TCF senior fellow and director of labor and employment policy Jody Calemine presented testimony to the U.S. House of Representatives Committee on Education and the Workforce, Subcommittee on Health, Employment, Labor, and Pensions on November 30, 2023 at a hearing, “Safeguarding Workers’ Rights and Liberties.” The testimony, which is reproduced below, explores the history and harms of “right to work” laws and underscores how national “right to work” legislation has everything to do with undercutting workers’ growing bargaining power and maintaining the status quo of an imbalanced, rigged economy, and nothing to do with workers’ rights.
My name is Jody Calemine. I am the director of labor and employment policy at The Century Foundation (TCF). I have worked in labor law practice and policy for the past twenty-four years, representing workers and unions for thirteen years (1999–2003, 2014–2023), staffing the House Education and Labor or Workforce Committee for eleven years (2003–2014), and now working with TCF starting a few months ago. TCF was founded in 1919 by American businessman Edward Filene. It is a progressive, independent think tank that conducts research, develops solutions, and drives policy change to make people’s lives better.
I thank the Subcommittee for the opportunity to submit the following testimony on H.R. 1200, the National Right to Work Act.
This bill would amend the National Labor Relations Act (NLRA) and the Railway Labor Act (RLA) to prohibit unions and employers from agreeing to what are commonly referred to as “union security” provisions. The proposal is a very bad idea, as explained below.
Background on Union Security
Union security provisions help ensure that a union has the resources to do the work of bargaining and enforcing a contract. If workers at a company unionize, their union becomes the exclusive bargaining representative of all the employees in the bargaining unit. Federal law imposes a duty of fair representation on the union, requiring the union to represent all employees in the bargaining unit fairly, whether they are members of the union or not. Federal law already prohibits the “closed shop,” which is an arrangement where the employer may only hire and retain union members. In other words, it is already unlawful to compel a worker to join a union as a condition of employment.
What federal law does allow is a “union shop,” which is where, after a grace period of at least thirty days, a newly hired employee must meet the core financial obligations of membership—that is, pay dues even if they decline actual membership. Federal law also allows an “agency shop,” wherein the employee must pay fees to the union if they opt to not become a full dues-paying member. There is very little operative difference between an “agency shop” and a “union shop.” Under either lawful union security clause, if an employee opts not to join the union and instead pay the agency fees or financial core obligations, federal law allows such a nonmember to opt out of paying any amount beyond costs that are germane to collective bargaining, contract administration, or grievance adjustment. Finally, since enactment of the Taft–Hartley Act of 1947, while federal labor law preempts state and local policy on most matters concerning private sector collective bargaining, NLRA Section 14(b) has allowed states to enact prohibitions on union or agency shop arrangements altogether. The Railway Labor Act, which covers airlines and railroads, contains no 14(b) equivalent and instead explicitly authorizes union or agency shops if a covered union and employer agree to them, regardless of the state.
Background on “Right to Work” Laws
The state-enacted prohibitions on union security, made possible by NLRA Section 14(b), have been termed “right to work” laws—a misnomer adopted by their early proponents. In a state without a “right to work” law, union security agreements are not required, but the union and employer may voluntarily enter into them. In a state with a “right to work” law, an employer and a union are prohibited from entering into a union security agreement. In other words, a right-to-work law is a governmental restriction on the freedom of contract. These state laws do not reach employers subject to the Railway Labor Act. And with respect to public sector employment, these laws were superseded by the 2018 US Supreme Court decision in Janus v. AFSCME, which effectively made “right to work” the law of the land for the public sector in every state.
In “right to work” states, where union security clauses are illegal, an employee under a union contract can opt to pay nothing for the union’s services while enjoying the benefits of the contract and being entitled to the union’s representation and enforcement efforts. That nonmember and non-fee-paying employee will receive the wage increases that the union—funded by his or her coworkers who are dues-paying union members—won at the bargaining table. If that nonmember and non-fee-paying employee is fired without just cause, the union will represent that employee in the grievance procedure and fight to win his job back. Receiving all these benefits without paying for them while others do pay for them creates a free-rider problem, i.e., why pay for the services if one can receive them for free?. The “right to work” restrictions on union security are not intended to give anyone the right to work as the phrase might be commonly understood. Instead, these laws are designed to deplete the union’s resources, weakening its bargaining power over time, in the hopes that it will eventually disappear from the workplace altogether or at least be dissuaded from organizing further in that state.
The Purpose of “Right to Work” Laws
The true anti-union, anti-worker character of a “right to work” law is apparent from the company it keeps. The states that initially rushed to adopt these laws in the 1940s and 1950s were not beacons of workers’ rights. On the contrary, other than some very rural states with little industrial union experience, such as South Dakota or Nebraska, they typically were states that were also seeking to hang onto various Jim Crow laws. “Right to work” snugly fit into that legal framework. For example, while the state of Florida was adopting its first-in-the-nation “right to work” law, it was actively defending a version of a state peonage law before the U.S. Supreme Court. Under that law, a worker could be arrested and prosecuted for fraud for quitting a job before the job’s term was over, restricting his right to work for another employer. Other states, such as Georgia, featured vagrancy laws that criminalized unemployment so that poor people, particularly poor Black men, could be charged with vagrancy and punished with work on a farm. One historian has quoted the Atlanta Constitution’s appeal to police to illustrate how the vagrancy law worked: “Cotton is ripening. See that the ‘vags’ get busy.” While these states were passing “right to work” laws, newspaper articles highlighted that, “due to a tight labor situation in some of the Southern States, emigrant-agent laws are being rigidly enforced.” These emigrant-agent laws were designed to control and curtail efforts of out-of-state or even out-of-county employers from recruiting workers to work elsewhere. These and other laws limited workers’ freedom to work where they wanted and their ability to negotiate, even on their own as individuals, better terms and conditions at work.
It was in this context that “right to work” laws were adopted by these states. An unemployed person could not hold out for a better-paying job, but rather would be forced to work on a farm under a vagrancy law. An employed worker could be entrapped by a form of peonage and be forced to work out the rest of his job on the employer’s terms under threat of arrest rather than seek out other employment. Employers could rest assured that any poaching of workers by outside employers would be limited, so they would not have to compete for labor and bid its price upward. And if workers exercised their federal right to collectively bargain, there was now an insidious way—through “right to work” laws—to starve the democratically elected union of resources, so that the union would not be able to bargain new terms effectively. Each of these laws in their own way was designed to ensure employer control over workers and the terms of their employment, not workers’ liberty.
There was an added racist element to the early campaign for “right to work” laws. One of the first champions of “right to work” laws, Texas businessman Vance Muse, argued that, with unions, “From now on, white women and white men will be forced into organizations with black African apes whom they will have to call ‘brother’ or lose their jobs.” For Muse, unions, with their ability to break down racial lines, were a threat to segregation and white supremacy, and “right to work” was a way of maintaining those lines.
While the misleading slogan “right to work” might have been penned by a white supremacist, the bills were pressed into law with vigorous support from employers and their associations, like the National Association of Manufacturers and the Southern States Industrial Council. As time went on, states outside the Old South became the battlegrounds for “right to work” laws. To adopt a “right to work” law was a mark of being “business friendly” in the interstate competition for capital investments.
There is a reason why big business supports “right to work” laws. It is the same reason why these laws have been more accurately described as “right to work for less” laws. These laws result in a reduced rate of unionization. And a reduced rate of unionization results in lower wages, lesser benefits, and generally cheaper labor costs. If “right to work” resulted in higher wages and benefits, then we would not see organizations aligned with big business spending millions to have these laws enacted.
Impact of “Right to Work” Laws
Studying the precise economic impact of “right to work” laws has proven notoriously difficult. These laws are never passed in isolation. They are part and parcel of a broader regulatory framework that includes other employment and tax rules. States have different industrial bases, varying access to markets, different populations, and so on. When determining whether “right to work” leads to job growth, a “business friendly” offer to an employer seeking to relocate might feature a “right to work” law but also generous taxpayer-funded subsidies. Depending on what factors you attempt to control for and time periods you consider, these studies can have wildly different results. But a few items appear undeniable or uncontradicted.
Studies have found that adopting a “right to work” law reduces a state’s unionization rate. A 1983 study published as a working paper by the National Bureau of Economic Research (NBER) found that there is a dramatic fall in organizing immediately after passage of the law and a less dramatic decline in later years, concluding that “right to work” reduces unionization by 5–10 percent. A 1995 study found that “right to work” laws “significantly affect union density in the private sector.” Another researcher agreed that adoption of a “right to work” law has an immediate negative impact on union organizing and a long-run depression of the unionization rate, reducing union density by 5–8 percent. A 2016 study focused exclusively on Oklahoma’s 2001 passage of “right to work” concluded that the law decreased the state’s private sector unionization rates. Most recently, NBER published a study of states that adopted “right to work” from 2011 to 2017 and found that these laws resulted in a 4 percent drop in unionization rates five years after enactment and a 13 percent drop in the most union-dense industries of construction, education, and public administration.
Given the union difference in wages and benefits won through collective bargaining, it is no surprise that studies find that passage of a “right to work” law reduces worker compensation from what it would otherwise be. On this issue in particular, researchers have battled over how to control for the many factors beyond “right to work” that might affect wage trends. For example, there is the reality that most states that adopted “right to work” laws already had low unionization rates and few legal protections for workers, so they began with lower wages than states that reject “right to work” laws. Likewise, these states have lower costs of living so one should consider how far the dollar goes, even if there are fewer dollars to spend. A 2011 study conducted an analysis that controlled for demography of the workforce, the job characteristics of the workers, state-level economic conditions, and cost of living. It found that, after controlling for all these variables, “right to work” states had 3.1 percent lower wages than free-bargaining states, 2.6 percent lower rates of employer-sponsored health insurance, and 4.8 percent lower rates of employer-sponsored pensions. Per this data, if “right to work” had been imposed on the rest of the country in 2011 with all these downstream effects on compensation, workers in free-bargaining states would have seen their annual salaries drop by $1,500; 2 million workers would have lost employer-sponsored health insurance; and 3 million workers would have lost access to a pension. Looking solely at wages and not benefits, the 2022 NBER-published study of the states that adopted “right to work” from 2011 to 2017 found that wages dropped by 1 percent five years after adoption and by more than 4 percent in the same timeframe for the union-dense industries of construction, education, and public administration. These researchers noted that, overall, there is a 20 percent difference in unionization rates between “right to work” and free-bargaining states, with “right to work” states having wages that are 7.5 percent lower than those in the free-bargaining states.
When it comes to employment and job growth, studies have likewise struggled to isolate the causal impact of a “right to work” law. Some have engaged in a spatial analysis, looking at counties that border one another across a “right to work” state/free-bargaining state divide. A great deal of attention is focused on manufacturing employment in particular, since, unlike services, manufacturing can be relocated, with “right to work” laws being part of a “smokestack chasing” strategy. Certainly, “right to work,” as shorthand for weak or nonexistent unions or a lesser chance of being organized, is one tool of many that a state might use to compete with other states for capital investments and plant relocations. This competition can result in job loss in one state and job gains in another, and, at least anecdotally, it depresses wages across the board. A refrain I often heard growing up in a manufacturing area in free-bargaining West Virginia was that unions needed to be careful about pressing the employer too hard on wage increases or the factory would move south. After decades of employers having “right to work” venues available, it is likely that the adoption of “right to work” by a particular state does not have as significant an impact on employment as it once did. Indeed, a study of the 2001 adoption of “right to work” by Oklahoma found no impact on total employment for the state, as well as no impact on wage rates. There are bigger issues at work. That state and many other “right to work” states have seen manufacturing jobs move overseas, to countries such as China that, with no regard for workers’ rights and freedoms, can outcompete even the most anti-union states on labor costs, if labor costs and collective bargaining are the principal concerns of an employer looking to relocate.
In any event, the subject of his hearing is a national “right to work” law, which would eliminate the differences between states on this question of union security. If the theories that proponents use state-by-state are correct – that adopting “right to work” in your state will attract investment from elsewhere or new investments that would otherwise go elsewhere – then the main employment impact of a national “right to work” law is to eliminate this particular incentive for an employer to locate or relocate to a “right to work” state, therefore hurting job growth outlooks for the current “right to work” states.
By reducing labor’s bargaining power nationally, a national “right to work” law would certainly have a negative impact on workers’ compensation. Conversely, by weakening the bargaining power of workers, a national “right to work” law would likely have a positive impact on executives’ compensation. A study of recent “right to work” enactments show precisely that inverse correlation between workers’ share and owners’ share of wealth. Within one year, “right to work” laws “eliminate a substantial fraction of real wage growth” for workers as their bargaining power decreases. As the firm’s bargaining power increases, it keeps more for itself. The study found that executives benefit from adoption of “right to work,” receiving “increases in base salary, the value of options granted, and other compensation, such as contributions to pension plans.” Moreover, dividend payouts increase, particularly by the third year after adoption of “right to work.” Firms headquartered in a “right to work” state are “more likely to decrease their labor force due to a negative industry-wide shock than firms located in” free-bargaining states. These behaviors of increasing executive compensation, converting more profits into dividends for shareholders, and treating the workforce as dispensable when there are any bumps in the road are all functions of an imbalance of power in dramatic favor of the company over the workers post-“right to work.” A 2009 study tracked this same result, finding per capita personal income and wages higher in free-bargaining states, and business owner’s income higher in “right to work” states. In the context of a weakened labor movement in general over the past several decades, national “right to work” would simply mean more of the same—more income inequality, bigger gaps between executives and rank-and-file workers, and more job insecurity.
Firms freed from union watchdogs and pursuing the narrow interests of executives are not necessarily good for the overall health of the nation. A 2015 OECD study, for example, found that growing income inequality hampers economic growth. One economist extrapolated from the study: “Using the same OECD estimates, if the United States could reduce its inequality to the level in Canada, U.S. GDP would rise about 0.9 percentage points per year. This is a large effect relative to the average annual growth rate since 1970 of U.S. inflation-adjusted GDP of about 2.8 percent.” It has been the lack of workers’ power—not the lack of corporate power—that has squeezed the American middle class in contemporary times, yet a national “right to work” law would plainly only exacerbate this imbalance.
By weakening unions, a national “right to work” law would reduce the many benefits that collective bargaining produces. A 2021 report by the Illinois Economic Policy Institute found that “right to work” states when compared to free-bargaining states:
- have 3 percent lower hourly wages for workers on average;
- have 5 percent less health insurance coverage;
- have 8 percent less retirement security;
- have larger pay penalties for workers deemed essential during the COVID-19 pandemic, including 16 percent for police officers and firefighters, 11 percent for construction workers, and 7 percent for registered nurses;
- have 11 percent fewer workers with bachelor’s degrees or higher;
- have 31 percent fewer registered apprentices per 100,000 workers;
- have 50 percent more on-the-job fatalities per 100,000 workers;
- have economic productivity that is 17 percent lower per worker;
- had economic growth that was 3 percent slower during the pre-COVID-19 economic expansion;
- have a consumer-debt-to-GDP ratio that is 26 percent higher;
- have higher loan delinquency rates;
- have 15 percent more of their households falling below the poverty line;
- have 10 percent more of their households receiving food stamps;
- have no discernible bump to employment, with “right-to-work” ranking outside of the Top 10 factors cited by corporate executives in business location decisions;
- have a life expectancy at birth that is two years lower;
are not in the top ten states with regard to life expectancy, while nine of the bottom ten states with the lowest life expectancy are “right-to-work” states;
- have infant mortality rates that are 28 percent higher, on average;
- have 3 percent fewer of their adults voting in national elections;
- have 18 percent fewer of their adults contacting their elected officials; and
- have 3 percent fewer of their adults donating to charities, schools, or other nonpolitical organizations.
Some of these phenomena may be attributable to more factors than just union strength, but unions make a clear difference. Having higher wages, access to affordable health care, and experience participating in a democratic organization can make a difference when it comes to poverty, infant mortality, and civic engagement.
And unions make a difference not only where they operate and not only for their own members. The U.S. Treasury Department recently issued a report on the value of collective bargaining, pointing out the spillover effects of unions on nonunion workers. For example, for every 1 percent increase in private-sector unionization rates there is a 0.3 percent increase in nonunion workers’ wages, with even larger effects for workers without a college degree. A clear example of a spillover effect occurred in the past few weeks within the auto industry. When the auto workers won their strike against the Big Three, foreign carmakers who located their U.S. plants in “right to work” states and have remained nonunion increased their wage rates to closer approximate that of workers who had exercised their full bargaining power at a collective bargaining table. Presumably, these foreign carmakers did not want workers to be incentivized to organize unions in the wake of the UAW victory. The pay raises, benefiting workers in southern “right to work” states, was termed “the UAW bump.” With a national right to work law, and a consequently less-resourced UAW, there would be less or no UAW bump. By weakening unions everywhere, a national “right to work” law would hurt union and nonunion employees alike, as nonunion employers would feel less pressure to keep wages up to attract and retain workers or fend off an organizing drive.
A national “right to work” law has equity implications. Collective bargaining has helped narrow the wage and wealth gaps between Black and white workers. Thanks to union contracts, unionized Black workers make 14.7 percent more than nonunion Black workers, while white workers make 9.6 percent more than nonunion white workers. That higher union premium for Black workers helps close the overall wage gap. A national “right to work” law, by weakening unions and reducing unionization rates, would therefore have a disproportionately negative impact on Black workers.
On Rights and Liberties Arguments
Setting aside the economic harms caused by a national “right to work” law, it may be that proponents simply value the liberty of individual workers to not pay a fee to a union expending resources to represent them more than the liberty of a union, democratically elected by its worker members, and the employer employing those workers to make an agreement on how to cover the costs of bargaining and enforcing a contract. But that makes “right to work” a liberty to free-ride. Not even libertarians favor this “liberty.” The Libertarian Party has long opposed “right to work” laws because they amount to government restrictions on the freedom of contract between employers and unions. Nevertheless, if the liberty to free-ride is a critical individual right, why did “right to work” proponents exempt police officers and firefighters from Michigan’s “right to work” law? Of all workers whose liberty was to be protected, why would police and fire be left out? The law’s supporters struggled to come up with a coherent answer. One “right to work” proponent said police officers and firefighters were exempted from “right to work” because they do not have a right to strike. In other words, in a double-whammy against police and fire, stripped of the right to strike, they are also stripped of the individual right to free-ride? The only way to make sense of the Michigan law was to be honest about the true nature of the legislation: it was not about expanding worker liberty but about curtailing worker power. As a concession to two politically-favored groups, understanding that these groups did not have the right to strike (one kind of worker power exercised at bargaining), legislators showed mercy and decided to not also strip them of their union security (resources for bargaining). Liberty has nothing to do with it. (Incidentally, Michigan has since voted to repeal its “right to work” law, while the US Supreme Court has since imposed “right to work” on all public sector employment, including police and fire.)
After all, the liberty to not pay fees for services rendered is not just an odd notion of liberty, but also an extraordinarily narrow one. This bill concerning a national “right to work” is focused on that fraction of workers with union representation with a union security clause in their contract working in one of the now-minority jurisdictions without a “right to work” law or for a Railway Labor Act employer. Noncompete agreements, for example, impose a far greater restriction on more workers’ right to work than a union security clause. A union security clause says nothing about who is hired. It merely says that, once hired, you can enjoy the benefits of this contract so long as you pay a fee to cover the cost of bargaining, administering, and enforcing that contract. A noncompete agreement, usually imposed on hires without negotiation, says you cannot be employed by any competitors for some period of time, directly prohibiting you from working in an industry except with the employer that imposed the agreement. There are less than 12 million union-represented workers in free-bargaining states, a large portion of whom are already exempted from union security provisions as public employees. But over 30 million workers are subject to a noncompete agreement restricting their actual right to work, and 38 percent of all workers are expected to be subject to a noncompete agreement sometime during their careers. Noncompete agreements do not appear to be a concern of “right to work” proponents, even though they directly restrict a person’s actual right to work. On the contrary, the U.S. Chamber of Commerce’s letter opposing a proposal to regulate noncompete agreements at the Federal Trade Commission reads like a who’s-who of big-business “right to work” proponents. There is one thing that a noncompete agreement, which can radically restrict a person’s actual right to work, and a misnomered “right to work” law have in common: they curtail worker power and enhance employer power.
Perhaps the concern of proponents is as narrow as a right to work without deductions from your paycheck. In that case, it should be noted that there are other lawful deductions made from workers’ paychecks in both “right to work” and free-bargaining states. For example, unless a state has restricted these deductions, an employer can deduct expenses from employees’ paychecks for the provision of uniforms, so long as the deduction does not result in pay below the minimum wage. These uniform deductions go to employers to compensate them for their expenses (employer power!), not unions to compensate them for theirs (worker power!), and that may make all the difference for “right to work” proponents.
Ultimately, “right to work” is narrowly focused on a particular mechanism of union funding. Its purpose is not worker liberty. Its purpose is to weaken workers’ bargaining power. Doing so results in great benefits for the already rich, and more loss for working people. And a national “right to work” law seeks to accomplish these ends on a national scale rather than a state-by-state scale. The great irony of national “right to work” is that the one talking point that “right to work” proponents had was a state-level talking point: a “right to work” state would be more business-friendly compared to other states and therefore attract jobs from employers who do not want to bargain collectively with their employees; that is, proponents were aiming for a state with more jobs at lower pay. With a national “right to work” law, proponents would not be able to claim that particular “business environment” advantage for any particular state. All states would be the same on this score. “Right to work” would no longer constitute a reason for businesses to move from one state to another. South Carolina would lose the claimed “right to work” advantage over, say, Washington state. So what is the point of a national law? The point is undermining funding for unions. With national “right to work,” unions will certainly have lost a form of funding—the agency fee in a minority of states—and will be that much weaker as a result, and the already-rich will have that much more power.
I respectfully recommend rejecting legislation that results in, among other things, weaker unions and higher executive compensation. As we speak, unions are proving their mettle at bargaining tables and on picket lines, winning record-breaking contracts for their members with positive spillover effects for workers throughout the economy. The American people see the difference that unions make in providing a voice for workers and countering the power of big corporations. The public is giving labor organizations their highest favorability rating in half a century. This support cuts across party lines. In every major labor dispute recently, the public has overwhelmingly sided with the workers over the employers. The country wants and needs stronger, not weaker, unions. Close to 50 percent of all nonunion workers have said they would join a union if they could, but the national unionization rate—10.1 percent—realizes only a fraction of that desire. That gap is a function of a broken labor law that sits in this Committee’s jurisdiction. Instead of considering H.R. 1200, I respectfully recommend Congress pass the Protecting the Right to Organize Act, which would strengthen Americans’ right to organize and give more workers a chance to bargain for a better deal for themselves and their families.