Union Contributions to the Historic Shift in Workplace Governance
At the same time that union density in the United States has declined (Figure A) and labor law has “ossified” and “drifted”, employment law has flourished (Figure B), proliferating at the state level and expanding into new substantive domains (Source: Galvin 2019).
For the vast majority of 21st century workers, what workplace rights and protections remain come not from labor law and the mechanism of collective bargaining (as the designers of the National Labor Relations Act of 1935 had hoped) but from employment laws and the mechanisms of regulation and litigation.
Employment laws establish minimum workplace standards (like the minimum wage and health and safety rules), which are enforced by state agencies, and provide individual rights and protections (like privacy rights and protection against discrimination), which are adjudicated in court. Labor law, in contrast, deals with collective bargaining, unionization, and related issues that arise between groups of workers and their employers.
Unfortunately, when it comes to protecting workers’ rights, employment laws are clearly inferior to labor law. They are notoriously under-enforced by state regulatory agencies, ignored by employers, invisible to many workers, and often circumvented by mandatory arbitration clauses in employment contracts. In contrast to labor law, they lack built-in mechanisms for generating collective action, solidarity, and expressing voice in the workplace. And because those with information and resources have greater access to employment law remedies—and due to a judicial process that exacts high costs on those plaintiffs who do file cases—they tend to exacerbate and reproduce racial, gender, and income inequalities among affected workers.
And yet, workers still receive valuable protection from employment laws. Until labor law is overhauled, they constitute workers’ last line of defense against exploitative employers.
The shift from labor law to employment law as the central locus of workplace governance therefore represents an historic shift in the primary institutions constituting workers’ rights—one at least as consequential as the advent of collective bargaining in the 1930s, if not more so—given the larger proportion of workers affected and the profound impact it has had on the substance and form of workers’ rights.
How, then, did this institutional shift in workplace governance occur?
Some scholars have suggested a link between the decline of labor unions and the rise of employment law. The eminent legal scholar Theodore St. Antoine, for example, posits that “part of the growth we have seen in employment law, as distinguished from labor law, is attributable to the decline of organized labor. Government has had to step in to fill the vacuum.”
The notion that state legislatures turned to employment laws to protect an increasingly non-unionized, vulnerable workforce has a certain intuitive appeal. But the role of organized labor remains unclear. Did labor unions sit idly by while state legislatures built new workers’ rights? Or did unions, even as they declined, push for greater protections and rights for all workers? Or did the growth of employment law actually contribute to the decline of unions, as some economists have suggested?
We know that labor unions have historically been strong advocates for universal social welfare programs and redistributive policies—but their interest in vigorous state regulation of the workplace is less clear. From Gompersian voluntarism to the “government substitution hypothesis,” there has long been a question of whether state intervention helps or hurts unions—in particular, whether protective employment laws might undermine workers’ incentive to unionize by providing for free what workers might otherwise get through their unions.
Interestingly, there isn’t a lot of empirical evidence to support this idea. As Richard Freeman has shown, there’s actually more evidence that causality works in the reverse direction, such that “more highly unionized states are…more likely to pass protective legislation.”
The problem is that we still do not know in which way the causal arrows point. It could be that states that are more likely to enact stronger employment laws are also more likely to have stronger unions, and there’s little direct relationship between the two.
In my recent paper published in the ILR Review, I try to dig a little deeper to flesh out the nature of the relationship. I adopt an integrated multi-method approach in which quantitative and qualitative analyses are used in a complementary fashion to move us incrementally closer toward drawing causal inferences with greater confidence.
A basic regression model incorporating known theory serves as the starting point; the residuals from the model are used to select “deviant” cases for in-depth analysis; and systematic process tracing then serves to explore hypothesized causal pathways, validate measurement, and identify potential omitted variables. In this approach, quantitative and qualitative analyses do not redundantly demonstrate the same outcome through different methods; rather, the strengths of each are leveraged to compensate for the weaknesses of the other. The explicit goal is not to resolve the causal inference puzzle definitively, but to make piecemeal progress toward clarifying causal relationships and building theory.
Although the resulting inferences are circumspect by design, the implications of the findings reach more broadly, to normative and empirical questions as old as the American labor movement itself. Does protective labor legislation complement or substitute for unionization? How compatible is trade unionism with political engagement? When and how do unions transcend the imperatives of organizational maintenance and advocate for all working people? In contemporary debates over how to best secure workers’ rights and redress power imbalances in the workplace, these are pressing questions.
Using more fine-grained data than Freeman had available to him, I ran a number of statistical tests which confirmed his basic finding. States with higher rates of union density do boast more robust employment law regimes: they maintain higher standards for work conditions and offer workers a wider range of substantive rights and protections. And while unionism predicts employment law enactments, employment laws do not appear to cause or accelerate union decline: All else equal, the higher the rate of union density and the less decline over time, the greater the state’s propensity to enact employment laws.
Using the residuals from one of the statistical models to select “deviant” cases – cases that are least-well explained by the model (as a rule of thumb, greater than two standard deviations away from the mean) – the model identifies Pennsylvania and Maine as good cases for further examination. Pennsylvania has many fewer major employment laws than the model would predict, given its high level of union density and other relevant factors, and Maine has many moreemployment laws than predicted, given its merely average rate of union density.
I then systematically examined hundreds of newspaper articles covering any and all campaigns in each state relating to the minimum wage, family and medical leave, discrimination, and employee misclassification over the last few decades. Using standard process tracing methods, I ask of the evidence: 1) Were unions integral to the policy process? And, 2) how integral were they? Were they considered key actors in all legislative campaigns, or just some? How much leadership did they exert? What tactics did they employ? In other words, what does the causal pathway actually look like, if one exists?
The case studies turned up voluminous evidence that unions did indeed play an integral role in campaigns to enact stronger subnational employment laws. Even when bills failed to pass or were ultimately watered down, organized labor was consistently on the front lines in both states, pushing for stronger protections for workers. (Except in efforts to expand LGBTQ rights in Pennsylvania, where organized labor was nowhere to be found—serving as a reminder of unions’ mixed history on issues of diversity and inclusion.)
My integrated mixed-method approach does not provide dispositive empirical evidence of a causal relationship or fully refute the notion that the relationship may yet be spurious in other cases. But the evidence assembled is difficult to square with alternative causal arguments and it lends concrete support to the hypothesized causal pathway connecting labor unions to employment laws. Although unions did not build these employment laws alone—multiple forces clearly played a role in each case—unions ought to be considered consistent contributing causes of their growth.
Even as union membership continued its seemingly inexorable decline and labor’s myriad positive “spillover” and “threat effects” vanished—along with their once-buoyant effects on the moral economy, as Bruce Western and Jake Rosenfeld have shown—labor unions were clearly working hard to enact many of the subnational employment laws that continue to structure and regulate the employment relationship today.
Case evidence suggests that many unions were motivated to fight for stronger employment laws for a variety of nonexclusive reasons, including their abiding commitment to advocate for all working people, not just union members, as Ahlquist and Levi (2014) have shown. Indeed, in most of the policy campaigns I examined, unions publicly trumpeted their role as guardians of all workers’ well-being. Future research should explore whether this public stance, and their ability to achieve policy victories, has been advantageous for these unions, perhaps resulting in increased membership or more favorable views of unionism generally. This is a question that has loomed over the Fight for $15 (‘‘and a union’’) movement: When, whether, and how do employment law campaigns complement unionization?
The contribution of this analysis is that it has allowed us to more confidently point the causal arrows in the right direction and move on to the next stage of analysis. We can now dig deeper into the role unions have played, and continue to play, in building the primary sources of workers’ rights and protections in the 21st century, inadequate though they may be. And we can state with greater confidence that the new federalism in work regulation constitutes one of organized labor’s most significant and enduring legacies.