The Effect of Right-to-Work on Unfair Labor Practice Charges: Synthetic Control Evidence From Indiana and Michigan

Abstract

This study investigates the effect of adopting right-to-work on unfair labor practice charges filed at the National Labor Relations Board. Most charges accompany union elections, which are expected to decrease under right-to-work. However, this work's synthetic control method results show that unfair labor practice trends in rates and success are generally unaffected by right-to-work adoption. Underlying these stable rates of filing and success is a more resource-intensive strategy where federations collaborate on charges with their constituent members. Using the cases of Michigan and Indiana, which both adopted right-to-work in 2012 and a unique dataset of all charges filed from 2000 to 2019, I show that legal mobilization is one union activity that adjusts under right-to-work as unions and federations look to protect workers.

Keywords

right-to-work unfair labor practice legal mobilization NLRB

Right-to-work laws impair the organizing of American workers (Hogler 2015; Bruno et al. 2015; Devinatz 2015). During steady union membership decline for the past half century, the recent adoptions of right-to-work laws emphasize labor's decline in the United States. The effects of right-to-work laws go beyond union membership–depressing wages and political participation (Kogan 2016; Feigenbaum, Hertel-Fernandez, and Williamson 2018; VanHeuvelen 2020). However, little attention has been paid to how these laws affect labor activities (Elliot & Huffman 1984). By leveraging a unique dataset of all unfair labor practice (ULP) complaints filed at the National Labor Relations Board (NLRB) from 2000 to 2019, this research investigates to what extent right-to-work laws impact legal mobilization by workers. One may expect ULP filings to decline commensurately with membership declines under right-to-work. However, this research shows that filing complaints remains largely unchanged. This result suggests that labor may lean on legal mobilization to protect their interests in an unfavorable legal environment. Additionally, federations may increase their attention to legal mobilization strategies in a new right-to-work context.

The survival of unions is predicated on the ability of labor to provide for workers in terms of benefits, employment protection, and wages. ULP charges are among the few options workers have to enforce these premiums gained through collective bargaining (Rosenfeld 2014; Galiatsos 2015). Yet little research has shown how right-to-work affects this type of legal mobilization. This research addresses this gap by asking how ULP's usage and composition change in a new right-to-work context. On the one hand, unions working to support their constituents with fewer resources may lead to more ULP charges, engaging the state to provide for workers. On the other hand, the declines in membership and union elections due to the law may incur related decreases in ULP usage as a consequence of fewer opportunities to use ULP complaints. This research aims to uncover what changes in legal mobilization have occurred under recent adoptions of right-to-work.

This research uses a synthetic control analysis to identify the effects of right-to-work adoption on legal mobilization behaviors for two states: Indiana and Michigan, which both adopted the law in 2012 (Abadie 2021). These two states are compared in this analysis to track differences in the local organizing contexts. Past research has highlighted how bases of strong unionism, like in Michigan, can take advantage of institutional resources relative to their counterparts in less unionized states like Indiana (Dixon 2008). It is unclear in the literature what sort of shape post-adoption trends in legal mobilization may take. Past work emphasizes that legal mobilization is not just a function of the underlying relationships between employers and employees (Galanter 1974; McCammon 2001). The decision to file ULP is tied up in filers’ economic constraints, unions’ organizational power, and political circumstances. Does the shift in these conditions attributable to right-to-work adoption change the calculus of legal mobilization?

This article aims to track the changes in legal mobilization in a new right-to-work context. Absent strong theoretical foundations for the outcomes of unfair labor practice suits, this work draws on theories of legal mobilization and prior work on ULP to frame the results. The analytic strategy builds on Ozkan and Ozbeklik (2016) by using the synthetic control method for right-to-work outcomes and compares those outcomes in two states. The results show that ULP rates remain largely unchanged after adoption. In the discussion, I develop this finding as evidence for using ULP as a tool to continue to offer protection to union members in a disadvantageous legal context. While right-to-work may increase free riders, unions are still required to cover those workers. And, ULP may be one avenue to secure this coverage with fewer resources. This paper proceeds as follows: a description of the unfair labor practice system at the NLRB, expectations of where right-to-work laws fit into ULP charges, the analytic strategy, results of the synthetic control, and concludes with a discussion of the extent to which right-to-work impacts ULP usage and composition.

Unfair Labor Practice

The National Labor Relations Act of 1935 allows private sector workers to file unfair labor practice complaints against employers who interfere with their rights to organize and collectively bargain. These interferences can take many forms, such as firing workers involved in organizing, discriminatory treatment by management, failure to uphold the contents of active bargaining agreements, and arguing in bad faith during contract negotiations (McDowell et al. 1976; Weiner 2005; Galiatsos 2015). Successful ULP charges can earn remedies court-mandated by the NLRB, like backpay and reinstated employment (Ferguson 2008). Rosenfeld (2014) shows that the use of strikes as the primary thrust of union power has declined for decades, while ULP complaints have increased in kind. As ULP becomes a tool for enforcing collective bargaining agreements, little is known about how this shift to legal mobilization as a primary action fares in a new right-to-work context.

The typical usage of ULP occurs around union elections and pushes for organizing. Organizing a workplace offers more opportunities for interference with the rights offered by the National Labor Relations Act by an employer compared to day-to-day work with or without a union. For example, since the first union organized at Starbucks was elected on December 9, 2021, 856 charges have been brought against Starbucks Corporation (NPR 2022). From 2000 to 2021, just eighty-three charges were brought against Starbucks. ULP charges tend to accompany organizing drives and elections (McCammon 2001).

Since ULP charges are likely around elections, they have a relatively limited scope. This limitation is partly due to the National Labor Relations Act, which establishes the rules for the NLRB running union elections and allows the NLRB to investigate and prosecute ULP cases. These functions of the law are bound together. Employers coercing workers against joining a union is more likely as a new union is being established at a firm location than when a union has been collectively bargaining for some time. This limitation is important in a new right-to-work context where union organizing drives are made more fraught with popular sentiment around the law and its adoption.

In addition to the limited scope, one critique among labor is that ULP charges are ineffective. NLRB decisions on ULP often follow lengthy legal battles that advantage employers (Dannin 2006). Even in a union win, sought-after remedies are viewed as weak: failing to enforce backpay settlements or reelection processes that stifle union support (Bronfenbrenner 2009). The NLRB in 2019 has pushed for shorter turnaround times, but the median time to consider a case resolved was 106 days in Fiscal Year 2018 (NLRB Activity Report 2019). Dannin (2006) makes the case that the letter of the law for the National Labor Relations Act is quite strong in its affirmation of a unionized workplace, but in practice and perception, some weaknesses of the law endure.

Despite these criticisms of ULP, it remains one of the few options to protect union representation in the workplace. In testimony in support of the PRO Act, Richard Trumka, President of the AFL-CIO from 2009 to 2021, cites these weaknesses of the NLRB in the enforcement of union representatives and the brinksman mistreatment of workers by employers up until ULP was filed (Trumka 2019). These legal and interpersonal conditions present a dilemma for organizers. ULP is one of the few viable options to enforce their contracts, and at the same time, ULP seems a weak tool to make these enforcements for its weak remedies and limited scope. Fantasia (1989) describes how bureaucratic approaches, like filing ULP, can hamper the militancy of locals and workers, and yet such bureaucratic approaches were particularly preferable to national unions and federations. These preferences highlight the broader priorities of federations compared to workers. As right-to-work is adopted, federations are looking to maintain coverage since a decrease in membership is expected to follow. Successful union elections may become a larger focus on adoption, and federations could see ULP as a path to ensuring elections run smoothly.

In new right-to-work states, labor is presented with shifting priorities to maintain worker protections. Clawson and Clawson (1999) suggested that the survival of unions in the pro-employer policy regime of the United States is built on the “ability to deliver wages, benefits, and a systematized defense of workers’ everyday workplace rights.” Militant actions like strikes may be less feasible to deliver those benefits given the within-state popularity of the right-to-work laws at adoption (Brooks and Manza 2007; Hogler 2015). Despite the issues with filing ULP, the de jure provisions of the National Labor Relations Act seem to provide these exact premiums suggested by the literature by having legal action available for ensuring employers will bargain with the union. How might right-to-work laws affect the usage of ULP charges considering this filing dilemma among workers?

Right-to-Work

Right-to-work laws prohibit union security agreements making mandatory payments to unions illegal. There are several findings about how this change in dues payment disrupts union organizing. One mechanism is free-riding; the decline in membership becomes inevitable upon the law's adoption because unions must continue to provide support for workers with a smaller monetary base (Ellwood and Fine 1987; Moore 1998; Hogler, Shulman, and Weiler 2004). This consequence of right-to-work undermines the solidarity of workplaces and the sometimes slim majorities that move the needle at union drives and the bargaining table. Another finding is that right-to-work is passed where there is a preference for the law, and diminished union rates following right-to-work adoption are just a realization of this preference (Devinatz 2011; Vedder 2011). Historically, as right-to-work is passed, union membership rates diminish, and fewer union elections occur (Ellwood and Fine 1987). This is important to ULP rates because violations of the NLRA that warrant a ULP charge often occur at the time of elections (McCammon 2001). There is little work on how these consequences of right-to-work relate to legal mobilization strategies. The connection between rates of elections and ULP may not hold under right-to-work as the political opportunity of union organizing changes. Successful union drives and contract bargaining for unions may necessitate legal mobilization where these mechanisms are in play.

While right-to-work laws impact workers specifically, their reach also signals a pro-business environment (Holmes 1998; Lichtenstein 1998; Rao et al. 2011). Employers can feel an added sense of safety from a strike or other union activity in a right-to-work context (Dixon and Martin 2012). Hiring replacement workers comes at a lower cost as unions weaken and the threat of union response is lower. The question, then, is how does ULP fit into this pro-business environment? ULP may decline since most ULP cases are filed around union elections, and organizing where pro-business attitudes coalesce into lawmaking may be a steeper ask of workers. However, ULP may increase as workers look for an avenue to advance their interests with a diminished threat of other actions. Bronfenbrenner (2009) suggests that as employers feel supported in an advantageous policy context, workers could be discouraged from legal mobilization because of threats, harassment, and other retaliatory attacks against organizing. One of the actions that ULP is designed to punish could deter filing ULP. The dilemma of filing is clarified under this theory. The lengthy wait times and weak remedies are not worth the possibility of punishment from employers who can now hire replacements with limited expectations of a response from workers (Dannin 2006). Therefore, ULP cases may decline through worker sentiment, employer behavior, or both.

In contentious times for labor, the symbolic value of protecting labor through institutional means can generate excitement among the rank and file as part of larger organizing pushes (Carson 2010). Legal mobilization has been a branch of organizing efforts in right-to-work contexts (Nissen and Churchill 2020). Further, Nissen and Churchill (2020) posit that legal strategies will fail without vigorous campaigning elsewhere. In the view of organizers, legal mobilization may provide retentive labor wins while policy-focused organizing occurs elsewhere (Dixon 2008). These strategies should be in connection with larger drives. Since unions are still expected to cover their members even as mandatory dues end, ULP filing may be one way to show the value of union membership to workers (McCammon 2001). ULP charges may become more prominent and potentially cushion a transition to a right-to-work context as part of a larger organizing strategy. The ULP mechanism offers support for workers while other political activities work to regain an advantageous policy context.

Legal mobilization by a union is constrained by instrumental resources (McCann 1994). These resources, like local political positions and skilled lawyers, are activated for ULP, particularly around organizing drives. Such mobilizations have been the subject of much legal mobilization research. One enduring insight from the legal mobilization literature is that the “haves” come out ahead (Galanter 1974). Legal systems offer advantages to repeat players, while one-off or infrequent players are disadvantaged. In this case, a repeat player is defined as a union assuming that the organization has dealt with the NLRB at least at the stage of becoming a union and has more access to legal resources than an individual filing a charge. The access to resources for unions is not evenly distributed, however. Historically, unfavorable legislation may push unions to seek outside connections to pool resources in response (Dixon 2010, 2020). Some ULP cases are supported by state and national federations, which could influence ULP outcomes with their larger resource pools. Right-to-work may exacerbate this legal inequality as the ULP dilemma is clarified such that collaborations with other unions and affiliations with larger federations are related to greater success than individual filer counterparts (Fantasia 1989; Bronfenbrenner 2009; Snead 2022).

In past research examining right-to-work's effects on labor union activity, geographical findings have been varied. Past work has found mixed results for right-to-work outcomes on union drives and employment across different states, suggesting that the interruption to organizing due to right-to-work is place-specific (Ferguson 2008; Bruno et al. 2015; VanHeuvelen 2020). Collins (2013) sums up these state-level variations as an argument against a national right-to-work law on the grounds that the observed economic effects are too fraught with the complexities of local economies to pattern in aggregate. To approach this geography question, a key insight comes from Ozkan and Ozbeklik (2016), who show what differences hold between states as a pair. Comparing a pair of states may highlight the findings attributable to differences in each state's organizing base and potential points of similarity in the law's impact.

Analytic Strategy

To estimate the effect of right-to-work laws on ULP charges filed by workers, this research uses a synthetic control methodology (SCM) which has gained traction for its use in policy impact studies (Keele and Minozzi 2013; Mcclelland and Gault 2017; Donohue, Aneja, and Weber 2017; Dunford 2021). SCM builds on traditional difference-in-difference and matching methods by using weighting techniques to build data-driven counterfactuals or “synthetic” units (Xu 2017; Abadie 2021). The interpretation of SCM results is much the same as traditional difference-in-difference methods, where we expect an obvious break in trends post-treatment. The key advantage for SCM over a difference-in-difference is that this break in trends is measured against a matched comparison case in the synthetic trend. The synthetic trend is a weighted average of other units in the sample (in this case, non-RTW states) to be as similar as possible to the treated unit in the pretreatment time such that one can calculate a view of how the outcome may have looked had the treated unit not received treatment. SCM is applied to Indiana and Michigan separately to assess the impact of adopting right-to-work laws in each state, allowing for the variation between states to be analyzed.

The SCM model used here constructs a counterfactual synthetic control unit by weighing control-group states who did not adopt right-to-work such that a weighted average of outcomes and relevant covariates in the pre-intervention period will resemble a treated state (i.e., a focal state that adopted right-to-work either Indiana or Michigan). This donor pool of control states is all states who did not have a right-to-work law during the time period (2000 to 2019). States that would eventually adopt the law during the sample period are dropped from the donor pool. The synthetic control outcome based on these weighted averages is then used as the counterfactual to the observed treated state. I define $α_{i t}$ as the effect of right-to-work adoption in state i at year t. Then I use $D_{i t}$ as a treatment identifier which satisfies
$D = {1 if R TW is adopted in state i, 0 otherwise} .$
The observed outcome variable $Y_{i t}$ can be shown as:
$Y_{i t} = Y_{i t}^{N} + α_{i t} D_{i t},$
where $Y_{i t}^{N}$ is a counterfactual outcome that would be found if there was no adoption of right-to-work by a state. A focal state that adopts right-to-work is the “treated” unit and assumes, this is, the only state $i = 1$ exposed to right-to-work after a year $t = T_{0}$ . Then $D_{i t} = 1$ if $i = 1$ and $t > T_{0}$ and $D_{i t}$ otherwise. I estimate $α_{i t}$ after $t = T_{0}$ like:
$α_{1 t} = Y_{1 t} - Y_{1 t}^{N} f o r t > T_{0} .$
Since $Y_{1 t}$ is observed, only $Y_{1 t}^{N}$ must be estimated to find $α_{1 t}$ . A synthetic control unit is a weighted average of units in the “donor pool”: the states that do not have right-to-work on the books. These donors construct the $Y_{1 t}^{N}$ . The synthetic control unit can be represented by $(J \times 1)$ vector of weights $W = (w_{2}, \dots, w_{J + 1})$ for states $j = 2, \dots, J + 1$ which satisfy $0 \leq w_{j} \leq 1$ , and $w_{2} + \dots + w_{j + 1} = 1$ to avoid extrapolation. The vector of optimal weights $W * = (w_{2} , \dots, w_{J + 1} )^{'}, Y_{1 t}^{N}$ is estimated as the weighted average of $Y_{j t}$ . Then $Y_{1 t}^{N}$ or the effect of the right-to-work adoption on the outcome in the treated state is estimated as follows:
$\hat{α_{1 t}} = Y_{1 t} - \sum_{j = 2}^{J + 1} w_{j} * Y_{j t} .$
The vector of optimal weights $W $ is calculated so that W minimizes the difference in the pre-intervention predictors of the treated unit and the weighted average of predictors for the states in the donor pool. I follow Abadie (2021) in calculating weights to minimize the mean-squared prediction error of the subsequent vector V of the weights predicting the pre-treatment outcome values of state i who adopted right-to-work. Further, Abadie, Diamond, and Hainmueller (2015) showed that the estimator of the SCM is unbiased if the treated unit and the synthetic control unit are matched in observed covariates and a long set of outcome variables in the pre-intervention periods assuming a linear factor model that allows for time-varying unobserved confounding effects.

The results of the SCM are presented in figures to highlight the break in trends after treatment. The first figure shows the changes in the counts of union membership and union elections to illustrate how general union outcomes fare after adoption in the focal states. Next, the ULP outcomes are covered for all cases, then for just collaborations. These outcomes’ findings are discussed with typical robustness checks (Abadie, Diamond, and Hainmueller 2015). Then, the between-state differences are discussed in view of the union power within each state. Finally, more general conclusions are drawn about the impact of right-to-work on ULP usage and success.

Case Selection

Five states have adopted right-to-work laws since 2012: Indiana, Michigan, Kentucky, Wisconsin, and West Virginia (Hogler 2015; Manzo 2015; Kogan 2016). This study focuses on Indiana and Michigan, the first among the recent adopters to bring right-to-work back into serious contention in the national legislative landscape, with only Oklahoma and Idaho adopting in the prior thirty years. These two states are selected for two reasons.

The first reason is methodological. The synthetic control method requires that the treated units have no interference between units. Since both of these cases adopted the law in the same year, there is less risk of interference in these two states compared to the latter three. The adjacency of the Indiana and Michigan may be of concern. However, union organizers in Michigan were surprised to have the law adopted (Hogler 2015). The law was not on Governor Rick Snyder's legislative agenda, and the Republican-led state legislature pushed the adoption in the final weeks of their lame-duck session (Devintaz 2015; Kaminski 2015). Given the Michigan adoption circumstances, the potential anticipation effects that transfer from Indiana are likely minimal. This same reason excludes the later states from the selection since their adoption follows a larger push for right-to-work throughout the U.S. Spillover in ULP filing practices among federations is much more likely for the later adopters. Moreover, these later adopters do not contribute to the donor pool either since their treatment status changes during the length of the panel.

The second reason for selecting Indiana and Michigan is the possibility of comparisons between a weak and strong resource position of labor. In Michigan, the birthplace of United Autoworkers, private-sector labor unions have been the forerunners in concerted actions for labor in the United States (Babson 1986; Smith 2001; Minchin 2017). Workers in Michigan consistently boast higher levels of union coverage on average compared to the rest of the United States (BLS 2019). National and institutional union strength has long been housed in Michigan (Lichtenstein 2002). The relative power position of labor in Michigan regarding influence in local government is higher than in Indiana. Indiana is much weaker regarding union strength and symbolic standing for the labor movement (Cantin 2012; Kahlenberg and Greene 2012). In the year leading up to right-to-work adoption, the local legislative climate seemed much more amenable to right-to-work than their counterpart in Michigan, with the housespeaker for Indiana calling the adoption of right-to-work the “No. 1 priority” (Manzo 2015). Their differences may highlight disparate ways right-to-work impacts legal mobilization and characterize expectations among other recent and future adopters. Further, Michigan is set to repeal its right-to-work law in 2024. The legal mobilization differences in each state may offer insight into the organizing strategy that went into this legislation change.

Data and Variables

This work uses a dataset of ULP charges at the NLRB from 2000 to 2019. This census of ULP filings from 2000 to 2019 was collected via web scraping of the NLRB website using the R package rvest. The initial scraping procedure required downloading all the cases available through the NLRB website, and then scraping the web page per case number to store the relevant data. The filing data available on the NLRB website only goes back in time to roughly 2008, depending on the state. Therefore, their old proprietary filing system, “CATS,” is recorded to reproduce the relevant information in the current NLRB filing system for the remaining years. The total unique ULP cases, including the scraped current data and the harmonized, historical CATS data total 541,033 separate ULP cases. While this data is at the case level, it is aggregated this data to the state-year level to take advantage of the panel functionality of the synthetic control method.

This case data only includes some information regarding each ULP charge in the interest of privacy. For each case, the entries for the charging party are filed as “union(s),” “individual,” or both with the possibility of support from federations. The names of individuals who are filing are censored, as well as the addresses of firms. I have also coded “collaborations,” defined as when federations are included in a charging party. Collaborations have been added to the analysis to make sense of charging party behaviors in line with the legal mobilization literature ideas of the haves coming out ahead (Galanter 1974). The data contains the type of charge alleged through the statutes under section 8(a) of the NLRA and the outcome of the charge: withdrawal, dismissal, informal settlement, formal settlement, Administrative Law Judge decisions, and Board orders. The opening and closing dates of the charge processed by the NLRB are included. The distribution of ULP charge usage, success, and charging party composition over time is analyzed by aggregating these cases to the state-year level.

This work examines several outcomes of interest to make sense of ULP charges after right-to-work adoption. Two outcomes aim to capture union organizing outcomes: the percentage of workers covered by a union and the count of union certification elections. Then the ULP outcomes are examined: total cases filed, total cases won, collaborations filed, and collaborations won. Each ULP outcome is a count variable coded per 1,000 establishments with employees in the state of interest. ULP is measured as a rate per 1,000 establishments to capture the relationship between violations of the NLRA and the contract in question. Since most ULP charges occur around union elections, the establishment is a useful unit of analysis.

Figure 1 shows the trend in counts of ULP charges filed against employers in each of the focal states compared to the average counts of charges for states that have had right-to-work since before 2000 and states that have not enacted the law during the sampling period of 2000 to 2019. Figure 1 also shows trends in the filing and success of ULP charges. The trends in Panel A draw attention to a historical artifact in the analysis. The Great Recession is associated with a spike in charges filed, particularly for workers in Michigan. To that end, a lagged predictor for each outcome is added to the analysis to adjust for this historical event.

Figure 1.
Unfair labor practice (ULP) trends.

Predictors included in this analysis focus on local economic context, political composition, and union resources. These predictors are used to create the synthetic comparison state for the results. The economic context is operationalized as GDP per capita, % unemployment, % of workers in manufacturing, taxes on production, benefits in dollars given to employees, and income per capita (Dixon 2008; Eren and Ozbeklik 2016). Political composition is demonstrated through the percentage of Democrats in the state legislature per year and the NLRB's political composition (Feigenbaum, Hertel-Fernandez, and Williamson 2018; Semett 2016; Devinatz 2015). Union resources are considered through % union membership, the number of unions at each OLMS size, and union assets in dollars drawn from OLMS data (Dixon and Martin 2012; Kimeldorf 2013). These predictors are drawn from various sources outlined in Supplemental Appendix Table A. Lags of the dependent variable are included for 2008 and 2010 to capture changes related to the Great Recession regarding legal mobilizations (Milkman and Luce 2017). Lags for union elections are also included for the last two years before adoption to capture the relationship between ULP and elections (McCammon 2001). All data in this analysis is aggregated to the state-year level to provide the panel data required for the synthetic control method.

Results

The thrust of this analysis implements the synthetic control method to derive the effect of right-to-work adoption on the usage and success of ULP by workers in Indiana and Michigan. For each outcome and each focal state, a graph presents the difference between the treated, observed values and the synthetic, counterfactual values. The numeric presentation of these graphs is reported in Supplemental Appendix Table B, the predictor weights in Supplemental Appendix Table C, and donor state weights in Supplemental Appendix Table D.

Figure 2 shows the results of the SCM on two organizing outcomes. These outcomes highlight the cooccurrence of expected changes to organizing with the ULP trends. The literature suggested that ULP charges are tied to elections, and these results show that certification elections remained stable compared to the synthetic outcome in both states. The stability of these trends aligns with estimates of unionization from Farber et al. (2021). With no obvious break in election trends, we expect ULP filings to remain stable after adoption. The mean square prediction error (MSPE) for the pre-treatment difference in synthetic and observed trends is compared to the post-treatment MSPE using a Fisher's exact test, reported in Table 1. This test asks: How well the observed and synthetic trends align in the pre-treatment time, then compares the quality of this alignment in the post-treatment time. The graphical difference in observed and synthetic trends is robust if the trends do not align in the post-treatment time (Abadie, Diamond, and Hainmueller 2015). The trend in union membership remains stable with the synthetic trend in Indiana. However, the decline in Michigan is significant only at the 0.1 level rather than the typical 0.05 level, which may make it susceptible to Type 1 error. In other words, the observed difference in Michigan may appear to be an effect, but it is likely noise. Overall, the organizing outcomes show no difference between observed elections and union membership and what we might expect had those states not adopted right-to-work.

Figure 2.
Synthetic control methodology (SCM) results for organizing.

Table 1.
Fisher's Exact Test p-Values.

Outcome Indiana^a Michigan^a

Union Membership 0.24 0.08

Certification Elections 0.16 0.32

ULP Cases Filed 0.24 0.64

Successful ULP Cases 0.88 0.36

Collaboration Cases Filed 0.68 0.04

Successful Collaboration Cases 0.56 0.04*

a
P-value for Fisher's exact test for a difference in the mean square prediction error (MSPE) in pre-treatment and post-treatment at each outcome.

*p < .05.

In Figure 3, the total ULP cases are reported. Across both states and both outcomes, the observed trend follows the synthetic trend. Across these outcomes and locations, the difference between the trends is not statistically significant. Therefore, right-to-work appears not to affect total ULP case rates or success. This is unsurprising given the relationship between union elections and ULP filing and how elections kept high rates after adoption.

Figure 3.
Synthetic control methodology (SCM) for unfair labor practice.

There is also the question of party compositions and whether the haves come out ahead. In Figure 4, the rates and success of collaborations, or parties that include a federation on the charging party, are reported. This composition of charging party became more common in about 2008 and continues to be used after adoption when it was expected to decline. Particularly in Michigan, collaborations have become a more regular occurrence. Additionally, such cases in Michigan see higher rates of success over time. The differences observed in the difference in trend lines for Michigan for both collaboration outcomes are statistically significant (p-values for both are .04). Indiana, by contrast, does not have sustained differences after adoption and is not statistically significant in the differences in trends pre- and post-treatment.

Figure 4.
SCM for ULP collaborations.

Robustness Checks

To assess the credibility of these results, a placebo test was conducted as described by Abadie, Diamond, and Hainmueller (2015). In the placebo study, right-to-work adoption is iteratively assigned to each donor state to assess if the observed effects in Michigan and Indiana are unexpected relative to these placebo states. That is, the bolded line is said to be a robust effect if it appears far away from the mass of gray lines. In Figure 5, each gray line is the observed outcome run through the SCM structure for each other state. This graphical check reaffirms the suggested findings for collaborations and the lack of differences in the other outcomes. Collaboration filing and success appear higher than placebos in Michigan. This, combined with the findings in Table 1, provides support for the increase in rates and success of collaborations in Michigan.

Figure 5.
Synthetic control placebos Indiana.

Discussion and Conclusion

Right-to-work laws increase barriers to organizing labor throughout the United States. Yet, few have identified the consequences of right-to-work on union legal mobilization. This research has endeavored to uncover the changes in usage and success of ULP charges after right-to-work adoption. I have suggested that the dilemma of filing ULP for union provisions may have a new layer of complexity in a new right-to-work context: federations and unions look to support their constituents with fewer resources, and local sentiment around unions diminished in the legislature. Faced with these constraints, legal mobilization is one valuable resource for meeting the legal obligation of covering workers. The results from this work show that in a diminished power position for labor, unfair labor practice remains in use. The organizing of workers remains stable with regular election rates and membership. Success rates and usage also track the flat trend in union elections. However, underlying these flat trends is a shift in charging party characteristics that are distinct to each state. Collaborations between locals and federations see more prominence and success (Galanter 1974). This innovation in Michigan highlights the cost of legal mobilization as well. Legal mobilization with federation resources may be more common as part of a larger organizing effort (McCann 1994). These results show that right-to-work spurred more attention by federations to legal mobilization options.

The collaboration strategy suggested by the aggregate rates is a curious situation. After conducting this analysis, the collaborations were given a close reading to make sense of what was different about these cases that warranted additional charging party members. In some cases, what appears from the NLRB data are particular lawyers working for the federation or state council that have a task. One example is multiple charges filed as collaborations where a particular lawyer covers multiple cases at different supermarket chains over time. A similar structure appears in the automotive industry as well. In contrast to a non-collaborative charging party that hires a lawyer for a one-off legal mobilization, this collaboration strategy may be taking advantage of “repeat players” to aid union members. While this provides some anecdotal evidence of intentionality in union legal mobilization in certain industries, further work needs to be done to assess specific strategizing by federations in Michigan.

This work has aimed to show that right-to-work may impact union-related legal mobilization in two states. Michigan and Indiana were selected for their differences in labor power in each state. These differences are clear in the current right-to-work status of each state. In March 2023, the Michigan state legislature repealed its right-to-work law. This repeal aligns with the expectation that union legal mobilizations provided retentive wins while organizing for a more advantageous policy context elsewhere. The inclusion of legal mobilization efforts during a larger push for union power appeared to work for Michigan. The continued collaborations of federations and locals after adoption in Michigan point to an effort by federations to maintain union coverage while other political mobilization occurred. Labor in Indiana with a weaker political position was unable to conduct such a push, which aligns with Nissen and Churchill's (2020) suggestion that such legal and technical mobilizations would fail without one-to-one organizing campaigns. This insight may explain the drop-off in collaboration usage in Indiana. Altogether, the differences in collaborations in each state suggest different priorities for state-level federations and legal councils in different political environments.

The results showed a great deal of stability in the trends of legal mobilization and other union outcomes. Farber et al. (2021) did not find an impact of right-to-work on union density or elections for historic right-to-work adoptions in the 1940s and 1950s. This is curious since the expectation is that union density would fall with adoption. However, the findings of this research may be an illustration of what has been found by VanHeuvelen (2020). VanHeuvelen shows that right-to-work limits the power of unions to influence inequality. The continued usage of collaborations in Michigan highlights how right-to-work reprioritizes union activity. The union response to a law expected to decrease membership could draw focus away from the kinds of activity that bring wage premiums and lowered inequality traditionally related to union activity (Western and Rosenfeld 2011). Legal mobilization in Michigan after adoption illustrates one way that union power to influence other outcomes is diminished.

Right-to-work laws disrupt union organizing, and in response, unions may change their legal mobilization strategies. I used the synthetic control method to show that right-to-work changed the usage and success of ULP charges intended to maintain union provisions in Indiana and Michigan. Hartley (1948), author of the amendment to the National Labor Relations Act that allowed states to enact right-to-work laws, was convinced that ULP filings by employees would not change as a result of right-to-work. I found this was generally the case. This research shows that the affirmative protections offered by the NLRA may become increasingly layered and costly to unions in a new right-to-work context. In comparison, the rates did not change, the associated strategy around them did, which is emblematic of what right-to-work seeks to accomplish. Union strategy changes under right-to-work in ways that may be subtly disrupting union activity.

Many scholars have pointed out the need for innovation and creativity to maintain labor unions and collective bargaining more broadly in the United States (Dixon 2008; Bronfenbrenner 2009; Milkman and Luce 2017). My findings support the value suggested by these authors through the innovative collaborations, which were virtually unused for much of the 2000s in the particular case of charges in Michigan and made more clear by the contrasting lack of support in Indiana as right-to-work was adopted. However, these findings also emphasize the active responses by labor that are often assumed to be absent in right-to-work studies. This research shows that innovation in legal mobilization may constitute a means that workers can continue to organize under a disadvantageous policy.

Supplemental Material

sj-docx-1-lsj-10.1177_0160449X231218982 - Supplemental material for The Effect of Right-to-Work on Unfair Labor Practice Charges: Synthetic Control Evidence From Indiana and Michigan

Supplemental material, sj-docx-1-lsj-10.1177_0160449X231218982 for The Effect of Right-to-Work on Unfair Labor Practice Charges: Synthetic Control Evidence From Indiana and Michigan by Breon Haskett in Labor Studies Journal

Outcome	Indiana^a	Michigan^a
Union Membership	0.24	0.08
Certification Elections	0.16	0.32
ULP Cases Filed	0.24	0.64
Successful ULP Cases	0.88	0.36
Collaboration Cases Filed	0.68	0.04*
Successful Collaboration Cases	0.56	0.04*

Footnotes

Acknowledgements

The author is indebted to Ben Glasner, Peter Catron, and the Harry Bridges Center for Labor Studies.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Partial support for this research came from a Shanahan Endowment Fellowship and a Eunice Kennedy Shriver National Institute of Child Health and Human Development training grant, T32 HD101442-01, to the Center for Studies in Demography & Ecology at the University of Washington. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Eunice Kennedy Shriver National Institute of Child Health and Human Development (grant number T32 HD101442-01).

ORCID iD

Breon Haskett

Supplemental Material

Supplemental material for this article is available online.

Author Biography

Breon Haskett is a PhD student at the University of Washington Department of Sociology. His current research interests focus on organizational behaviors that levy inequality in wage attainment and population processes. He is a NIH T32 Data Science and Demography Training Fellow at the Center for Studies in Demography and Ecology at the University of Washington.

References

Abadie

Alberto.

2021. “Using Synthetic Controls: Feasibility, Data Requirements, and Methodological Aspects.” Journal of Economic Literature 59 (2): 391‐425. https://doi.org/10.1257/jel.20191450

Abadie

Alberto

Diamond

Alexis

Hainmueller

Jens

. 2015. “Comparative Politics and the Synthetic Control Method.” American Journal of Political Science 59 (2): 495‐510. https://doi.org/10.1111/ajps.12116

Babson

Steve.

1986. Working Detroit: The Making of a Union Town. Detroit: Wayne State University Press.

Bronfenbrenner

Kate.

2009. “No Holds Barred: The Intensification of Employer Opposition to Organizing.” Economic Policy Institute Briefing Paper # 235.

Brooks

Clem

Manza

Jeff

. 2007. Why Welfare States Persist: Public Opinion and the Future of Social Provision. Chicago: University of Chicago Press.

Bruno

Robert

Zullo

Roland

Manzo

Frank

Dickson

Alison

. 2015. “The Economic Effects of Adopting a Right-to-Work Law.” Labor Studies Journal 40 (4): 319‐61. https://doi.org/10.1177/0160449X15619539

Bureau of Labor Statistics. 2019. Union Members in Michigan – 2019. Washington, DC: BLS News Release.

Cantin

Étienne.

2012. “The Politics of Austerity and the Conservative Offensive Against US Public Sector Unions, 2008–2012.” Relations Industrielles 67 (4): 612‐32. https://doi.org/10.7202/1013197ar

Carson

Jenny.

2010. “‘Taking on Corporate Bullies’: Cintas, Laundry Workers, and Organizing in the 1930s and Twenty-First Century.” Labor Studies Journal 35 (4): 453‐79. https://doi.org/10.1177/0160449X10379642

10.

Clawson

Dan

Clawson

Mary Ann

. 1999. “What Has Happened to the US Labor Movement? Union Decline and Renewal.” Annual Review of Sociology 25 (1): 95‐119. https://doi.org/10.1146/annurev.soc.25.1.95

11.

Collins

Benjamin.

2013. “Right to Work Laws: Legislative Background and Empirical Research.” Right to Work Laws and Federal Labor Relations Statutes: An Overview 1–19.

12.

Dannin

Ellen J.

2006. Taking Back the Workers’ Law: How to Fight the Assault on Labor Rights. Ithaca, NY: ILR Press/Cornell University Press.

13.

Devinatz

Victor.

2011. “The Continuing Controversy Over Right-to-Work Laws in the Early Twenty-First Century.” Employee Responsibilities and Rights Journal 23 (4): 287‐93. https://doi.org/10.1007/s10672-011-9185-z

14.

Devinatz

Victor G.

2015. “Right-to-Work Laws, the Southernization of U.S. Labor Relations and the U.S. Trade Union Movement’s Decline.” Labor Studies Journal 40 (4): 297‐318. https://doi.org/10.1177/0160449X15622702

15.

Dixon

2008. “Movements, Countermovements and Policy Adoption: The Case of Right-to-Work Activism.” Social Forces 87 (1): 473‐500. https://doi.org/10.1353/sof.0.0076

16.

Dixon

2010. “Union Threat, Countermovement Organization, and Labor Policy in the States, 1944–1960.” Social Problems 57 (2): 157‐74. https://doi.org/10.1525/sp.2010.57.2.157

17.

Dixon

2020. Heartland Blues. New York: Oxford University Press.

18.

Dixon

Martin

. 2012. “We Can’t Win This on Our Own: Unions, Firms, and Mobilization of External Allies in Labor Disputes.” American Sociological Review 77 (6): 946‐69. https://doi.org/10.1177/0003122412460649

19.

Donohue

John

Aneja

Abhay

Weber

Kyle

. 2019. “Right to Carry Laws and Violent Crime: A Comprehensive Assessment Using Panel Data and a State-Level Synthetic Control Analysis.” Journal of Empirical Legal Studies 16 (2): 198‐247. https://doi.org/10.1111/jels.12219

20.

Dunford

2021. tidysynth: A Tidy Implementation of the Synthetic Control Method. R package version 0.1.0. https://CRAN.R-project.org/package=tidysynth

21.

Ellwood

David T.

Fine

Glenn

. 1987. “The Impact of Right-to-Work Laws on Union Organizing.” Journal of Political Economy 95 (2): 250-73. https://doi.org/10.1086/261454

22.

Eren

Ozkan

Ozbeklik

Serkan

. 2016. “What Do Right-to-Work Laws Do? Evidence from a Synthetic Control Method Analysis” Journal of Policy Analysis and Management 35 (1): 173‐94. https://www-jstor-org-443.web.bisu.edu.cn/stable/43866565

23.

Fantasia

Rick.

1989. Cultures of Solidarity: Consciousness, Action and Contemporary American Workers. Berkeley, Calif.: University of California Press.

24.

Farber

Henry S.

Herbst

Daniel

Kuziemko

Ilyana

Naidu

Suresh

. 2021. Unions and Inequality Over the Twentieth Century: New Evidence from Survey Data.” Quarterly Journal of Economics 136 (3): 1325-85. https://doi.org/10.1093/qje/qjab012

25.

Feigenbaum

James

Hertel-Fernandez

Alexander

Williamson

Vanessa

. 2018. “From the Bargaining Table to the Ballot Box: Political Effects of Right to Work Laws.” NBER Working Paper No. 24259: 70-70.

26.

Ferguson

John-Paul.

2008. “The Eyes of the Needles: A Sequential Model of Union Organizing Drives 1999–2004.” ILR Review 62 (1): 3‐21. https://doi.org/10.1177/00197939080620010

27.

Galanter

Marc.

1974. “Why the ‘Haves’ Come Out Ahead: Speculations on the Limits of Legal Change.” Law & Society Review 9 (1): 95‐160. https://heinonline.org/HOL/P?h=hein.journals/lwsocrw9&i=97

28.

Galiatsos

Caroline B.

2015. “Beyond Joint Employer Status: A New Analysis for Employers’ Unfair Labor Practice Liability Under the NLRA.” Boston University Law Review 95: 34. https://heinonline.org/HOL/P?h=hein.journals/bulr95&i=2121

29.

Hartley

Fred

. 1948. Our New National Labor Policy: the Taft-Hartley Act and the next steps. New York: William S Hein & Company.

30.

Hogler

Raymond L.

2015. The End of American Labor Unions: The Right-to-Work Movement and the Erosion of Collective Bargaining. Santa Barbara, California: Praeger.

31.

Hogler

Raymond

Shulman

Steven

Weiler

Stephan

. 2004. “Right-to-Work Legislation, Social Capital, and Variations in State Union Density.” Review of Regional Studies 34 (1): 95‐111. https://doi.org/10.52324/001c.8371

32.

Holmes

Thomas.

1998. “The Effect of State Policies on the Location of Manufacturing: Evidence from State Borders.” Journal of Political Economy 106 (4): 667‐705. https://doi.org/10.1086/250026

33.

Kahlenberg

Richard

Greene

Jay

. 2012. “Unions and the Public Interest” Educ Next, Winter 2012.

34.

Kaminski

Michelle.

2015. “How Michigan Became a Right to Work State: The Role of Money and Politics.” Labor Studies Journal 40 (4): 362‐78. https://doi.org/10.1177/0160449X15619542

35.

Keele

Minozzi

. 2013. “How Much is Minnesota Like Wisconsin? Assumptions and Counterfactuals in Causal Inference with Observational Data.” Political Analysis 21 (2): 193‐216. https://doi.org/10.1093/pan/mps041

36.

Kimeldorf

Howard.

2013. “Worker Replacement Costs and Unionization: Origins of the U.S. Labor Movement.” American Sociological Review 78 (6): 1033‐62. https://doi.org/10.1177/0003122413509627

37.

Kogan

Vladimir.

2016. “Do Anti-Union Policies Increase Inequality? Evidence from State Adoption of Right-to-Work Laws.” State Politics & Policy Quarterly 17 (2): 180‐200. https://doi.org/10.1177/1532440016677217

38.

Lichtenstein

Nelson.

1998. “Taft-Hartley: A Slave-Labor Law?” Catholic University Law Review 47: 763‐89. https://heinonline.org/HOL/P?h=hein.journals/cathu47&i=779

39.

Lichtenstein

Nelson.

2002. State of the Union. Princeton, NJ: Princeton University Press.

40.

Manzo

Frank

IV . 2015. “Promises Unfulfilled: Right-to-Work’s Early Economic Track Record in Indiana.” Labor Studies Journal 40 (4): 379‐95. https://doi.org/10.1177/0160449X15617405

41.

McCammon

Holly

. 2001. “Labor's Legal Mobilization.” Work and Occupations 28 (2): 143-75. https://doi.org/10.1177/0730888401028002002

42.

McCann

Michael.

1994. Rights at Work. Chicago, IL: University of Chicago Press.

43.

McClelland

Robert

Gault

Sarah

. 2017. The Synthetic Control Methods as a Tool to Understand State Policy. Washington, DC: Urban Institute Research Reports.

44.

McDowell

Douglas

Huhn

Kenneth

McGuiness

Kevin

Janus

Peter

. 1976. NLRB Remedies for Unfair Labor Practices. Philadelphia, PA: University of Pennsylvania Press.

45.

Milkman

Ruth

Luce

Stephanie

. 2017. “Labor Unions and the Great Recession.” RSF: The Russell Sage Foundation Journal of the Social Sciences 3 (3): 145. https://doi.org/10.7758/rsf.2017.3.3.07

46.

Minchin

Timothy.

2017. Labor Under Fire: A History of the AFL-CIO Since 1979. The University of North Carolina Press.

47.

Moore

William J.

1998. “The Determinants and Effects of Right-to-Work Laws: A Review of Recent Literature.” Journal of Labor Research 19 (3): 444‐69. https://doi.org/10.1007/s12122-998-1041-z

48.

Nissen

Bruce

Churchill

Candi

. 2020. “Unionism in a Right-to-Work Environment: United Faculty of Florida from Stagnation to Crisis Mobilization to Power Building.” Labor Studies Journal 45 (4): 331‐426. https://doi.org/10.1177/0160449X20911710

49.

Rao

Hayagreeva

Yue

Lori Qingyuan

Ingram

Paul

. 2011. “Laws of Attraction.” American Sociological Review 76 (3): 365-85. https://doi.org/10.1177/0003122411409698

50.

Rosenfeld

Jake.

2014. What Unions No Longer Do. First Edition. Cambridge, Massachusetts; London, England: Harvard University Press.

51.

Semett

Amy.

2016. “Political Decision-Making at the National Labor Relations Board: An Empirical Examination of the Board’s Unfair Labor Practice Decisions through the Clinton and Bush II Years.” 71.

52.

Smith

Mike.

2001. “‘Let’s Make Detroit a Union Town’: The History of Labor and the Working Class in the Motor City.” The Michigan Historical Review 27 (2): 157-73. https://doi.org/10.2307/20173931

53.

Snead

Warren.

2022. “The Supreme Court as an Agent of Policy Drift: The Case of the NLRA.” American Political Science Review, 1‐14. https://doi.org/10.1017/S0003055422000685

54.

Trumka

Richard.

2019. “Testimony of Richard L. Trumka on Deterring Unfair Labor Practices and the Protecting the Right to Organize Act.” Retrieved May 6, 2021.

55.

VanHeuvelen

Tom.

2020. “The Right to Work, Power Resources, and Economic Inequality.” American Journal of Sociology 125 (5). https://doi.org/10.1086/708067

56.

Vedder

Richard.

2011. “The Economic Effects of Right to Work Laws.” Employee Responsibilities and Rights Journal 23 (4): 305‐10. https://doi.org/10.1007/s10672-011-9182-2

57.

Weiner

Michael.

2005. “Can the NLRB Deter Unfair Labor Practices? Reassessing the Punitive-Remedial Distinction in Labor Law Enforcement.” UCLA Law Review 49: 1579-626. https://heinonline.org/HOL/P?h=hein.journals/uclalr52&i=1593

58.

Western

Bruce

Rosenfeld

Jake

. 2011. “Union, Norms and the Rise in U.S. Wage Inequality.” American Sociological Review 76 (4): 513‐37. https://doi.org/10.1177/0003122411414817

59.

2017. “Generalized Synthetic Control Method: Causal Inference with Interactive Fixed Effects Models.” Political Analysis 25 (1): 57‐76. https://doi.org/10.1017/pan.2016.2

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

6.96 MB