Читать книгу Inequality and the Labor Market - Группа авторов - Страница 27

Our legal system is failing workers. Not only have circumstances and economic change outpaced the law’s response to new challenges, but they also have revealed a gaping hole in competition law where labor is concerned. There are many factors at play, but it is increasingly clear that diminishing worker power and lower labor shares are, at least partially, the result of antitrust law’s failure to address labor markets.¹

The law has always taken a siloed approach to the world. Lawyers use the Latin term lex specialis, which means that, if both a specific and a general body of law applies to a situation, the specific body of law will govern. In looking for regulation of labor markets, antitrust law—the more general body of law—cedes the field to employment law—the more specific. As a consequence, antitrust enforcement tends to ignore issues of worker power to labor law, and instead has been laser-focused on product competition. Nowhere is this more evident than in antitrust’s historical failure to protect labor from buyer-side market power. In fact, buyer-side power in labor markets (sometimes called monopsony power) and, until recently, restraints on worker mobility or wage coordination, have been entirely ignored by antitrust law.² Though antitrust statutes were enacted to protect both consumers and workers, the law has devolved toward considering only the interests of consumers to the exclusion of all else. It has operated to undermine worker power, against the intentions of the drafters (Vaheesan 2019).

In recent history, antitrust law and economic theory have largely presumed that labor markets are competitive, and have drawn an arbitrary line between products and labor (Open Markets Institute 2018). Antitrust enforcement on mergers and conduct have scrutinized product markets from the purchaser’s and seller’s sides, but, until recently, the effect of a merger on labor markets was rarely if ever considered. Even blatantly problematic mergers, where workers might go from having two potential employers to just one, a merger to monopoly—the most alarming possibility in mergers analysis—have barely even provoked review of the labor market effects (Naidu, Posner, and Weyl 2018). This leaves workers who want to stay in their chosen field or geographic area with no leverage to bargain about their employment.

Economic theory is partly at fault, but legal developments have exacerbated the problem. For example, courts developed a legal standard for analyzing antitrust harm called the “consumer welfare standard” (Wilson 2019). The idea was that not all competitive harms would really hurt the American public, and that judges should be prudent about when to interfere with the economy. The intuition that we should cross-check potential violations to make sure there is some larger harm to the economy makes sense, but it has been over-literalized. Today, unless you can trace a particularized and direct price increase to consumers—and not to workers, small businesses, or any other conception of the American stakeholder—courts can decline to even examine antitrust violations. This myopic view often leads judges to ignore violations that are extremely harmful to consumers in the medium term, and to the economy as a whole in the long term.

When two companies merge and there is potential harm to their markets, enforcers tend to examine only the direct measurable effects on consumers of the products in those markets. When they analyze the companies’ buying power in the upstream markets, they examine only the product inputs. And, since the ultimate yardstick for whether there is a problem at all is outcomes for consumers, sometimes merging parties can point to their upstream buying power—including power over the labor market—as resulting in potential savings for the consumer. Merging parties have pointed to reduced labor input costs as an efficiency of the merger, and reduced labor costs have generally been taken as a neutral or perhaps even a good thing. Without analyzing the competitive dynamics of the relevant labor market, it has been assumed that lower input costs for labor unambiguously help the economy, even where those costs are inefficiently low or subcompetitive. As long as labor becomes less expensive, there has been a general sense that consumers will pay lower prices.³

Intuitively, it seems likely that less expensive inputs or lower wages would mean savings for firms to pass on to the consumers. But it turns out that inefficiencies and lack of competition in upstream markets have ripple effects that can harm everyone. In a competitive market, employers pay the market wage; when there are vacancies, a marginal increase in pay will follow so employers can fill those vacancies. Labor monopsonists have different incentives. If they raise pay to fill a marginal vacancy, they might also have to raise pay for their existing employees. The small increase in pay needed to attract one more worker could mean a massive swing in overall labor cost (Krueger 2017). So even if growth would generally be good for the company, they might not be able to add the workers they need specifically because of the special dynamics of controlling too much of the market.

This is an extreme example, but the same general principle applies when employers have the market power to depress wages below competitive levels. When the marginal cost of filling vacancies and growing one’s business to efficient levels diverges from the firm’s individual incentives for doing so, firms are constricted and leave jobs unfilled. Constraining inputs like labor leads to constrained outputs, and if firms are producing less of the products that consumers want, then prices for those products go up. After all, supply constraints and price increases are two sides of the same coin, economically. Fewer workers ultimately means fewer goods, and fewer goods means higher prices for the limited amount of goods available.⁴ Over time, this problem is magnified because fewer workers are incentivized to enter the field at all. The supply of qualified workers will go down, further reducing the firm’s ultimate output below efficient levels. In the end, everyone suffers except the firm with market power, which captures outsized profits.

Think: Why does America have a chronic undersupply of nurses or teachers, as well as stagnant wages (Council of Economic Advisers 2016)? In a competitive market, undersupply would lead to higher wages and increased entry to the field. If wages are inefficiently underpriced, we end up without enough nurses and ballooning healthcare costs. (Not to mention that, in the case of nurses, we end up with worse health outcomes for consumers!) This is part of the reason it is so problematic to interpret the consumer welfare standard to mean that short-term consumer prices are increased: presumed price effects could be irrelevant or misleading as to the overall effect on consumers.

Antitrust enforcement is supposed to be dynamic and to be able to keep up with the state of economic theory.⁵ But this cross-pollination is not in evidence. For example, even though inefficiency anywhere in the supply chain leads to worse outcomes for consumers, product market cases outnumber labor market cases by a factor of nearly 15, and in mergers by closer to 35. Moreover, no recent merger has been blocked on the basis of labor market effects alone (Levi 1948, 540, fn10). A quick foray into how antitrust law has developed follows.

When a company has market power—the ability to depress what it pays for goods or to increase its price above competitive levels—it causes a lot of harms, including harm caused by some firms capturing profits that would normally be spread across the economy, such as when a monopolist firm captures so much of the market that it becomes profitable to raise prices above their competitive levels, to the detriment of consumers. At its core, antitrust is concerned with preventing these harms, and allowing the competitive market to incentivize and reward innovation. However, the Sherman Act itself merely forbids “agreements in restraint of trade,” and it has been left to courts to figure out what is harmful and what is beneficial for the economy. Courts have created many cabining principles to help distinguish good agreements from bad ones. One such rule separates per se violations—that is, conduct that is so difficult to detect and likely to cause harm that it is outright forbidden⁶—from conduct that might have anticompetitive effects but that might be beneficial, although not every time.

Per se standards are critical, in part because antitrust violations are easy to cover up and highly profitable. For example, horizontal competitors—rivals in a given market—who purport to compete for the same customers, want customers to think they are fiercely competing to get their business. So, even if it were not illegal, they would naturally want to conceal a scheme to raise prices. It is good business to hide that scheme, and would be profitable to work together because rivals can charge monopoly prices.

Regular monopolists have powerful incentives to keep rivals out, especially when they have better or cheaper products. However, the only people likely to know what is going on are the same people profiting from abusing their dominance. Antitrust violations are so potentially harmful, and are so often effectively concealed, that deterring them requires heavy penalties and lower standards of proof—that is, proof that the defendant was engaging in the conduct and the liability inquiry is over (Bork 1965). And procedural short-cuts like this make a tremendous difference in how often conduct is deterred, and, therefore, how likely people are to think they will get away with it.

Throughout history, a great deal of conduct has been per se illegal. For example, it was per se illegal for companies to engage in a vertical restraint, such as when manufacturers set prices for their goods across retailers by means of an agreement between nonrivals, such as suppliers and their purchasers. Today, effectively the only agreements that are per se illegal are horizontal price-fixing agreements. Add the slightest argument that something else is going on or some wrinkle to their relationship, and you trigger the less-stringent rule of reason.

As more laissez-faire economic theories have gained ascendancy, these per se rules have been eroded. Vertical price restraints were ruled to sometimes have procompetitive justifications.⁷ Joint ventures⁸ and intra-brand restraints on trade (which likely cover a great deal of fissuring conduct) were also taken out of the per se category.⁹ Thus, more and more cases have been moved to the more complex realm of non–per se analysis.

Judges determined that, for non–per se potential antitrust violations, the courts should use a burden-shifting regime called the “rule of reason” (Bork 1965). Thus, under the rule of reason, if there is a prima facie case for anticompetitive conduct, individualized factually intensive (and resource-intensive) inquiry into the history, aims, and ultimate likely outcome for competition overall must be conducted.¹⁰ Then the defendants are allowed to offer procompetitive justifications that they assert counterbalance the anticompetitive harm. Finally, the judge is tasked with weighing the negative consequences against the positive, and to determine whether the conduct was ever an antitrust violation.

Sounds simple, right? Why wouldn’t we want more-accurate weighing of the consequences of conduct we are not sure is problematic? Well, aside from creating some uncertainty about what is legal, rule of reason is often a free pass for antitrust violators. Rule of reason cases are so complicated and expensive that they are exponentially more difficult to pursue. By the way, under current law monopoly or single-firm conduct cases are always rule of reason, because the law does not want to deter healthy, vigorous competition or success, and some monopolies are simply the result of a superior product or process. So, the law has taken the position that per se rules in this arena have too much potential for over-enforcement.¹¹

The same legal system that is so preoccupied with sparing defendants from the complexity and expense of combatting these cases is quick to multiply the complexity and expense of such actions. Ironically, this problem is exacerbated by our ever-increasing sophistication in economic analysis: the more we could theoretically prove about a situation with just a little more data and analysis, the more difficult and expensive the plaintiff’s burden becomes. Enough complication or novelty in an antitrust violation can serve as functional immunity from our antitrust laws. Judges are expected to analyze the specific factual outcomes of business arrangements as well as to do so in a way that keeps up with the state of economics, sifting through theoretical justifications for conduct, perhaps those created in hindsight. It is a vicious cycle because rule of reason cases make judges’ jobs more difficult, making them more reluctant to allow them to proceed. It is not a recipe for rational legal principles or for functional enforcement.

Labor relationships have evolved rapidly in the past 50 years from being a direct, clearly defined relationship between employees and large companies, to a blurred and complicated set of relationships between various mediating legal relationships and layers of companies, also known as workplace fissuring. In his book The Fissured Workplace, David Weil (2014) uses the example of the hotel chain business model to demonstrate fissuring. Instead of Marriott or Hyatt or another chain name above the door that employs all the workers in the hotel, in the fissured business model the hotel chain employs few workers. Instead, the hotel’s housekeepers are employed by a temp agency that contracts with the hotel chain. The front-desk workers are employed by a different contractor. And the landscapers might be independent contractors. Although all of these workers share a common workplace and probably all wear a similar uniform with the hotel chain’s name, legally they are not coworkers.

Workplace fissuring is, itself, a result of failed antitrust enforcement mechanisms around monopsony and monopoly power. Fissured entities blur the all-important line between horizontal relationships between competitors who are rivals in the same market and vertical competitors who are counter-parties in one supply chain. As a practical matter, without strong per se rules about specific conduct, firms are rightly emboldened to violate the antitrust laws without fear of enforcement. This is a large part of why fissuring has escaped enforcement: it is complicated and does not fit neatly into the few remaining per se rules. When courts took joint ventures, as well as vertical and intrabrand restraints, out of the per se category, it enabled firms to create their own captive markets where they themselves make the rules, centrally. And antitrust’s focus on products further reduces the chance that these arrangements will even be examined.

That is why a company can evade employment law by creating layers of independent contractors and firms, then regulating how they compete with each other without regard to competition law. A more reactive and aggressive antitrust regime would make employers choose: you get to be one firm and control pricing, hiring, the terms of internal competition—everything—but then you are regulated as functionally one single firm. You can be treated as a joint employer and so be responsible for the conduct of the whole enterprise. Or you can truly operate at one level of the market and your independent contractors and firms in your supply chain must be allowed to genuinely compete with each other—one franchise location must be allowed to poach employees from another location, and central rule-setting functions across supposedly independent economic firms can and should be regulated. This is the private ordering of markets, outside of our democratic controls, and therefore subject to competition law.

Today, many firms evade both regimes. They contend that they are many independent units for employment law, but face only rule of reason (at best) for their private market ordering, and little attention to labor markets from enforcers in any case. Thus, they are able to operate with virtual impunity, in contravention of the basic principles of free competition. Slap a brand name on it and direct your conduct at workers, and you have taken yourself out of the category of antitrust violations that are being enforced vigorously. Their monopsonistic labor market conduct also renders merger analyses inadequate to capture the harms of market power in the fissured workplace—intrabrand becomes a veil behind which the law has made little attempt to reach.

But turning a blind eye to anticompetitive mergers and conduct among employers is not the only way antitrust has failed workers. There is a rich history of public and private enforcement that perverts the intended purpose of antitrust law by attacking worker organizing (Vaheesan 2019). Some worker organizing is immunized from antitrust. Sections 6 and 20 of the Clayton Act establish that the Sherman Act does not apply to the existence of labor unions or to the legitimate activities of unions, such as strikes, boycotts, and pickets. These exemptions were interpreted by the Supreme Court in the 1940s to protect the activities that unions undertook to address competition in the labor market, but not competition in the product market (Greenslade 1988). Powerful interests have carefully monitored the borders of this immunity and pounced on collective action that crosses the line. And labor law does nothing to protect independent contractors; by some estimates, 30 percent of these workers are employees misclassified as contractors (Greenslade 1988, 809).

The best recent example of this perversion of antitrust law is the Seattle ordinance to provide Uber and Lyft drivers collective bargaining rights. In December 2015, the Seattle City Council passed an ordinance that set forth a process through which Uber and Lyft drivers could choose a union or nonprofit organization to be an exclusive representative for them in negotiations with the companies. Because drivers are treated by Uber and Lyft as independent contractors, they had been denied collective bargaining rights under the National Labor Relations Act.¹² The U.S. Chamber of Commerce challenged the ordinance on a number of grounds, including that it violated antitrust by allowing price-fixing among independent contractors. The Ninth Circuit held in U.S. Chamber of Commerce v. City of Seattle that the ordinance did violate the Sherman Act on its face and failed to meet the requirements of Parker immunity from the act.¹³ Thus, the effect was to leave Uber and Lyft drivers without any recourse to acting collectively to countervail the companies’ power, and to leave them stymied by both labor and antitrust law.¹⁴

Powerful interests have always been able to avail themselves of powerful tools like antitrust law to challenge their opponents, and antitrust law is certainly a powerful cudgel. After all, it allows for public and private enforcement, both criminal and civil liability, joint and several liability, civil penalties, and treble damages. And labor has rightly sought protection from these abuses over the years. But it would be a mistake to overlook the tremendous potential that antitrust law offers labor to take on consolidation, redress fissuring by rationalizing the borders of what constitutes a firm under employment law, and, with a nudge in the right direction, to ensure their mobility—a central lever in maintaining the balance of power between employers and workers. Now, when the antitrust establishment is waking up to the pernicious effects of noncompetes,¹⁵ no-poach agreements, and other wage-suppressing conduct (U.S. Department of Justice 2018), it would be a shame to throw the baby out with the bathwater.