In this article, I will describe this success. I will first outline the goals of the Wagner Act (the NLRA’s progenitor legislation), and then explain how the NLRA achieved those goals. I will conclude by explaining why it’s not surprising that those successes would result in declining union membership.

GOALS OF THE NLRA

The first of the Wagner Act goals was, and is, industrial peace. The preamble of the Act states that the “denial by some employers of the right of employees to organize” and bargain collectively had led “to strikes and other forms of industrial strife or unrest.” On one level, that goal means reducing the number of strikes or the economic effects of strikes. But that barely scratches the surface of that goal. Industrial strife in the late 19th and early 20th centuries went far deeper, raising the question of whether the employees would agree to work within a capitalist system.

Prior to 1932, there was no federal legal right to strike, even peacefully, and many strikes were illegal under state law or the federal common law. Employers often required that workers agree not to join a union or be involved in union activities during the term of their employment, and the federal courts held such agreements binding. Concerted activity by employees was not protected. If workers went out on strike and did not return to work when served with a state court–ordered injunction, the striking workers were in contempt of court. When confronted by police or Pinkerton guards, strikes would often turn violent. The next move in many strikes was for the governor to call out the National Guard to restore order.

In the Great Railroad Strike of 1877, federal troops were deployed in major cities in six states, including Baltimore, Pittsburgh, Chicago, and St. Louis. Striking workers often resisted, resulting in considerable violence and many deaths. Certainly one could understand President Rutherford Hayes’ concern that a revolution against the government itself might be in the making. Hence, when I use the term “industrial peace” to describe what Congress was seeking, my focus is—and Congress’s focus was—on the unrest that led to riots and the eventual use of police or military force to restore order.

Equality of bargaining power / The second goal was, and is, to redress “inequality of bargaining power.” In the words of the Act, “[t]he inequality of bargaining power … substantially burdens and affects the flow of commerce, and tends to aggravate recurrent business depressions, by depressing wage rates and the purchasing power of wage earners in industry and by preventing the stabilization of competitive wage rates and working conditions within and between industries.” Whereas the goal of industrial peace is easily stated, the same is not true of the equality of bargaining power.

The second goal is complex because it has both procedural and substantive elements. On the procedural element, the legislation’s author, Sen. Robert F. Wagner (D‑N.Y.), said that the goal was satisfied if workers were represented by unions. I will adopt Senator Wagner’s interpretation by equating the procedural element with workers’ achievement of collective bargaining status. This provides a clear and measurable goal. The greater the percentage of workers belonging to unions and engaging in collective bargaining, the more successful is the Act. Hereafter, I use the economic term “union density,” which means the percentage of workers who belong to unions.

The substantive element is raising wages, which it was hoped would reduce the likelihood or severity of depressions. The traditional indicator of whether unions raise wages is the union wage premium—that is, the percentage difference between the union wage and the nonunion wage for workers doing similar work. Collective bargaining and higher wages were linked. It was always understood that the collectively bargained wage would be higher than the wage achieved in the nonunion sector.

Under the traditional industrial relations view, the procedural goal is achieved when workers join unions and engage in collective bargaining and the substantive goal is achieved when the collectively bargained wage is set above otherwise-prevailing wages in an unorganized labor market. But that understanding of the second goal of the Act is problematic because it is internally inconsistent, based on flawed and outdated theories of wage determination and of business cycles.

The labor market analysis at the time of the Great Depression was still rooted in the theories of Thomas Malthus and John R. Commons. Malthus claimed that population growth would always leave a pool of unemployed workers that would keep wages at the subsistence level. Commons, one of the original giants of industrial relations, extended the claim, saying that “cutthroat competition” among workers set the market wage at the wage that the “cheapest laborer” would be willing to accept. To remedy the problem, unions were needed to address the inequality of bargaining power.

The modern concept of competitive labor markets was undeveloped at this time. It was not until 1932 that John Hicks published The Theory of Wages and laid the framework for the neoclassical theory of wage determination, and it was several decades later before it became widely known and accepted. In the modern theory of wage determination, the competitive wage is the wage that equates supply and demand. Both employers and employees are “price takers”—neither exercises bargaining power. The competitive wage may be a depressed wage in terms of some norm of acceptable living conditions, but it is the market outcome. But the conventional wisdom among policymakers when labor law was being developed in the 1930s was that of Commons and not Hicks.

The business cycle language of the Act also creates problems in light of modern neoclassical economic theory. The statutory language looks to unions to raise wages to counter an ongoing deflationary cycle where declining wages result in under-consumption and increased unemployment. The under-consumption story was a neat one but there was never any solid economic support for it, and it was gradually being replaced by Keynesian economics even as the Act was passed. Keynesian economics posits that a combination of countercyclical fiscal and monetary policy could reduce the severity and length of downturns in business activity.

Elevating peace / Two alternative stories can be told in evaluating the two goals. The first story is the one told by the framers of the Wagner Act. Industrial peace is an important, clear, and coherent goal of the Wagner Act. Replacing industrial strife and unrest with industrial peace makes both employers and employees better off and has enormous benefits for social welfare. On the other hand, a violent regime of illegal strikes, riots, and the recurring exercise of police power represents a failed industrial relations system. In this story, the goal of equalization of bargaining power seems to fit neatly with the goal of industrial peace. Equality of bargaining power was required for labor disputes to be resolved peacefully. The two goals are thus complementary.

The second story reaches a very different conclusion. As a threshold economic issue, the procedural and substantive aspects of the goal of equalizing bargaining power are inconsistent. The higher the union wage, the lower the level of employment in the union sector. Thus the substantive goal of a high wage pulls in one direction, while the procedural goal of more workers covered by collective bargaining pulls in the other. There also is a potential inconsistency between the substantive goal of higher union wages and the goal of industrial peace. High union wage, relative to the nonunion wage, means greater management opposition to union demands (and to union organizing efforts generally), thus the greater likelihood of strikes.

The complexities and potential inconsistencies inherent in this second goal of the NLRA are one reason for emphasizing the more straightforward goal of industrial peace. But another reason lies in the dramatic statutory revisions of 1947. In the Taft-Hartley Amendments to the NLRA, making greater progress in achieving industrial peace was the paramount goal. At the same time, certain tactics and conduct during labor disputes were restricted, even at the obvious cost of curbing unions’ bargaining power. With the Taft-Hartley amendments, the goal of industrial peace clearly became paramount.

THE FOUR LEGAL REGIMES

To evaluate the success of the NLRA, I will treat it as one of four alternative legal regimes that have each existed in the United States at some time since the beginning of the New Deal. I will then ask which of the four is most likely to achieve the two goals. In terms of terminology, I note that the NLRA has been amended several times since the original Act—the Wagner Act—was first passed in 1935. When I use the term “NLRA,” I refer to the labor law as it exists today.

The NIRA / The first of the alternative legal regimes is the National Industrial Recovery Act (NIRA) of 1933, which was the first attempt in the United States to give workers the right to act in concert without employer interference and to encourage collective bargaining. The theme of the Act, as stated in its preamble, was to replace “free competition” with managed “fair competition.”

The NIRA’s legal structure is known as corporatism. Corporatism emphasizes cooperation among interest groups or constituencies—especially labor and capital—and between those constituencies and the government. One problem with this scheme was that unions represented only a small percentage of the private labor force at the time. Without labor unions that broadly represent employees’ interests, industry codes would likely be unbalanced, reflecting only the interests of business. To provide a countervailing power to corporations, the NIRA actively encouraged unionization. As a result, union membership grew exponentially in the period following the adoption of the NIRA.

Essentially, the economic goal of the NIRA legal regime was to cartelize industry in order to prevent price and wage competition from feeding deflation. Although the term “cartel” was not explicitly used, it was this feature that encouraged corporations to participate. At the heart of the NIRA’s labor policy was section 7(a), which required that each code recognize the rights of employees “to organize and bargain collectively through representatives of their own choosing free from employer interference.” Section 7(a) was breakthrough legislation for the union movement, providing labor the right to organize and to do so without interference from employers.

Most importantly, the NIRA held out the promise of a truly cooperative relationship between labor and capital. By equalizing pay across competing firms, unions took wages out of the price competition. The incentive to reduce wages and then prices to gain market share was no longer possible or profitable. Cutting prices would violate the NIRA’s codes of conduct. Moreover, the extra revenue provided by the higher prices could pay the higher union wages. High union wages thus did not put unionized firms at a competitive disadvantage. Management associations and labor unions worked together to form codes of behavior that prevented competitive cost cutting.

The NIRA experiment ended when the Act was declared unconstitutional by the U.S. Supreme Court. In the case Schecter Poultry Corporation v. United States, the Court held that the code-making authority conferred by the NIRA was an excessive delegation of legislative power and therefore unconstitutional. With its emphasis on fair competition, the NIRA is against the spirit of free competition embodied in neoclassical economics. And as a legal regime, the NIRA failed to work effectively. Rather than attempting to reform the NIRA, President Franklin Roosevelt made the important political decision to let it become a brief footnote in American economic history.

Wagner Act / After the NIRA was declared unconstitutional, a second attempt to calm industrial strife was made with the 1935 passage of the National Labor Relations Act, better known as the Wagner Act. The heart of the legislation, Section 7, was largely a carryover from Section 7(a) of the NIRA. Workers were given a right to join labor organizations, bargain collectively, and engage in concerted activity such as strikes without “interference, restraint, or coercion” by management. The NLRA provided a detailed set of rules for both union recognition and collective bargaining. It forbade many employer tactics that discouraged unionization, set up machinery for determining the union designated as the bargaining agent of the employees, and directed employers “to bargain collectively” with the chosen representatives in good faith.

To achieve its goal of promoting industrial peace, the Wagner Act provided for a legal strike mechanism that channeled concerted activity into a peaceful form. Employees were given the right to strike, but that right was required to be exercised in a peaceful fashion.

The Act favored collective bargaining as the preferred form of the employment relationship and favored spreading collective bargaining throughout the economy. With the Wagner Act favoring unionization, if all firms in an industry were to unionize, the competitive pressures for firms to compete based on differences in costs would be reduced. The advantages of higher prices and higher wages, promised by the NIRA, might be gained without the cumbersome structure and questionable constitutional legality of the NIRA. Again, higher prices would fund the higher wages with the associated economic inefficiencies promoted by the NIRA.

The cooperative spirit between employers and unions envisioned by Senator Wagner, however, was an impossible dream from the beginning. Without the government assistance to cartelize labor and product markets provided by the NIRA, the Wagner Act could only take wages out of competition if all firms in the industry were unionized and if wages were bargained at the industry level. Under the best of circumstances that would take time to develop. Thus, from the outset, staying nonunion under the Wagner Act gave firms much lower labor costs, negating any incentive to cooperate with unions.

Taft-Hartley Amendments / Most of the changes brought by the Taft-Hartley Amendments of 1947—the third of the four legal regimes—reduced the scope and effectiveness of the economic weapons available to the union in organizing new workers. Critically, Taft-Hartley replaced the “closed shop” of the Wagner Act with the “union shop.” Under the new rules, prospective employees did not need to be members of a union as a condition of employment. Instead, the collective bargaining agreement could require that an employee join the union and was given at least 30 days from the date of hire to join. Although the loss of closed shop status was important to unions, it was minor compared to the effect of the “open shop.” The new section 14(b) permitted states to pass “right to work” laws mandating the “open shop.” In the open shop, employees hired into a bargaining unit job did not have to join the union or pay dues. The effect of the right-to-work laws, which were especially popular in the South, was to make it much more difficult for a union to organize and sustain bargaining units across an entire industry. Competing on wages was back as a business strategy.

In addition, Taft-Hartley added a list of unfair labor practices by unions to balance the list of unfair labor practices by employers in the Wagner Act. Taft-Hartley also added a new Section 8(c) to clarify that employers had the right to express their views about unionization in response to a union organizing drive.

The main effect of Taft-Hartley was to limit the spread of unionization. The relative difficulty of organizing, as well as the ban on the “closed shop,” guaranteed that there would be a vibrant nonunion sector, especially in the “right-to-work” states that required an “open shop.” Consequently, the legal regime of the Taft-Hartley Act has a nonunion sector competing actively with a union sector.

The nonunion sector / The fourth legal regime is the patchwork of employment laws that regulate today’s nonunion sector. The employees in this sector do not have the benefits of an enforceable contract, or of “just cause”–type job security, and they do not have a bargaining agent to represent their interests before the employer. This legal regime has two components. The first is the employment-at-will doctrine, which governs the norms of the workplace. The second is a set of government mandates such as the Fair Labor Standards Act (FLSA); the Occupational, Safety, and Health Act (OSHA); and the Employee Retirement Income Security Act (ERISA); as well as Title VII and other antidiscrimination laws.

The employment-at-will doctrine effectively operates as a jurisdictional boundary. The effect of allowing an employer to discharge a worker at-will is that the employee cannot contest that decision in court. If taken literally, this rule may appear to promote employer opportunism and unfairness. Yet employment-at-will survives.

What explains the almost universal fact that the nonunion employment relationship works without use of an enforceable contract for most of its terms? One possible answer is that employers are able to exploit their superior bargaining power over employees and impose this unfair arrangement. But that begs the question of why the nonunion sector seems tranquil today, rather than having a labor force eager and ready to unionize. Perhaps nonunion employers can offer a pay and job security package that is attractive because it has lower costs than the unionized firms with which they compete. Understanding this point takes us to the neoclassical theory of the firm.

The key point here is that in addition to higher union pay, collective bargaining is a high-cost mechanism for providing worker protection against employer opportunism. Simply put, collective bargaining is very expensive because of its high transaction costs. In general, when transaction costs are high and contract enforcement is expensive, the economic relationship is brought inside the firm, where the parties are governed by the intra-firm hierarchical governance structure. From the perspective of transaction cost theories, the decision to bring relationships within the firm is the decision to opt for the intra-firm governance structure over contractual governance within markets.

With the single exception of the collective bargaining contract, the decision to bring an activity inside the firm means that the activity will not be governed in most of its particulars by contract terms. What then explains the fact that employers do not use the employment-at-will doctrine to act opportunistically and take advantage of their work force? As an empirical matter, employment-at-will is today an accepted part of the nonunion employment relationship, at least to the extent that it is not a serious topic of labor law reform at either the national or state level.

What explains the relative lack of employer opportunism in today’s nonunion sector? The answer is to be found in the unique nature of the employment relationship: it is an intensive repeat-play game. Monitoring is costly and thus incomplete. It is now well known that informal norm governance works best in such situations because self-help methods are much more effective. In this situation, it is the firm that arguably lacks bargaining power, since the remedy—increased monitoring—can be prohibitively expensive for the same reasons that contract writing is prohibitively expensive.

The employment relationship is typically marked by the parties investing in their match. Firm-specific investments create a wedge between the employee’s value to her current employer versus her value to a new employer. The contract is self-enforcing because both sides then lose their investment if the relationship is terminated early. In addition to the self-enforcing structure of norms, other factors are also at work. For instance, reputational effects can be a strong deterrent to employer opportunism.

The ultimate deterrent to employer opportunism, however, is the threat effect of unionization. A nonunion firm will become much less profitable if unionized. Wage and benefits will likely be raised above competitive levels and the firm will have the transaction costs of negotiating a collective bargaining agreement that will also impose restrictions on its ability to manage its work force unilaterally.

The second component of the nonunion legal regime is the extensive set of government mandates such as the FLSA, OSHA, and ERISA, as well as Title VII and other antidiscrimination laws. Mandates such as ERISA and OSHA serve to remedy potential problems of information asymmetries. Mandates such as minimum wages, child labor prohibitions, and discrimination-free employment serve a different function. Rather than correcting a market imperfection, they impose a public moral standard. Such regulations impose minimum standards on the theory that some market-determined outcomes are unacceptable as a matter of national policy.

ACCOMPLISHING THE GOALS OF THE NLRA

Which legal regime best accomplishes the goals of the NLRA? The NIRA receives some credit for being the first federal labor law legislation to provide for the right to engage in lawful concerted activity: both to unionize and to strike without interference from employers. As a practical matter, however, the NIRA failed on the ground, and the problems showed up almost immediately. Price-fixing proved difficult to accomplish. Noncompliance begot further noncompliance, as code-abiding business executives began to feel the pinch of competition from cheating firms. The hoped-for stable higher prices were not achieved.

The NIRA was no more successful in labor relations than it was at fixing prices. In the NIRA framework, unions and business were expected to exercise self-restraint in their bargaining demands in order to support national priorities. That did not happen. The historical record of strike activity during the brief NIRA era illustrates the failure of the law to reduce industrial strife. Instead of providing for greater labor stability, the number of workdays lost to strikes tripled over the first three years of the NIRA.

The NIRA does much better with the goal of equalization of bargaining power. On the procedural element, the NIRA succeeded because the percentage of workers from the private sector belonging to unions increased. On the substantive element, The NIRA was also successful. Although the exact premium differs by industry and over time, the evidence uniformly supports the existence of a high union wage premium over the entire period studied here.

Overall, the NIRA scores high as the first major legislation to grapple with the problems of industrial strife and unequal bargaining power. Much more statutory work needed to be done, but the NIRA was a good first attempt.

Achieving the Wagner Act’s goals / One would expect that the Wagner Act would be successful in achieving its own goals. In fact, the record turned out to be mixed. With respect to industrial peace, the Wagner Act created a legal strike mechanism that turned many strikes from violent ones to non-violent ones. This was an important change. Although less threatening to the established order, industrial strife—which had already increased during the NIRA years—increased further under the Wagner Act. Rather than bringing industrial peace, the number of strikes and lockouts nearly doubled under the Wagner Act.

There are several explanations for the worsening in industrial strife. First, particularly in the late 1930s, many new unions were forming, undertaking their organizing drives and bargaining for their first contract. Second, the legal regime was particularly favorable to unions. For example, the fact that there were no unfair labor practice standards restricting union action meant that the strike weapon could be used freely except as constrained by state law. Third, the aspirations of union leaders and workers increased along with the more favorable legal regime, and rising aspirations translated into more costly bargaining demands that were difficult to resolve without strikes.

The jump in industrial strife went along with a sharp increase in union membership. So while the Wagner Act was unable to reduce industrial strife, it was able to increase union representation. That is, while the first goal was proving unattainable, the second goal was being achieved. This underscores one of the themes of this article: the goals of the Wagner Act were potentially inconsistent. A potential inconsistency in the Act turns into an actual inconsistency once the substantive goal of equalizing bargaining power is taken into account. Concomitant with the increase in union density, the newly organized union members were able to achieve higher wages and thus gained the union wage premium. Herein lies the problem: who would pay for the higher wages?

Under the Wagner Act, and unlike the NIRA, firms would pay for the higher wages through reduced profits. Although firms might be able to pass on some of the wage increases to consumers, there is no reason to suppose that they could pass on the bulk of the increase. Consequently, employer opposition to unions was built into the Wagner Act.

In summary, the Wagner Act scores high on the goal of equalizing bargaining power. With respect to the key goal of industrial peace, however, the Wagner Act was not a success. Strikes did become less violent compared to the strikes of the late 19th century, but violence was still a frequent feature of strike activity. In addition, the level of strike activity increased dramatically, and this—combined with the continuing incidence of violence—was eventually deemed to be unacceptable. Whatever its success in promoting the bargaining power of workers, it was doomed to be replaced because it failed to achieve industrial peace.

Taft-Hartley and the NLRA’s goals / The Taft-Hartley Amendments transformed the original Wagner Act into a very different regime. It certainly changed the Wagner Act’s balance between employers and unions in favor of employers. It also supported the development of a vibrant nonunion sector in almost every industry, thus raising the likelihood of direct labor cost competition between union and nonunion companies.

With respect to the goal of industrial strife, the post-Taft-Hartley NLRA has been much more successful than the Wagner Act. The average number of strikes, adjusted for the size of employment, dropped throughout the decades following passage of Taft-Hartley. During the 1970s, there was an average of 289 strikes per year involving 1,000 or more workers. This same figure was roughly 83 during the 1980s. Since 2000, the average number of strikes per year has been around 20.

The Wagner Act increased industrial strife, but union representation grew strongly. This has reversed under the Taft-Hartley legal regime; industrial peace has been achieved, but not the equalization of bargaining power.

What accounts for the success of Taft-Hartley in reducing industrial strife? One of the theses of this article is that a key underlying factor is the growth of the nonunion sector. Nonunion companies became a factor in nearly every industry. The growth of the nonunion labor force takes us back to the second goal of the Wagner Act and my thesis that the goals are in conflict with each other. While industrial peace was finally being achieved, the gains in the equalization of bargaining power were being undone.

In summary, the Wagner Act increased industrial strife, but union representation grew strongly. This has been reversed under the Taft-Hartley legal regime; industrial peace has been achieved, but not the equalization of bargaining power. Instead, a vibrant union sector has been replaced by a vibrant nonunion sector.

Nonunion sector and the NLRA’s goals / In analyzing the success of the NLRA as amended by Taft-Hartley, one needs to address the nonunion sector as well as the union sector. The short answer to the question of whether the nonunion sector achieves the goals of the NLRA would seem to be no, at least with respect to inequality of bargaining power. There is, perhaps ironically, a longer answer that affords the nonunion sector more credit for fulfilling the public policy of the Wagner Act, and it makes the nonunion sector one of the great success stories of the NLRA. The longer answer starts by recognizing the importance of industrial peace; it ends by questioning whether nonunion employees truly lack bargaining power.

With respect to industrial peace, unlike the decades prior to the passage of the NLRA, the nonunion employment relationship is no longer a source of industrial strife. Employees apparently are not so frustrated by their inability to organize a union and get employer recognition that they take to the streets, which they did in large numbers before the Wagner Act. What has changed? One obvious answer is that the nonunion worker can now trigger the union option if the employer proves untrustworthy. Employees have a legal right to organize whenever they choose to do so. Consequently, the threat to unionize is a powerful deterrent that has likely caused the nonunion employer to act in a more trustworthy manner, living up to the accepted norms of the workplace. However, the threat effect of unionization cannot be the entire story, especially as union density in many sectors of the labor market approaches zero.

The employment relations practices of nonunion firms have also likely improved over time. Self-governing norms take time to develop and be tested for effectiveness. In the wake of the decline in union density, a consulting industry has been established that can give employers either an off-the-shelf set of norms or norms targeted to their specific employment relationship. Those norms are embodied in employee handbooks as well as much of modern human resources practice. Since labor costs are such a large component of total costs, the efficiency of the nonunion employment relationship is big business.

With respect to the equality of bargaining power, the nonunion sector lacks the collective bargaining apparatus, but it can make other claims to satisfy some aspects of the second goal of the NLRA. The nonunion sector has a governance structure in the form of the self-enforcing norms that constrain management. Self-enforcing norms work silently, through the invisible hand as it were, in terms of their adoption and retention. Although there is no formal “offer/​acceptance” process, employees show constructive acceptance when they consent to employment and then do not quit with knowledge of workplace norms. Quit rates, in the form of workers voluntarily leaving an employer, are highly concentrated in the first few years of employment. This suggests that workers do search and reject jobs that they do not like. Similarly, employers show adherence to workplace norms when they respect them, even though it is costly to them in the short run. Moreover, there is evidence that the norms of the nonunion workplace do change over time in a way that reflects changes in social norms.

Conclusion

According to my analysis, the NLRA has been strikingly successful in achieving its explicit legislative goals. My positive assessment of the NLRA rests in part on the notion that the overriding goal of the Wagner Act was to achieve industrial peace. The NLRA, as amended by Taft-Hartley, solved the problem of industrial warfare by creating a legalized regime of union representation elections and a legalized strike weapon that has been choreographed into a peaceful series of steps between the union and the employer.

Critical to the success of the NLRA is the transformation of the nonunion sector from a dysfunctional labor relations system that was an incubator for riots and violence, into one in which employees can trust the employer most of the time to enforce the norms of the workplace. The NLRA should get much of the credit for the transformation of the nonunion sector, however ironic that may be to union adherents.

In a very real sense, the union sector is a victim of the success of the NLRA in achieving industrial peace and incentivizing the emergence of a viable nonunion employment relationship. Although a goal of the NLRA was to create a vibrant union sector, it seems to have created a vibrant nonunion sector instead. Whether from the threat of unionization or simply the realization that acting opportunistically toward one’s work force is unproductive, the nonunion sector has emerged as a central component of the NLRA’s striking success.

Could a different system have worked better in generating high union employment and industrial peace? The United States tried one of the legal regimes that would have made it all work: corporatism as developed in the NIRA. In the corporatist regime, all workers could be unionized—or at least covered by the major economic terms of union agreements. That would secure both the procedural and substantive elements of the equalization of bargaining power. The workers could bargain collectively at the firm or industry level and their national or federation union would have a seat at the highest level of policymaking, thus securing industrial peace. But that was not the choice that President Roosevelt and the American voters made.

Readings

• Corporatism and the Rule of Law: A Study of the National Recovery Administration, by Donald R. Brand. Cornell University Press, 1988.
• “Labor Unions: A Corporatist Institution in a Competitive World,” by Michael L. Wachter. University of Pennsylvania Law Review, Vol. 155 (2007).
• “The Enforceability of Norms and the Employment Relationship,” by Edward B. Rock and Michael L. Wachter. University of Pennsylvania Law Review, Vol. 144 (1996).
• The State and Labor in Modern America, by Melvyn Dubofsky. University of North Carolina Press, 1994.
• “Understanding the Employment Relation: The Analysis of Idiosyncratic Exchange,” by Oliver E. Williamson, Michael L. Wachter, and Jeffrey E. Harris. Bell Journal of Economics, Vol. 6 (1975). Reprinted in The Economic Nature of the Firm, edited by L. L. Putterman and R. S. Kroszner; University of Cambridge Press, 1996.
• “Unions, Dynamism, and Economic Performance,” by Barry T. Hirsch. In Research Handbook on the Economics of Labor and Employment Law, edited by Cynthia L. Estlund and Michael L. Wachter; Edward Elgar Publishing, 2013.
• “Why Workers Still Need a Collective Voice in the Era of Norms and Mandates,” by Cynthia L. Estlund. In Research Handbook on the Economics of Labor and Employment Law, edited by Cynthia L. Estlund and Michael L. Wachter; Edward Elgar Publishing, 2013.