THE DEVELOPMENT AND DECLINE OF LABOR LAW

The rise and fall of U.S. labor unions and the ossification of labor law are topics that have been so carefully discussed by legal scholars, historians, sociologists, political scientists, and activists that in retreading the same ground, one runs a high risk of redundancy, committing sins of omission, or both.Footnote ²⁶ The following sketch therefore prioritizes brevity, emphasizing only the dynamics most relevant for the ensuing analysis.

The National Labor Relations Act (NLRA or Wagner Act), enacted at the height of the New Deal in 1935, remains the primary federal law governing labor relations in the United States. Its stated purpose is to redress “the inequality of bargaining power” in the workplace by protecting the right of workers “to organize and bargain collectively” over the “terms and conditions of their employment.” If left unaddressed, the statue notes that this inequality would produce a “diminution of employment and wages,” decrease the “purchasing power of wage earners,” put downward pressure on working conditions, and threaten the flow of commerce.Footnote ²⁷ Once the Supreme Court affirmed the constitutionality of the Act in 1937, it began to generate strong feedback effects on interest groups, government elites, and mass publics. Indeed, it offers a model case of how “new policies create a new politics.”Footnote ²⁸

Its most significant feedback effects can be observed in the development of private-sector labor unions. Many unions, of course, were already in existence and instrumental in securing the law's passage in the first place.Footnote ²⁹ But by formally recognizing unions as workers’ representatives and establishing procedures for union elections, the Wagner Act accelerated their growth and influenced their organizational development.Footnote ³⁰ The share of nonagricultural private-sector workers belonging to a union more than doubled between 1936 and 1938, from 13 to 27 percent, and did not dip below a quarter of the workforce again until 1975. Despite the challenges posed by the Taft-Hartley amendments of 1947, union density remained above 30 percent for the seventeen-year stretch between 1943 and 1960 (Figure 1).

Sources: 1880–1995 data are from Richard B. Freeman, “Spurts in Union Growth: Defining Moments and Social Processes,” No. W6012 (Cambridge, MA: National Bureau of Economic Research, 1997); 1996–2017 data are from Barry T. Hirsch and Macpherson, “Union Membership and Coverage Database from the Current Population Survey: Note,” Industrial and Labor Relations Review 56, no. 2 (2003): 349–54 (updated annually at www.unionstats.com).

Fig. 1. Union Density, 1880–2017.

The collective bargaining system established through national labor law also had profound feedback effects on workers’ and employers’ political capacities. As Theda Skocpol writes, policies can “affect the capabilities of various groups to achieve self-consciousness, organize, and make alliances.”Footnote ³¹ Labor law encouraged workers to view their interests as collective interests that could be pursued through self-organization and collective action.Footnote ³² Indeed, as the collective bargaining system matured, the range of imaginable alternatives for reform narrowed: As Nelson Lichtenstein, Ira Katznelson, and others have shown, by the early postwar period, more radical possibilities had been eliminated from the realm of possibility, and the collective bargaining regime had become largely taken for granted.Footnote ³³ Both workers’ advocates and business interests structured their political activities around the new system, seeking to gain sufficient political power to control the National Labor Relations Board (NLRB) and shape its common law.Footnote ³⁴ In these ways and more, the Wagner Act “made” labor politics in America.Footnote ³⁵

As a legal matter, the NLRA has long maintained exclusive governing authority over an extremely broad domain. In 1959, the Supreme Court ruled in San Diego Building Trades Council v. Garmon that the NLRA barred any and all state and local efforts to legislate in the areas of labor law it covered, preempting regulation in areas even “arguably” protected or prohibited by the NLRA.Footnote ³⁶ Unlike the Fair Labor Standards Act (FLSA) of 1938, which allowed states to enact stronger protections and set higher minimum standards than federal law provided, the “Garmon doctrine” prohibited state efforts to do anything similar in the field of labor law. Subsequent Court decisions narrowed the ability of states and localities to intervene further still—for example, the Machinists case (1976) ruled that subnational regulation was preempted even in areas Congress ostensibly meant to leave to “the free play of economic forces,” thereby “virtually banish[ing] states and localities from the field of labor relations,” with precious few exceptions.Footnote ³⁷ States were permitted to design their own labor laws only for those workers excluded from coverage under the Wagner Act, such as public-sector and agricultural workers. If they sought to provide stronger rights and protections to a wider range of workers, they would need to find alternative routes for doing so; the Wagner Act was to be the primary, authoritative, centralized labor law in the United States that would enjoy an effective monopoly on the process of collective bargaining and unionization for the vast majority of workers.

Although the Taft-Hartley Act of 1947, the Landrum-Griffin Act of 1959, and numerous business-friendly Court decisions and NLRB rulings over the years chipped away at unions’ power, undermined workers’ right to strike, and generally enhanced employers’ prerogatives, they also served to further ratify the Wagner Act's exclusive governing authority over private-sector labor-management relations.Footnote ³⁸ Indeed, these instances of what Kathleen Thelen might term “conversion” (the redirection of existing rules to advantage actors pursuing different purposes) did not alter the basic structure of the law or disturb its core features or mechanisms, which remain essentially as they were in 1935.Footnote ³⁹

Consequently, despite its formal institutional persistence, national labor law became an ever more glaring “mismatch” with changing economic conditions, including the changing composition of industries, new patterns of capital investment, and the changing nature of work.Footnote ⁴⁰ Amid the “bewilderingly complex proliferation of employment relationships that structure work in the modern city, college, or company” featuring “layer after layer of subcontractors and vendors,” labor law's manifold weaknesses became all the more glaring.Footnote ⁴¹ As fewer and fewer workers were able to take advantage of its procedures for establishing rights, representation, and voice in the workplace, employees in all industries became more vulnerable to exploitation and abuse. Employment became more precarious, income inequality widened, wages stagnated, and the power asymmetry in the workplace grew.

But this was no accident or oversight: Despite attempts by labor unions and their allies to update the law to better reflect changing economic conditions—by streamlining union election procedures (“card check”), increasing penalties for employer interference, eliminating the NLRA's preemption of state labor laws to allow experimentation, and more—opponents (chiefly business-funded Republicans but also some Democrats) successfully blocked reforms from advancing through the legislative process.Footnote ⁴²

As Kate Bronfenbrenner and Tom Juravich have shown, employers benefited greatly from the status quo. Anti-union employers regularly exploited labor law's loopholes and blind spots and even blatantly violated the law in order to deter and quash unionization drives. They used “a combination of legal and illegal approaches” such as illegally discharging workers for participating in union activity, “captive-audience meetings, employer leaflets and mailings, supervisor one-on-ones, and illegal wage increases.”Footnote ⁴³

Labor's opponents also became more politically engaged over the years, developing a multipronged approach to promoting their interests and curtailing workers’ rights. As Alexander Hertel-Fernandez has shown, the business community began in the 1970s to build a “durable ‘conservative-corporate’ subnational coalition” and a formidable organizational infrastructure to advance its legislative aims, “systematically changing the state policy landscape in ways that disadvantaged their political opponents over the long run.”Footnote ⁴⁴ For example, the pro-business American Legislative Exchange Council (ALEC) developed “model bills,” many explicitly aimed at “weakening labor unions,” and was often successful, especially in states with low policy capacity and high ideological conservatism.Footnote ⁴⁵

To be sure, many additional factors contributed to the growing inability of labor law to provide workers’ rights and protections to the vast majority of workers. These include the conservatism and mismanagement of union leaders, the changing attitudes of workers, the “changing relation of class forces,” and more.Footnote ⁴⁶ Studies have also emphasized how institutional and ideational abrasions between the emergence of civil rights consciousness, activism, and enforcement, on the one hand, and labor rights and collective commitments, on the other, contributed to the enervation of labor law and the labor movement more broadly.

For example, as Paul Frymer has shown, as a result of a “bifurcated system of power that assigned race and class problems to different spheres of government,” the enforcement of civil rights laws “unintentionally weakened national labor law,” ultimately producing a more “diverse but weakened labor movement.”Footnote ⁴⁷ Nelson Lichtenstein has similarly argued that it was the unsustainable ideational conflict between “the concept of rights and that of solidarity, between the civil rights consciousness that has proven so potent during the last half century and the collective institutionalism that stands at the heart of the union idea” which contributed to labor's decline. “As the former became a near hegemonic way of evaluating the quality of American citizenship,” he summarized, “the latter atrophied. Indeed, in its most extreme interpretation, rights consciousness subverts the mechanisms, both moral and legal, that sustain the social solidarity upon which trade unionism is based.”Footnote ⁴⁸

The causes of labor's decline are thus multifaceted and complex, and its consequences are equally multiform. But for the purposes of the present analysis, two key features of labor law and its decline stand out as especially important. First, even as labor law became increasingly ineffective, its stubborn persistence, broad authority, and preemption of subnational innovation limited the options of workers’ rights activists and reformers. Efforts to promote and defend workers’ rights were (and continue to be) formulated in light of those limits; reformers must either work within or around an immobile national labor law. By cordoning off potential pathways for reform, labor law has powerfully shaped the substantive content and institutional form of the alternative policy instruments that have emerged, as I will elaborate further below.

Second, labor law's ossification and drift altered the balance of power among organized interests. Those with the greatest stake in matters of labor-management relations did not disappear, but rather came to embrace qualitatively different kinds of politics. Business interests—the “losers” in the New Deal—gradually shifted from offense to defense, and their main task became preserving the legal status quo. In the absence of formal policy change, anti-union employers could continue to exploit labor law's weaknesses to make unionization more difficult while letting changes in the broader economic context further enervate and fractionalize workers, undermining what remained of the labor movement.

Meanwhile, labor unions and their allies—“winners” in the New Deal but increasingly “losers” in the post-Taft-Hartley period—gradually shifted from defense to offense (some say too gradually).Footnote ⁴⁹ Organizationally mature, resource-rich, and politically experienced, many labor unions responded to the growing problems facing workers by developing innovative organizing strategies and constructing new policy instruments in alternative venues to achieve many of the same purposes as labor law, albeit through different means and mechanisms.Footnote ⁵⁰ And as new nontraditional workers’ advocacy groups gradually emerged in this organizational space to meet the needs of an increasingly fragmented and vulnerable workforce, their goals dovetailed with unions’ goals, but their approach was quite different (in part because as nonprofit, non-union groups, it legally had to be): These new organizations experimented with creative strategies to help those workers left behind by labor law's ossification while accommodating the realities of the changing legal and economic landscape. Many of the new and emergent tactics in the labor movement have thus evolved as responses to old but persistent institutional constraints and the growing obsolescence of old organizing tactics. As I will discuss further below, synergies, conflicts, and ongoing negotiations between allies new and old now constitute the labor movement.

To fully understand the political development of workers’ rights—past, present, and future—these complex institutional and organizational dynamics must be disentangled and analyzed more closely. A severely delimited institutional landscape channeled and constrained the responses of both workers and business while flipping the balance of power between them. The new institutions, organizations, and strategic repertoires that have emerged in this space have, consequently, durably altered the politics of workers’ rights. This new politics did not emerge on a blank slate, but was rather spawned from old configurations, with old problems and purposes insinuated into the new formations. New and old now operate simultaneously, sometimes in harmony and sometimes in conflict, generating complex incentives, constraints, and challenges.

THE “CHANGING OF THE GUARD” FROM LABOR LAW TO EMPLOYMENT LAW

In 1988, renowned law professor Clyde Summers observed a “changing of the guard” from labor law to employment law:

Summers's rather functionalist perspective notwithstanding, the notion that there has been a historic shift in the primary “guardian” of workers’ rights from labor law to employment law has been confirmed and elaborated by numerous legal scholars.Footnote ⁵² In her influential work, for example, Katherine Van Wezel Stone writes that “the emerging regime of individual employee rights represents not a complement to or an embellishment of the regime of collective rights, but rather its replacement.”Footnote ⁵³ Cynthia Estlund likewise affirms that “the New Deal collective bargaining system has been supplemented, and largely supplanted, by other models of workplace governance: a regulatory model of minimum standards enforceable mainly by administrative agencies and a rights model of judicially enforceable individual rights.”Footnote ⁵⁴ A number of other scholars, including Kate Andrias, Nancy MacLean, and Benjamin Sachs, have helpfully brought workers back in to the story, with Sachs vividly describing this shift as a “hydraulic” process driven by workers themselves:

Federal employment laws represent only the tip of the iceberg, however—as several scholars have observed, most of the action has actually been at the state level. Richard Bales, for example, finds that “in addition to enacting state statutes that parallel the federal statutes … state legislatures have passed legislation protecting employees in a wide variety of other circumstances,” including in areas as diverse as wrongful discharge, whistleblowing, employee testing, and workplace safety and health.Footnote ⁵⁶ Although Theodore St. Antoine's focus is on state courts’ role in reshaping the doctrine of at-will employment, he also keenly observes that “the most important and dramatic development in employment law over the last couple of decades came at the state level, not the federal.”Footnote ⁵⁷

In other words, the fact that employment law has grown in volume, reach, and importance has been noted by attentive scholars. But to date, no empirical evidence has been systematically marshaled to substantiate the claim. Nor, consequently, has any existing work been able to analyze its substantive complexity or begin to hazard an explanation for its variation across space and time. Thus, while we think we know employment law has grown, we do not know what it consists of, where it has grown most, or who has pushed for it. For example, St. Antoine's claim that “part of the growth we have seen in employment law, as distinguished from labor law, is attributable to the decline of organized labor,” is an empirical question that we have yet to answer.Footnote ⁵⁸

Part of the reason for this empirical lacuna is practical, given that tens of thousands of new laws are enacted every year at the state level, and systematically obtaining reliable, equivalent information on the many different types of employment laws passed in fifty states over many decades appears daunting. Luckily, the Department of Labor has continuously monitored, recorded, categorized, and summarized employment law enactments at the state level every year for many decades. Until recently, expert staff members wrote and published summary reports of all laws passed during the prior year along with short descriptions of each in the Bureau of Labor Statistics’ Monthly Labor Review (MLR).Footnote ⁵⁹ The format and authorship of the reports were remarkably stable over time, with more than 90 percent of the reports issued between 1960 and 2014 authored by the same five individual authors, with some overlap between them.

With a small team of research assistants, I used these reports to construct a data set of every legislative enactment between 1960 and 2014, coded according to the categories listed in the reports (“wages,” “hours of work,” “plant closings,” “child labor,” “whistleblower,” and so on).Footnote ⁶⁰ During this period, states enacted 7,336 employment laws in thirty-three categories, and the relative attention paid by state legislatures to employment law grew steadily over the years as well. As Figure 2 indicates, across all states, employment laws as a share of all enacted bills more than quintupled over those fifty-five years. At precisely the same time that national labor law atrophied and private-sector union membership plummeted, state legislatures became increasingly preoccupied with employment law.

Sources: Employment law counts are from the Bureau of Labor Statistics, “State Labor Legislation Enacted in [Previous Year].” Total legislative enactments in each state each year (including overrides, minus vetoes) are from the Book of the States. American Legislators’ Association Council of State Governments, The Book of the States (Lexington, KY: Council of State Governments, 1935–2017).

Fig. 2. Employment Laws as a Share of All Enacted Laws by Session.

Of course, new laws are not enacted in a vacuum. Many build on extant laws, addressing weaknesses that have become apparent in the course of events, and are often designed with the state's existing administrative capacities in mind. Once a given state passes a pay equity law, for example, it need not pass another law to establish the same standard, but it may develop additional laws to bolster the enforcement of the primary law, add new categories of covered occupations, and so on. Many new laws thus build upon previous laws and alter their operation in ways both big and small. Figure 2 should therefore be interpreted as an aggregate count of laws that are actually often cumulative in nature. Still, as discussed further below, the simple measure of the volume of state employment laws enacted over time—including both minor and major enactments together—is highly correlated with multiple alternative measures of the strength of statewide protections for workers and remains quite useful in over-time analyses.

The scope of employment laws has clearly grown over time as well. New topics and issues have been added to the list of core employment laws each decade, indicating a gradually expanding issue space (Table 1). Whereas in the 1960s the vast majority of employment laws dealt principally with wages and child labor, the 1970s saw a surge of attention to discrimination and equal employment opportunity, and the 1980s saw the emergence of new employment laws pertaining to parental leave, plant closings, privacy and drug testing, whistleblower protections, and more. In the 1990s, state legislatures expanded their attention to issues of genetic testing and the complex standards for discharging employees; and in the twenty-first century, legislatures have tackled all of the traditional issues plus independent contractor issues, the eligibility of undocumented immigrants to work, workplace violence, and more.

Table 1. Categories and Counts of Employment Laws Enacted by Decade

* “Wages” includes minimum wages, overtime, wages paid, and prevailing wage. “Family leave” includes parental leave and family issues. “Undocumented workers” includes protections for immigrant workers, migrant workers, and penalties for work-related human trafficking.

The relative attention paid to major categories of employment law also reveals instructive temporal dynamics. Placing the thirty-three types of employment laws into four relatively distinct groups—wages, hours and leave, discrimination and retaliation, and terms and conditions of employment (see Table 2)—we see that employment laws dealing with wages became less common over the years relative to other types of laws, especially those pertaining to terms and conditions of employment (Figure 3). Laws covering matters of discrimination and retaliation notably grew and then fell over the decades, consistent with the historical trajectory of the “rights revolution,” while fairly steady attention was paid to laws pertaining to hours and leave.Footnote ⁶¹

Fig. 3. Issue Attention by Decade.

Table 2. Groups of Employment Laws

These over-time trends appear to reflect the emergence of new problems and vulnerabilities in the workplace as the nature of the employment relationship changed and unions declined. Laws dealing with wages—constituting more than 40 percent of all state employment laws in the 1960s and 1970s—generally raise the floor on wages and deal with the procedural requirements of where, when, and how employees are paid. They establish a baseline above which unions could conceivably negotiate other terms and conditions of employment. But as fewer workers were represented by unions and a greater share of the workforce became vulnerable to new forms of exploitation (e.g., using genetic tests to discriminate, the rising use of independent contractors and misclassification of employees), attention shifted to more nuanced and complex issues regarding the terms and conditions of employment: who is eligible to work, how workers must be or cannot be treated, and who is considered an employee.

The content of these laws thus appears to reflect the changing nature of work in the United States and the growing need for workers’ rights and protections in areas that might have otherwise been guarded by unions and collective bargaining agreements. In addition to wages, hours, and other terms and conditions of employment, the substantive reach of state-level employment laws has extended to core rules and workplace relations such as the eligibility to work; discrimination in hiring, firing, and relations in the workplace; the supply of workers; the legal status of workers; the ability to take temporary leave; the privacy of employees at work; employers’ responsibilities to employees; the legal bases for employee discharge; and enforcement mechanisms for all of the above. The very same substantive issues that labor law was originally designed to address through union representation and collective bargaining processes are now addressed through state-level employment laws, implemented through different processes, and enforced through different mechanisms.

The powerful constraints imposed by labor law's drift are also evident in what is conspicuously absent from these state laws. Due to labor law's preemption of subnational labor laws pertaining to private-sector labor-management relations, state employment laws do not (and cannot) address issues related to unionization or collective bargaining in the private sector. If workers’ advocates had been able to enact state labor laws altering union election procedures, raising the penalties for employer interference and intimidation, allowing for industry-based unions, or constructing other alternative arrangements, they surely would have (indeed, many have tried).Footnote ⁶² Many states did, of course, enact a variety of extensive labor laws governing industries explicitly falling outside the Wagner Act's reach: especially in the public sector, but also in agricultural and domestic work. But as long as national labor law remained in place and exerted exclusive authority over private-sector labor relations, the possibility of regulating labor-management relations in the private sector through state-level labor laws was foreclosed. Instead, workers and their advocates turned to employment law to achieve the same substantive ends through different means.

In other words, as a practical matter, governance of the employment relationship has shifted to alternative venues (states rather than the national level), taken on alternative institutional forms (employment laws rather than labor law), and employed alternative enforcement mechanisms (regulation and litigation rather than collective bargaining). Rather than determine workers’ wages, hours, and terms and conditions of employment through union-negotiated collective bargaining contracts (as labor law seeks to do), employment laws mobilize the regulatory instruments of the state to enforce higher standards and provide workers with a private right of action. That is, they pursue the same substantive outcomes as labor law but use different institutional pathways to get there. With respect to their timing, form, and content, then, these laws reflect the delimiting force of the increasingly antiquated but still authoritative labor law.

These insights have been gleaned from a great distance and at the aggregate level, however, and many questions remain. Not least, what explains the variation in employment law activity across states? And within states, are there patterns to be observed in who was pushing for the laws and what state-level conditions were more conducive to their enactment? The next section digs deeper into the data to begin sketching out these dynamics.

LABOR UNIONS AND THE GROWTH OF STATE EMPLOYMENT LAWS

Geographically, the volume of employment law enactments varied significantly across all fifty states and the District of Columbia between 1960 and 2014 (Table 3 and Fig. 4), with California an extreme outlier, enacting 513 employment laws (compared to the median of 122). Fine-grained historical case studies are clearly needed to examine why different states produced more or less robust employment law regimes across the states—for as Margaret Weir has written, states are “political arenas with their own distinctive capacities and political logics that must be understood in terms of earlier reform efforts.”Footnote ⁶³ Still, to begin exploring why some states enacted more employment laws than others, some theoretical possibilities can be examined statistically.

Note. Shades represent quartiles, with darker shades indicating more employment laws.

Fig. 4. Map of Total Employment Laws Enacted by State, 1960–2014.

Table 3. Total Number of Employment Laws Enacted, 1960–2014

Of particular interest here is the relationship between labor unions—the primary group beneficiaries of labor law's early feedback effects—and employment laws. Was part of the growth in employment law “attributable to the decline of organized labor,” as St. Antoine has posited?Footnote ⁶⁴ If so, why and how? Were employment laws enacted as substitutes for disappearing unions, were labor unions themselves instrumental in enacting those same laws, or both?

While statistical analyses cannot answer these questions definitively, they can reveal associations that may serve as the starting point for further investigation. I build on preeminent labor economist Richard Freeman's examination of the relationship between state-level protective legislation and union density but use the state-year MLR employment law data as my dependent variable.Footnote ⁶⁵ Specifically, I use a biennial measure of the total number of employment laws enacted in each state during the preceding legislative sessions (regular and special) and add three theoretically relevant political explanatory variables.Footnote ⁶⁶ The modified model estimates the following equation:

$$\eqalign{{\rm Employment}\; {\rm Laws}{\rm} & = a + b\;UnionDensity + c\;MFG \cr & \hskip-7pc + d\;Unemployment{\rm} + {\rm} e\;Income{\rm} ;+ {\rm} f\;LegislativeProductivity \cr & \hskip-7pc + g\;Democratic{\rm} + {\rm} h\;MassEconLiberalism{\rm} + {\rm} i\;CA{\rm} \cr & \hskip-7pc+ {\rm} j\;RTW + u} $$

Explanatory variables include UnionDensity, which measures the percentage of nonagricultural workers in each state who are members of unions, since a greater share of unionized workers may indicate a more politically powerful labor movement, which may push for the enactment of stronger employment laws or, alternatively, reduce their necessity.Footnote ⁶⁷ MFG is the share of the state's nonagricultural workforce employed in manufacturing: As a state's manufacturing industry declines, it may be encouraged to enact stronger protections for the workforce that remains.Footnote ⁶⁸ Unemployment is the level of nonagricultural unemployment in the state: When a state's unemployment rate is higher, legislators may be more reluctant to enact stronger protections for workers out of fear that such laws may deter investment in state and cause capital to flee.Footnote ⁶⁹ Income measures per-capita personal income in the state, since economic modernization is related to policy innovation more generally, and states with richer populations may be more likely to enact employment laws.Footnote ⁷⁰ LegislativeProductivity is the overall number of laws enacted by the state legislature during the previous two-year term (minus vetoes plus overrides), since states that typically enact more laws in general may be more likely to enact employment laws than states that are generally less legislatively productive.Footnote ⁷¹ Democratic is a dummy variable indicating Democratic control of both legislative houses in the state, since these states may be more likely to pass worker-friendly laws.Footnote ⁷² MassEconLiberalism is the Caughey-Warshaw state-year measure of the mass public's economic policy liberalism, included because states with more ideologically liberal populations on economic matters may also be more likely to enact worker-friendly employment laws.Footnote ⁷³ CA dummies for the state of California because it is an extreme outlier. RTW is a dummy variable indicating whether the state adopted a “right-to-work” law, since these states may be less inclined to pass worker-friendly employment laws. Right-to-work is also sometimes used as a proxy for the political strength of business.Footnote ⁷⁴

Using these variables, I conducted a time-series cross-section analysis and estimated the model with and without year and state fixed effects, since unobserved time-invariant characteristics of states—such as political culture, political history, industrial mix, demographic composition of the workforce, and the like—likely also matter. Results are reported from each specification (Table 4). The Driscoll and Kraay standard errors are heteroskedasticity- and autocorrelation-consistent and robust to general forms of cross-sectional dependence, all of which is necessary, given the structure of the data.Footnote ⁷⁵ Data limitations confine the time-series cross-section analysis to the period 1976–2013.Footnote ⁷⁶

Table 4. Time-Series Cross-Section Models

Note. Standard errors in parentheses: *** p < 0.01, ** p < 0.05, * p < 0.1.

Model 1 in Table 4 includes both year and state fixed effects, while Models 2 and 3 include one but not the other. In all three models, union density is positive and statistically significant, meaning higher union density predicts the enactment of more employment laws. This relationship holds both within and across states, even when controlling for time trends. In addition, states that are more legislatively productive overall and states with Democratic legislative control are also positive and significant, as expected. The share of nonfarm employment in manufacturing is negative and significant in each model, suggesting that the tendency to enact more worker protections is greater when the share of manufacturing jobs in the state declines. The relationship between employment laws and both personal income and mass economic liberalism depends on whether one is looking within or across states, and right-to-work is negative and significant when looking across states. Taken all together, then, these results indicate that union density and employment law enactments are positively associated and suggest that political factors matter as well, thus recommending further examination of why (see discussion below).

The same results obtain when we step back and consider aggregate outcomes in the states over time: Why did some states pass more laws over multiple decades than others? Why did some pay more attention to employment law relative to other items on their legislative agenda? Why did some tackle more categories of employment law than others? Using only fifty observations (for each state), I ran a similar set of models to see which factors were most correlated with a state's overall volume, relative attentiveness, and scope of employment laws (number of categories out of thirty-three possible) passed between 1974 and 2014 (Table 5). These models use the same covariates as above, but they disaggregate union density into two variables: the baseline level of union density prior to the period under investigation and the percentage point change in union density over the period (to account for the fact that the change in union density may affect states differently depending on the level at which they started). I also include baseline level and percentage point change variables for MFG for the same reason. A measure of the total contributions to state office candidates from the general business sector is also added to capture the relative political strength of business in the states, in addition to RTW.Footnote ⁷⁷

Table 5. Snapshot Models (50 States, 1974–2014)

Note. Standard errors in parentheses: *** p < 0.01, ** p < 0.05, * p < 0.1.

The results confirm that above and beyond all other factors, union density is strongly related to state employment laws: Both the base level of union density (average, 1969–1973) and the percentage point change in union density over the forty-year period have substantively large coefficients and are highly statistically significant. That is, the higher the starting point and the less decline in union density over time, the more laws enacted. Legislative productivity and right-to-work are also strongly related, in the expected directions.

But what about different types of employment laws—have labor unions pushed for certain categories of laws more than others? Looking at the relationship between union density and groups of employment laws while including the same covariates as above, it is apparent that union density is positive and significantly related to each group, but it explains some groups better than others (Fig. 5). Specifically, when we look only at laws pertaining to wages, hours, and leave, union density is significant at the p < 0.000 level; it is significant but only at the p < 0.10 level when explaining discrimination/retaliation and terms/conditions of employment. In sum, then, the relationship between labor unions and state-level employment laws is strong, but there is still some variation that cannot be fully explained quantitatively and from a distance.

Note. point estimates are for Union Density variable in models that include all the same covariates as above. Error bars represent 95 percent confidence intervals.

Fig. 5. Relationship between Union Density and Groups of Employment Laws.

Of course, despite its ubiquitous use in the literature, the standard measure of union density provides a rather crude proxy for the political strength of the labor movement: Union density may or may not be related to the political and legislative clout of labor unions at the state level. In Table 6, the first model therefore uses an alternative measure labor union political involvement as a substitute for union density: Like the business strength measure above, labor contributions measures the total contributions to state office candidates from contributors in the labor sector for available years.Footnote ⁷⁸

Table 6. Alternate Measures of Labor Strength

Note. Standard errors in parentheses: *** p < 0.01, ** p < 0.05, * p < 0.1.

The standard union density measure also makes no distinction between private and public-sector unions, despite their substantial differences. This is problematic, for as Alexis Walker and others have discussed, public-sector unions are outgrowths of state labor laws and are therefore subject to, and may contribute to, different political dynamics than private-sector unions.Footnote ⁷⁹ Indeed, the U.S. labor movement has long been bifurcated between the public and the private sectors, and this divide has grown increasingly consequential for the labor movement over time. In many states, public-sector unions now considerably outspend private-sector unions in electoral and lobbying activities, and their members have become increasingly engaged in political activities. Indeed, state-level anti-union conservatives, often backed by the Koch brothers’ Americans for Prosperity, have in recent years targeted public-sector unions in strategic efforts to demobilize the left.Footnote ⁸⁰ Model 2 in Table 6 thus disaggregates UnionDensity into public and private-sector union density measures (state-level data available only since 1983).

The first model shows that labor unions’ political contributions are positive and significant—lending additional support to the notion that the strength of labor unions, whether measured as the percentage of unionized workers in a state or as a function of labor's campaign contributions, helps to explain employment law enactments at the state level. The second model indicates that public-sector union density is significantly related to the enactment of employment laws—and although private-sector union density is positive and significant when used alone, the inclusion of public-sector union density in the model overpowers the explanatory weight of private-sector union density. This positive and significant relationship between public-sector union density and employment laws may therefore suggest yet another reason why opponents have sought to undercut collective bargaining rights in the public sector.

Caution is warranted in interpreting the results, however, since the strength and size of public-sector unions, as well as their institutional power (such as collective bargaining rights, agency fees, dues check-off, etc.) depend on state labor laws, which are products of the same legislative process that produce state-level employment laws (our dependent variable). Including a separate measure of public unions’ strength on the right-hand side of the equation may invite endogeneity into the analysis. But this, too, is an empirical question that requires further investigation. Moreover, making strong distinctions between public- and private-sector unions may obscure more than it reveals. State AFL-CIOs are usually composed of a mix of both public- and private-sector unions and workers—as are many of the larger unions like Service Employees International Union (SEIU)—and they self-consciously represent the interests of both in the legislative arena. In Rhode Island, for example, even as public-sector workers grew into the majority of workers represented by the state AFL-CIO, the state federation's legislative agenda remained remarkably consistent over the years, suggesting that the interests of public- and private-sector union workers in state politics may not be easy to distinguish.Footnote ⁸¹ Partnerships between public-sector unions, private-sector unions, community groups, and other workers’ advocates constitute the organizational core of the labor movement, and the relative strength of each cannot easily be disentangled with quantitative measures of union density. Further investigation is clearly needed, and additional case studies should help unpack the relationships further.

For the purposes of cross-validating the analyses above, I tested three alternative dependent variables that theoretically capture the same phenomenon of interest (variation in the strength of state employment laws), albeit in quite different ways. The first two measure the extent to which state laws are considered by employers and business advocates to be hospitable environments for investment and growth. The first uses a Chamber of Commerce study in which researchers examined thirty-four types of laws on the books in 2009 and graded each state on the extent to which its laws increased the regulatory burden on business and opened the door to litigation. The second uses the Index of Worker Freedom compiled by the conservative Alliance for Worker Freedom in 2009, which tracked fifteen laws that it claimed obstructed workers’ freedom and drove away high-quality workers.Footnote ⁸² Note that the sign is flipped on this measure. The third, compiled by the author, measures the number of eleven major state-level employment laws on the books in each state in 2014, as tracked by two authoritative sources: the National Conference of State Legislatures and the Department of Labor.Footnote ⁸³ All three alternative measures are highly correlated with the employment law data examined above, and as Table 7 shows, union density is strongly related to each.

Table 7. Alternate Dependent Variables

Note. Standard errors in parentheses: *** p < 0.01, ** p < 0.05, * p < 0.1.

Finally, a note on the “direction” of the main dependent variable used here (the MLR-derived data set of employment laws). A plain reading of the MLR employment law summaries indicates that the overwhelming majority of the enacted laws are designed to advance workers’ rights and provide statutory protections from exploitation. Still, the purposes of some laws are difficult to discern, and some are clearly more employer friendly—aimed at reducing the regulatory burden on employers, limiting their liability, making it more difficult for workers to sue, and so on. A qualitative effort at hand-coding the summaries indicates that about 15–20 percent of the laws fall into this category.Footnote ⁸⁴ As a share of all employment laws, these laws have grown slightly over time while roughly tracking partisan shifts, which is consistent with recent research on growing state-level policy polarization (see Figs. 6 and 7).Footnote ⁸⁵ The peak years in the late 1970s and mid-1990s, for example, correspond to major electoral gains for the GOP at the state level (357 legislative seats in the 1978 elections, 514 seats in the 1994 elections). By state, the volume of employer-friendly laws is correlated with employee-friendly laws, however (corr: 0.79), suggesting that some states are simply more legislatively active in this area than others and may be considered more lively sites of political contestation on employment issues.

Fig. 6. Number of Employer-Friendly Laws Enacted by Year, 1973–2010.

Fig. 7. Employer-Friendly Laws as a Share of All Laws Enacted.

Still, to be as conservative as possible, the above analyses include all employment laws, effectively biasing the analyses against the findings that more Democratic states with stronger labor unions enacted more laws. But Table 8 provides robustness checks. The first model excludes any law that could reasonably be considered employer friendly (e.g., reducing the regulatory burden on employers, making it more difficult for workers to sue, and so on). The second model uses as its dependent variable only those laws that are highlighted in the MLR’s opening pages, in which the major trends in employment law enactments in the previous year are summarized, arguably providing a measure of “significance.” The third model uses as its dependent variable only employer-friendly laws. Union density is positive and significant in Models 1 and 2 but not in Model 3, as expected.

Table 8. Robustness Checks

Note. Standard errors in parentheses: *** p < 0.01, ** p < 0.05, * p < 0.1.

It bears emphasizing that none of these cross-sectional models deal fully with unobserved heterogeneity, such as states’ deeply rooted structural or cultural attributes, which may explain both union density and employment law activity. Strategically selected case studies are clearly needed to help flesh out the nature of the relationship further. But as a first step, the robust statistical relationship deserves greater theoretical consideration. Why might there be such a strong correlation between union density and state-level employment laws, above and beyond other “usual suspect” explanations?

THE RISE OF ALT-LABOR

As the structure of the employment relationship changed and millions of low-wage immigrant workers entered the workforce between the 1990s and 2010s, labor law's fixity and stagnation left a growing number of workers either formally denied collective bargaining rights or unable to realistically access them. Filling precarious jobs that left them vulnerable to exploitation and abuse, many of these workers sought help. Traditional labor unions, however, faced significant legal and practical limits on how they could confront these problems and advocate for these workers. Into this void stepped a new generation of workers’ advocates (many of whom sought principally to aid the growing number of immigrant workers in the United States, not advance the purposes of the labor movement per se), and many community-based organizations began to grow from the bottom up. These emergent social actors began to develop new organizational forms, tactics, and strategies to advocate for workers’ rights and address other shared problems.Footnote ¹⁰⁶ These nontraditional forms of worker organization and inchoate social movements they built, sometimes called “alt-labor,” began to multiply in locations across the country.Footnote ¹⁰⁷

As the moniker implies, alt-labor groups are not labor unions—many of their members are forbidden from unionizing, and as nonprofit groups, they have no collective bargaining rights under the NLRA. They do sometimes receive funds from traditional unions and almost always stand shoulder-to-shoulder with unions in policy campaigns and street-level protests.Footnote ¹⁰⁸ They include worker centers like the Restaurant Opportunities Centers United and more than 200 other community-based worker centers across the nation; workers’ alliances like the National Taxi Workers’ Alliance; employee associations like OUR Walmart (Organization United for Respect at Walmart); associate member groups formally affiliated with unions, like Working America; nonprofit organizations like the Freelancers Union; faith-based groups like Interfaith Worker Justice; digital platforms for collective action like Coworker.org; and difficult-to-characterize movements and groups like the #RedforEd teachers’ strikes and the Fight for $15.

Like the employment laws discussed above, both the substantive aims and organizational forms of these alt-labor groups reflect the constraints and opportunities imposed by the still authoritative national labor law. In terms of their substantive foci, they self-consciously target workers that have been left behind as labor law has ossified. Their primary members include workers who are either virtually impossible to organize given the nature of their occupations or their geographic dispersion (like temp workers, fast-food workers, and taxi drivers); those who are legally excluded from labor law's provisions (such as domestic workers, independent contractors, farm workers, and day laborers); and those who do not know their rights or feel unable to assert them without fear of legal trouble (such as some nonnative English speakers and undocumented immigrants). They represent workers who are “either by law or practice excluded from the right to organize in the United States.”Footnote ¹⁰⁹

And because they are not structured nor do they claim to serve as employees’ exclusive bargaining representatives with employers (as per the rules of national labor law), their approach to protecting and promoting workers’ individual and collective rights has tended to be more confrontational, involving street-front protests, boycotts, and the generation of negative publicity for low-road employers (and favorable promotion of high-road employers). Although many worker centers do take advantage of Section 7 of the NLRA, which protects “concerted activities,” and often encourage workers to unionize, they typically focus more on assisting and empowering individual workers, helping them to navigate the bureaucratic process, connecting workers with others experiencing similar problems in similar industries or geographic locations, and pushing for the enactment and enforcement of stronger state and local employment laws, alongside many other varied activities.Footnote ¹¹⁰ They also emphasize connecting workers with private attorneys and building broader coalitions with other like-minded organizations. Thus in virtually every way, the structures and operations of alt-labor groups reflect the constraints imposed by the persistence of national labor law.

Thanks to the pioneering work of Janice Fine and her collaborators, we know that worker centers, the organizational core of alt-labor, have grown dramatically in number over the last three decades, from only five in 1992 to 139 in 2005 to approximately 214 in 2012 (see Figure 8).Footnote ¹¹¹ As worker centers multiplied, they became increasingly politically active, “successfully [placing] labor standards enforcement on the public policy agenda at the state and national levels.”Footnote ¹¹² For example, worker centers and other alt-labor groups took the lead in pushing for the enactment of major “wage-theft laws” in twelve states between 2006 and 2013.Footnote ¹¹³ In most cases, policy advocates built new coalitions composed of alt-labor groups, unions, legal clinics, and other organizations. Many continued to collaborate after the laws were passed, expanding their policy agendas to tackle new issues and continue building their coalitions. And those were only the successful cases: As a tally taken by Interfaith Worker Justice shows, dozens of other wage-theft campaigns failed or stalled in places with less favorable partisan majorities.Footnote ¹¹⁴ Alternative labor organizations are now central to the organizational and strategic outlook of the labor movement in the early twenty-first century.

Sources: Janice Fine, Worker Centers: Organizing Communities at the Edge of the Dream (Ithaca, NY: ILR Press/Cornell University Press, 2006); Janice Fine, “New Forms to Settle Old Scores: Updating the Worker Centre Story in the United States,” Relations Industrielles/Industrial Relations (2011): 604–30; Janice Fine and Nik Theodore, “Worker Centers 2012: Community Based and Worker Led Organizations,” (Washington, DC: Center for Faith-Based and Community Partnerships, U.S. Department of Labor, 2012); Interfaith Worker Justice, http://www.wagetheft.org/find-a-worker-justice-organization.

Fig. 8. Growth of Worker Centers.

A helpful illustration of how this new organizational innovation and political mobilization has been shaped by economic change amid labor law's inertia is found in the case of homecare workers. As Eileen Boris and Jennifer Klein detail in their magisterial study, low-wage homecare workers ineligible for NLRA's collective bargaining rights found their exclusion to be, itself, galvanizing.Footnote ¹¹⁵ As these disproportionately female, minority workers grew in numbers and came to see themselves as part of a broader group of workers facing the same sorts of challenges in disparate workplaces, they began to organize collectively to contest their exclusion (eventually forming the alt-labor National Domestic Workers’ Alliance in 2007). In a process akin to what Chloe Thurston describes as the work of “boundary groups” in discovering and contesting the state's role in authorizing exclusionary public-private policies, homecare workers found that although they could not change the labor laws, they could push for access to employment law protections.Footnote ¹¹⁶ After decades of grassroots organizing, coalition building, and a long and tortured legal process, homecare workers won Domestic Workers’ Bills of Rights (coverage under state labor and employment laws) in New York in 2010, Hawaii in 2013, California in 2014, and Illinois in 2016, and finally achieved eligibility for federal minimum wage and overtime pay in 2015.

Employment laws have also been instrumental for alt-labor groups: Campaigns for stronger laws have served as a means of generating collective action, building solidarity across disparate occupations, and awakening ordinary workers’ collective consciousness. In recent years, the Fight for $15 has been the most highly publicized alt-labor movement to raise wages for all workers and bring together those who would not necessarily have a chance to interact otherwise.Footnote ¹¹⁷ Although initially focused on fast-food workers, the movement quickly grew to include “home health aides, federal contract workers, childcare workers, and airport workers … employees at gas stations, discount outfits, and convenience stores,” and more, eventually morphing into a “broad national movement of all low-wage workers.”Footnote ¹¹⁸ The SEIU invested $2 million to support the first set of protests in New York in 2012, and by 2018 some estimated it had invested tens of millions in the growing movement.Footnote ¹¹⁹ Not wedded to a single tactic and led by mostly non-union workers, the Fight for $15 employed day strikes, public relations campaigns, community organizing, corporate campaigns, and legal battles to convince both public officials and employers to raise the minimum wage to $15 per hour. By 2018, the National Employment Law Project estimated that the multifaceted campaigns had resulted in $68 billion in raises for 22 million low-wage workers across the nation.Footnote ¹²⁰

Despite its wide-ranging activities and myriad accomplishments to date, alt-labor's room to maneuver has been remarkably narrow. Labor law sharply delimits what non-union groups can do to influence workplace relations; tax law presents an additional set of limitations even as it affords certain political opportunities; financial resources are in short supply, and the beneficiaries of alt-labor's activities can easily free-ride; and as the Fight for $15 illustrates, many of alt-labor's hopes of moving forward and scaling up depend on the strategic choices made by already heavily resource-constrained traditional labor unions that are themselves struggling to maintain their “core competencies” in the workplace. Still further challenges and tensions are discussed in the next section.

Yet these new types of workers’ groups have managed to fill an important organizational need and are now widely recognized as important players in both the labor movement and in Democratic Party politics more broadly, with politicians and party activists often overtly seeking to curry their favor. In 2016, for example, the national Democratic Party platform adopted Fight for $15's call for a national $15 minimum wage in an unconcealed effort to harness the energy of the movement, and the runner-up for the presidential nomination, Senator Bernie Sanders, publicized the (worker center) Coalition of Immokalee Workers’ national Campaign for Fair Food in a much-discussed 5-minute campaign commercial.Footnote ¹²¹

During the Trump presidency, up-and-coming Democratic politicians and presidential aspirants continued to try to tap into the growing activism of cross-cutting workers’ movements that blur traditional boundaries between union and non-union, private sector and public sector, and economic, racial, social, and political issues—exemplified in the wave of #RedforEd teachers’ strikes in 2018 and 2019;Footnote ¹²² in the #MeToo strikes and walkouts at McDonald's, Google, and elsewhere to protest sexual harassment in the workplace; and in locally rooted coalition-building efforts between unions and community groups seeking to “dismantle systems of racial discrimination and oppression; reclaim our democracy; reclaim our economy; reclaim our government [and] build the power of workers and workers’ organizations.”Footnote ¹²³ Gradually, more Democratic office seekers began to embrace the intersectional politics these movements embodied, “born out of [workers’] shared experiences of racism, sexism, poverty, and exclusion from the political process.”Footnote ¹²⁴ What Sanchez Ocasio and Gertner call “common-good unionism” increasingly views issues of workplace justice as inextricably linked to social, racial, gender, and citizenship injustices and inequalities, and its policy agenda is formulated in that light, reflecting workers’ simultaneous marginalization along multiple dimensions. They write:

As one Fight for $15 worker-activist aptly summed up: “Racial justice and economic justice are really just two sides of the same coin here. As a woman of color, I've been passed up for promotions and higher-paying positions.”Footnote ¹²⁶ It is important to point out, however, that politicians are not just responding to developments on the ground: The overlapping problems that galvanize these movements are increasingly constitutive of politicians’ political identities and organizing strategies—epitomized by the rise of Rep. Alexandria Ocasio-Cortez (D-NY) in 2018.Footnote ¹²⁷ In a similar vein, Michelle Chen writes that the inspirational work of the Coalition of Immokalee Workers served as both a “metaphor for the Sanders’ campaign” in 2016 and spoke “to a deeper grassroots ethos of social change from which his politics emerged.”Footnote ¹²⁸ In short, alt-labor and other still hard-to-define movements, forged in the shadow of an antiquated but still powerful labor law, constitute major new political developments that have altered actors’ incentives, shaped the substance of political conflicts, and reconfigured the political-organizational universe.

DISCUSSION: OLD PROBLEMS, NEW PROBLEMS

Reactions to labor law's ossification and drift thus include (1) the emergence of new employment laws at the subnational level designed to address the very same problems labor law no longer effectively addressed, but in different forms and through different mechanisms; (2) strategic efforts by labor unions to embrace employment laws and pursue their expansion; and (3) the emergence of new groups and movements to advocate for those workers who have been left behind amid the decline of the national collective bargaining regime. The persistence of national labor law has been found to powerfully shape the form, content, and timing of these institutional and organizational developments.

At the same time that labor law was stagnating, union density was falling precipitously, right-to-work laws were spreading to more states, and both private- and public-sector unions were coming under increasing attack. The developments discussed here cut in the other direction, invigorating the labor movement and creating new pathways to collective action and the assertion of workers’ rights. Yet as the new institutions and organizations emerged in parallel to old institutions and organizations, new problems emerged in the space between them. Put differently, not only were the new institutions and organizations unable to solve the problems produced by the increasingly antiquated labor law, but the coexistence of multiple institutions and organizations geared toward solving the same problems in different ways created a new set of unforeseen problems.