This paper answers the title question positively. Labor would fare better if the United States reduced federal preemption of private sector labor relations law and devolved the legal regulation and enforcement of freedom of association and collective bargaining to the states. This answer runs counter to the standard story in U.S. labor history and traditional views of pro-labor scholars. U.S. history warns against trusting states to protect labor rights. The southern states had slavery until the Civil War and used state laws to suppress blacks for over a century thereafter. Given the opportunity, twenty-one states enacted right-to-work legislation that outlawed union security clauses in collective contracts. By contrast, labor histories often credit the National Labor Relations Act (NLRA) for the growth of unionism in the United States (even though many workers organized outside the Act). And from the 1960s through 2000, Congress enacted a host of labor bills—outlawing discrimination, regulating occupational health and safety, and insuring employment retirement plans.

In labor relations law, on which I focus, courts have held that the federal legislation preempts state activity. States can legislate wages above the federal minimum and enact and enforce employment law but cannot act in areas covered by the Labor Management Relations Act (LMRA). The impetus for federal preemption came from liberal Supreme Court justices, influenced by prolabor legal scholars such as Archibald Cox, who argued for federal preemption, although there is no explicit congressional statement in this regard (Gottesman 1990). Since the Supreme Court decision in Garmon (1959) courts have held that the NLRB has exclusive jurisdiction over conduct regulated by the LMRA. At the same time, the Court has interpreted state sovereign immunity to bar suits by state employees for violations of federal employment rights (Landau 2002).

The starting point for my analysis is that national labor law has failed to give U.S. workers ways to obtain the labor representation and participation that they want (Freeman and Rogers 2006), which range from unions to nonunion committees. Taking that as given, I argue that the country would profit from state legal regulations on the basis of four propositions: (1) state regulation allows valuable experimentation and divergences in policies; (2) in the areas in which they have authority, states have enacted diverse labor laws, that affect state collective bargaining and other labor outcomes; (3) labor would gain more in the states that are favorable to collective bargaining than it would lose in states less favorable to collective bargaining; (5) state law offers the best opportunity for the United States to experiment with new forms of worker organizations, including company-sponsored committees.

Proposition 1. The Logic for Federalism in Labor Relations Law

"It is one of the happy incidents of the federal system that a single courageous state may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country." —Justice Brandeis (1932)

The classic argument for federalism is that it allows for experimentation at the state or local level so that the country can better assess the impact of different policies and choose the one that works best in accomplishing the public purpose. States are well-suited to serve as laboratories of experiment because they are numerous, have boundaries that cover populations with similar attributes and views, and have the large scale necessary for assessing major policy initiatives. There are two possible responses to state "experiments." States with comparable goals to the experimenting states can imitate the most successful initiatives, so that good practices spread rapidly. States with different goals can go their own way. In both cases, welfare is enhanced over a single "least common denominator" national regulation.

Liberals often object to federalism for fear that mobility of capital will lead states into a race to the bottom, as each state reduces its protections for labor to attract investments (Donahue 1997). But the politics could go the other way. Citizens in a state with worse labor relations policies might see how well people are doing in a state with better labor regulations and move to the more desirable state or campaign for their state to copy the preferable legislation. Federalism in welfare services and medical services for the poor have not led to any race to the bottom among states. Because states deliver services to citizens subject to hard budget constraints, governors and legislators tend to be less ideological and divisive than representatives and senators in Washington. Historically, many of the reforms adopted by the New Deal were first anticipated and developed in states, such as Wisconsin.

While the conceptual case for federalism has attracted increasing support in the past decade (Boston Federal Reserve 1998), national legislation also has its advantages— including incorporation of externalities across state borders and economies of scale in setting policy. Whether labor will do better under state law requires evidence on how states have regulated labor relations when they have authority and of the impact of their actions on workers' gaining the participation and representation at their workplace that they want.

Proposition 2. Variation in State Labor Relations Law

The U.S. labor relations system is well-suited for examining the effect of state as opposed to federal regulation of labor relations because states have authority over labor relations for some workers—those employed by state and local governments. Thus, every state has private sector workers covered by federal law and public sector workers covered by state law. Since state law can vary greatly, this provides a "natural experiment" to contrast outcomes for the treatment group—public sector employees covered by that state's public sector laws—against the control group of private sector employees in that state covered by the ubiquitous federal law.

To see how states regulate the state and local public sector workers over whom they have jurisdiction, I have examined the Kim Rueben update of the NBER Valletta-Freeman state public sector labor law data set (http://www.nber.org/publaw/). This data set categorizes public sector labor laws from 1955 to 1985 for five groups of workers (state employees, local police, local firefighters, local teachers, other local employees) in terms of their legal treatment of collective bargaining and the right to strike. I have categorized the legal status of public sector collective bargaining in 1996, the last year of the Rueben update, into two broad categories: having a favorable legal environment for public sector collective bargaining, where laws explicitly require or imply that public sector employers have a duty to bargain with unions; and other legal environments, including cases where the law either explicitly outlaws bargaining or contains no provision for bargaining. The data show that 64 percent of state occupation groups have laws favorable to collective bargaining, concentrated in twenty-seven proøcollective bargaining states.

I have also compared the policy of states toward labor in eight other domains covering private sector labor, ranging from state minimum wages to right to work (RTW) laws to providing Medicaid to persons above the poverty line. In these areas, state policies differ in ways that are closely associated with the state's stance toward public sector bargaining. For instance, twenty-one states have enacted RTW laws. Seventeen of them have public sector labor regulations that are not favorable to collective bargaining. Similarly sixteen states had above federal minimum wages in 2005, and all sixteen were states with favorable public sector labor laws. From this analysis, I conclude that states' policy stances toward private sector labor vary in ways associated with the way they regulate public sector collective bargaining. Some states are favorable to labor interests and likely to enact or enforce laws favorable to labor in the private sector, as they do in the public sector labor domains over which they have authority. On the basis of their legal stance, I would expect these states to enact policies that would create favorable legal environments for collective bargaining in the private sector. On the other side, the states that are unfavorable to public sector collective bargaining are likely to create a more hostile legal environment for collective bargaining than under federal legislation. How these two forces would work out for the country as a whole depends on the impact of potential changes in laws/policy on collective bargaining and on the numbers of workers affected.

Proposition 3: State Regulations Affect Unionization Outcomes

To assess the impact of differing state labor regulations on collective bargaining, I have compared the difference between collective bargaining coverage in the public sector and in the private sector within states differentiated by their public sector bargaining laws. Since private sector workers are covered by federal law, the difference in the legal treatment of workers in this analysis comes from the variation in state regulations of public sector collective bargaining. The results of such an analysis show that states with favorable public sector collective bargaining laws have roughly 20 percentage point higher coverage in the public sector relative to the private sector than states with less favorable laws. To be sure, this cross section relation does not establish any causality. To obtain insight into the likely causal impact of favorable laws, I direct readers to studies of public sector unionization before and after changes in state labor laws in the 1970s and 1980s that showed that the laws or their timing had an independent effect on union growth in the public sector (Farber 1988; Saltzman 1988; Freeman and Valletta 1988); and to Farber (2005), who uses the within-state variation in laws by type of worker to find that union density is significantly higher where unions are allowed to negotiate union security provisions (for example, agency shop) and where employers have a legal duty to bargain with labor unions.

There are reasons for public and private sector union density to diverge even under the same legal regime. The profit motive ought to induce greater opposition to unions among private sector employers than among public sector employers, for whom union members and their allies can be an important part of the electorate. Workers may have different desires for unions in the two sectors, depending on civil service regulations and market or political pressures. It is a "reach" to extrapolate differences in public sector coverage associated with different public sector stances to private sector outcomes under the assumption that state regulation would produce outcomes in the private sector comparable to those in the public sector. But it is a reach I have to make to estimate how state regulation of labor outcomes might affect the private sector.

Proposition 4. The Net Outcome on Collective Bargaining

There is one way in which devolving labor law from the federal government to state government would guarantee that labor gains. This would be if the devolution followed the Taft-Hartley model of allowing states to go beyond federal rules in one direction only. Just as the Taft-Hartley law allows states to outlaw union security provisions but does not give them any way to enact laws to help protect workers' rights of association or to strengthen collective bargaining, Congress could give states the right to pass or enforce laws only in ways that enhanced workers' collective bargaining rights. Some states would surely act on this and would go beyond the LRMA to give workers a better chance of gaining union representation when they want it. In the 1980s Wisconsin directed its Department of Industry, Labor, and Human Relations to bar firms who violated the LMRA over a five-year period from state procurement. In 1986 the Supreme Court unanimously ruled that by prohibiting state purchases from repeat labor law violators, the state was acting as a regulator rather than as a purchaser and held that the LMRA preempted the Wisconsin statute (Wisconsin 1986). In 2000 California prohibited employers from using funds received by the state to oppose union organizing (California 2000). The emphasis was on the state's rights as a purchaser rather than as an entity seeking to enforce the LMRA. Employers challenged the California law on the notion that it limited the legal rights of employers to speak against unionization, and in 2002 the U.S. District Court for California ruled against the state. Absent a Supreme Court decision reversing this and other preemption decisions, Congressional action would be needed to create space for states to operate in ways that they have demonstrated they would like to do to limit antiunion activity.

But given the authority, some states might go beyond the Wisconsin and California efforts. Some might introduce card checks for union recognition. Others might try the quick elections that the Dunlop Commission recommended. If all of the states with favorable legislation toward public sector collective bargaining were to take policy stances favorable to private sector collective bargaining, and if those efforts had approximately the same impact on private sector bargaining as their favorable laws have on public sector bargaining, I estimate that private sector coverage would increase by 4 million workers.

But if Congress were to give states authority over private sector labor relations, I would anticipate that it would allow states that are not favorable to unionism to undertake policies that could weaken collective bargaining. These states might enact laws that made it easier for employers to oppose unions, eliminate duty to bargain clauses, place greater administrative regulatory burdens on unions, or discriminate against unionized firms in its state contracts, and so on. Collective bargaining coverage is likely to fall in these states. To get some notion of how much coverage would fall, I assume that all states with unfavorable public sector labor laws would enact unfavorable laws in the private sector that would have the same negative impact on coverage as their unfavorable laws do in the public sector. This calculation implies a loss of 0.5 million union members. Thus, on net, my analysis suggests that unions would gain some 3.5 million members in the private sector if states regulated private sector labor law.

Why would unions gain more in states favorable to collective bargaining than they would lose in states unfavorable to collective bargaining? I use the same estimated elasticity of laws on outcomes in the two cases. The difference is simple. As Table 1 shows, there are millions of non-union workers who could be covered by collective bargaining in states likely to take a more positive stance toward collective bargaining, whereas unions have virtually no members to lose in the private sector in the states that are unfavorable to collective bargaining. Total private sector membership in those states is 1.2 million workers.

Finally, Canada's experience with provincial determination of private sector labor legislation is consistent with my claim that organized labor does better when labor laws are set at lower levels. Provinces have changed labor regulations, with card-check legislation in some areas and times that U.S. unions would regard as dream "pie in the sky" to get from the U.S. Congress; however, there is less favorable legislation in other areas and times. The changes in legislation in the same province over time show that the provincial legislation affects unionization rates.

Proposition 5: The Potential for Other Reforms

Other advanced English-speaking countries such as Canada, Australia, New Zealand, the United Kingdom, and Ireland allow management or workers to set up non-union organizations for management and labor communication that the Wagner Act has outlawed as company unionism. Management has no great desire to establish such organizations at present. No one knows whether they would turn into unions or become a works counciløtype substitute for unions. Given findings from diverse surveys that many workers would like alternative modes of representation and participation to collective bargaining, but which in the absence of experiments cannot be well determined, it is natural to look to the states for experimentation. While there are no guarantees, turning the law regulating private sector labor relations and/or its enforcement over to the states cannot be much worse than the U.S. labor law is now. Washington has failed. Why not see if Sacramento and Bismarck, Albany and Oklahoma City, Des Moines and Detroit, Salt Lake City and Madison can do better?

Acknowledgements

I benefited from discussions with Joel Rogers and from the comments of William Gould, Jon Hiatt, and Dan Mitchell.