers and teaching (2009): 481-489.↩︎
Hoxby’s (1996) difference-and-difference and instrumental variables (IV) estimates suggest that teacher salaries in unionized districts are 5 percent higher, on average, than those in non-unionized districts. West and Mykerezi (2011) provide similar estimates through their IV approach, finding that the presence of a CBA is associated with a 3.9 percent increase in average teacher salaries. Baugh and Stone (1982) explicitly evaluated teachers’ unions’ influence on salary levels by estimating cross-sectional wage-level and wage change regressions on a national sample of teacher data. They find that the estimated premium for joining a union was 12 percent in the 1977-1978 school year (up from 4 percent in the 1974-1975 school year). Brunner and Squires (2013) employ a fixed-effects approach that controls for time-invariant unobserved sources of omitted variable bias, but may not fully account for all time-varying impacts, and find that teachers’ unions increase overall teacher salary levels by 4.5 percent.↩︎
Jason E. Baron, “The effect of teachers’ unions,” 40-57; Andrew, Litten “The effects of public unions on compensation: Evidence from Wisconsin,” Unpublished Paper (2016).↩︎
Eric Brunner, Joshua Hyman, and Andrew Ju, “School finance reforms,” 473-489.↩︎
Tuan D. Nguyen, J. Cameron Anglum, and Michael Crouch, “The effects of school finance reforms on teacher salary and turnover: Evidence from national data,” AERA Open 9 (2023): 23328584231174447.↩︎
Jason A. Grissom and Katharine O. Strunk, “How should school districts shape teacher salary schedules? Linking school performance to pay structure in traditional compensation schemes,” Educational Policy 26, no. 5 (2012): 663-695.↩︎
John V. Winters, “Teacher salaries and teacher unions: A spatial econometric approach,” ILR Review 64, no. 4 (2011): 747-764; Eric J. Brunner and Tim Squires, “The bargaining power of teachers’ unions and the allocation of school resources,” Journal of Urban Economics 76 (2013): 15-27; Heather Rose and Jon Sonstelie. “School board politics, school district size, and the bargaining power of teachers’ unions,” Journal of urban economics 67, no. 3 (2010): 438-450.↩︎
Barbara Biasi, “The labor market for teachers under different pay schemes,” American Economic Journal: Economic Policy 13, no. 3 (2021): 63-102.↩︎
Eunice S. Han, “The Effects of Teachers’ Unions on the Gender Pay Gap among US Public School Teachers,” Industrial Relations: A Journal of Economy and Society 59, no. 4 (2020): 563-603.↩︎
Barbara Biasi and Heather Sarsons, “Flexible wages, bargaining, and the gender gap,” The Quarterly Journal of Economics 137, no. 1 (2022): 215-266.↩︎
Randall W. Eberts, “Union effects on teacher productivity,” ILR Review 37, no. 3 (1984): 346-358; Michael A. Zigarelli, “Dispute resolution mechanisms and teacher bargaining outcomes,” Journal of Labor Research 17, no. 1 (1996): 135-148.↩︎
Caroline Minter Hoxby, “How teachers' unions affect education production,” The Quarterly Journal of Economics 111, no. 3 (1996): 671-718.↩︎
Eric J. Brunner and Tim Squires, “The bargaining power of teachers’ unions,” 15-27; Clayton W. Hall and Norman E. Carroll, “The effect of teachers' organizations on salaries and class size,” ILR Review 26, no. 2 (1973): 834-841; Heather Rose and Jon Sonstelie, “School board politics,” 438-450; Stephen A. Woodbury, “The scope of bargaining and bargaining outcomes in the public schools,” ILR Review 38, no. 2 (1985): 195-210.↩︎
Dan Goldhaber, Michael DeArmond, and Scott DeBurgomaster. "Teacher attitudes about compensation reform: Implications for reform implementation." ILR Review 64, no. 3 (2011): 441-463.↩︎
Eunice S. Han, “The myth of unions’ overprotection of bad teachers: Evidence from the district–teacher matched data on teacher turnover,” Industrial Relations: A Journal of Economy and Society 59, no. 2 (2020): 316-352.↩︎
Jason E. Baron, “The effect of teachers’ unions,” 40-57.↩︎
Jonathan Roth, “Union reform and teacher turnover: Evidence from Wisconsin’s Act 10,” Harvard Kennedy School (2017).↩︎
Barbara Biasi, “The labor market for teachers,” 63-102.↩︎
Jason E. Baron, “Union Reform, Performance Pay, and New Teacher Supply: Evidence from Wisconsin's Act 10,” AEA Papers and Proceedings 111 (2021): 445-49.↩︎
Bradley D. Marianno and Katharine O. Strunk, “The bad end of the bargain?,” 93-106.↩︎
Bradley D. Marianno, Paul Bruno, and Kathrine O. Strunk, “The effect of teachers’ union contracts on school district efficiency”↩︎
Jason Cook, Stéphane Lavertu, and Corbin Miller, “Rent-Seeking through collective bargaining”↩︎
C. Kirabo Jackson, Rucker Johnson, and Claudia Persico. The effect of school finance reforms on the distribution of spending, academic achievement, and adult outcomes. No. w20118. National Bureau of Economic Research, 2014.↩︎
Eric Brunner, Joshua Hyman, and Andrew Ju, “School finance reforms,” 473-489.↩︎
Michael F. Lovenheim and Alexander Willén., "The long-run effects of teacher collective bargaining." American Economic Journal: Economic Policy 11, no. 3 (2019): 292-324.↩︎
Jason E. Baron, “The effect of teachers’ unions,” 40-57.↩︎
Melissa A. Lyon, “Heroes, villains, or something in between?”↩︎
Barbara Biasi, “The labor market for teachers,” 63-102.↩︎
Bradley D. Marianno, "A negotiated disadvantage? California collective bargaining agreements and achievement gaps." Educational Researcher 50, no. 7 (2021): 451-462.↩︎
Joshua Bleiberg, Eric Brunner, Erica Harbatkin, Matthew A. Kraft, and Matthew G. Springer. Taking teacher evaluation to scale: The effect of state reforms on achievement and attainment. No. w30995. National Bureau of Economic Research, 2023.↩︎
Melissa Arnold Lyon, Matthew A. Kraft, and Matthew P. Steinberg. The Causes and Consequences of US Teacher Strikes. No. w32862. National Bureau of Economic Research, 2024.↩︎
Melissa Arnold Lyon, "Current perspectives on teacher unionization, and what they’re missing." Educational Policy 37, no. 5 (2023): 1420-1443.↩︎
McCartin, Joseph A., Marilyn Sneiderman, and Maurice BP-weeks. "Combustible convergence: Bargaining for the common good and the# RedforEd uprisings of 2018." Labor Studies Journal 45, no. 1 (2020): 97-113.↩︎
Marianno, Bradley D. (2025). "Teachers' Unions and Collective Bargaining," in Live Handbook of Education Policy Research, in Douglas Harris (ed.), Association for Education Finance and Policy, viewed 04/11/2025, https://livehandbook.org/k-12-education/workforce-teachers/unions-and-collective-bargaining/.