Workers, Wages, and Mergers: A Back-to-Basics Guide

In recent years there have been several articles arguing that high industry concentration results in lower wages. While some papers have focused only on broad relationships between concentration and wages, others have begun to try to link wage changes specifically to merger activity. These analyses have been accompanied by calls for antitrust enforcement to consider more closely the effects of mergers on labor. For example, a recent request by the Federal Trade Commission and the Department of Justice for comments on updating the agencies’ merger guidelines explicitly asks about the appropriate treatment of monopsony power and labor markets. Consistent with the increase in public-facing expressions of interest in labor issues, practitioners are increasingly fielding questions from regulators about potential labor effects from mergers.

This article discusses the ways in which a merger might affect workers and their wages in order to identify the necessary—although not always sufficient—conditions where a substantial lessening of competition could arise. It serves as a starting point for practitioners to understand and evaluate potential theories of harm based on both economic theory and observations from empirical literature. This article is especially focused on the importance of evaluating the complete picture of any proposed transaction—in both output markets and labor markets—prior to drawing conclusions about the overall impact of a transaction on workers or their wages.

Market Power and Monopsony Power

We begin by describing the basic economic theory of monopsony. The term “monopsony” describes a market with a single buyer. In a perfectly competitive market, the price firms pay to acquire inputs is not affected by how much of the input each firm acquires. A monopsonist, by contrast, can affect the price of the input by choosing the quantity it purchases. While a market with a small number of buyers in which each buyer has a degree of power to affect price through its choice of input quantity might technically be called an oligopsony or oligopsonistic, in this article we use the term “monopsony” and “monopsony power” more generally to also cover the multiple-buyer scenario.

Monopsony can occur in any input market, including labor. For example, as Carlton and Perloff explain:

If there is a competitive labor market, each firm takes the wage rate as given, and the marginal cost of hiring one more worker is simply the wage rate. Now suppose there is only one local employer (buyer of labor services): a monopsony. . . . [It] faces an upward-sloping supply curve for labor. In order to hire an extra worker, the monopsony must not only pay that worker a slightly higher wage rate but also pay all its other workers a slightly higher wage rate, because only by raising the wage can extra labor be induced into the marketplace. . . . Because the marginal outlay curve lies above the supply curve, the monopsony hires fewer workers . . . than would a competitive market . . . The monopsony wage rate . . . is below the competitive wage rate.

The Horizontal Merger Guidelines’ (“HMG”) discussion of the effect of mergers on the buying side of the market applies to firms that purchase labor inputs by hiring workers. As the HMG notes:

Mergers of competing buyers can enhance market power on the buying side of the market, just as mergers of competing sellers can enhance market power on the selling side of the market. Buyer market power is sometimes called “monopsony power.”

The HMG also provides a useful example:

Merging Firms A and B are the only two buyers in the relevant geographic market for an agricultural product. Their merger will enhance buyer power and depress the price paid to farmers for this product, causing a transfer of wealth from farmers to the merged firm and inefficiently reducing supply. These effects can arise even if the merger will not lead to any increase in the price charged by the merged firm for its output.

Although the provided example focuses on physical inputs, labor is, in economic terms, an input, and the HMG discussion applies equally to labor. Applying the HMG example to labor, one could picture two firms that are the only employers in the relevant geographic market for a particular labor input (noting that the labor market can be quite different than the downstream product market, as we discuss in the next section). Their merger could enhance market power and depress the price (i.e., wages) paid to labor, causing a transfer of wealth from workers to the merged firm and reducing the supply of labor. It is important to note that both price and quantity are affected by monopsony, just as both price and quantity are affected by monopoly. As noted in the HMG, those effects on the upstream market (labor) need not be associated with any change in downstream price.

Although the HMG indicates that the agencies employ essentially the same framework in analyzing input markets as for traditional downstream monopoly concerns, economic analysis of labor markets raises some important distinctions as we discuss below.

Labor Market Definition

Just as with a traditional downstream product-market analysis, the standard starting place to analyze a merger that could affect labor is to consider whether there is a meaningful change in concentration in a properly defined labor market. Although the effect of a merger on the downstream product market may ultimately be informative to the overall analysis of the net effect of the merger on workers and wages, it is inappropriate and could lead to incorrect conclusions to assess potential labor effects based only on analysis of a downstream product market. The relevant market for purposes of analyzing a merger’s impact on labor is not the downstream product market, but the relevant market for labor. Labor markets and product markets can and often do have meaningful differences.

First, upstream labor markets need not look anything like the downstream product markets, and may well be much broader or much narrower. A downstream product market is defined relative to the alternatives available to consumers. An upstream input market is defined relative to the alternatives available to suppliers—in this case, to workers. The alternatives available to employees of a given manufacturer may be largely unrelated to the products made by the manufacturer. For example, almost all firms of any size have payroll departments. Consider a merger of the only two companies that manufacture widgets, and assume that widgets constitute a relevant product market. The fact that there are only two widget manufacturers, and that they are merging to monopoly, may be completely irrelevant with respect to the employees in the payroll departments. There may be a large number of other firms that also have payroll departments and hire payroll workers in that labor market. The skills of payroll employees are in no way unique to widget manufacturing. Thus, the merger of the two widget manufacturers is unlikely to have any meaningful impact on the market for payroll employees. Indeed, the labor market for such employees may well be much broader than payroll departments, as there are any number of other types of firms that employ, for example, accountants.

Second, even if a given firm only competes in one downstream product market, it may well have employees in more than one labor market. For example, hospitals may employ orderlies, nurses, and doctors. The employment alternatives available to each of them likely differ. Labor requiring less specialized skills in particular may have a broad range of alternatives available. As the labor skills required for a particular profession become more specialized, however, those alternatives may narrow. As they narrow, the possibility that a merger within that industry could affect employer market power may increase.

Third, the geographic labor market may differ from the geographic product market. The geographic labor market could be broader, but could also be narrower. For example, markets for concrete might be small in scope geographically because of the limited radius within which concrete can be transported (setting aside the ability of firms outside that geography to enter), but employees at one concrete firm may be easily able to move within a city, or between cities. On the other hand, a software firm with a single physical location may offer its products worldwide through digital distribution, but face a more localized labor market. Or, with remote work being what it has been since the COVID-19 pandemic, the firm may face a national or even global labor market.

These potential distinctions between the labor and product market(s) have a variety of implications for merger analysis and how a merger might affect workers—both positively or negatively. Of particular interest are the effects of the transaction on the output market as well as the effects of the transaction on the market for labor and how those effects may be interrelated (or not). Below we consider a few stylized examples to highlight various considerations that may arise in different cases: (1) a merger that creates monopoly power in an output market, (2) a merger that creates monopsony power in a labor market, (3) a merger that results in non-labor production efficiencies, (4) a merger that results in labor-specific production efficiencies, (5) competition in the output market from firms with alternative production technologies, and (6) instances where the monopsony conditions may not be met.

Market Power in an Output Market

Consider a merger between two neighboring corn farms: Acre Farm and Barn Farm to create Corn Co. Assume that the merger increases the market power of the combined firm in an output product market for corn. Economic theory suggests that, as a result and absent countervailing effects such as efficiencies, the price of corn goes up, and the quantity of corn produced goes down. But what happens to labor?

First, there is an important distinction between the impact of a transaction on workers of the combining firms versus the impact on labor overall. In a scenario where all else is equal and there are no efficiencies resulting from the merger, we would expect Corn Co. to need fewer workers to produce a reduced quantity of output. We would also expect the price for corn to rise, since the combined firm by assumption has market power. The impact on wages and overall employment, however, is less clear. If the labor market in which the workers at the farms participate is competitive, there should be no or minimal impact on wages. A key insight here is that the labor market in which the farmworkers participate need not be limited to firms participating in the product market for corn. For example, the corn farms might compete for labor with farms growing wheat or any number of other crops. It is also possible that not all firms participating in the product market are competing for the same pool of labor. For example, firms might have competing products but use substantially different production technologies that call for different skill sets. Perhaps corn is in the same product market as other fruits like tomatoes, but corn producers may draw from a different labor market than tomato producers because one can be picked with machinery that workers need to know how to operate, while the other is still picked by hand. (More on this in a later example.) Whether it is possible for the merger to affect wages depends, therefore, on the degree of competition in the relevant labor market, which may differ in both degree and participants from the product market. In other words, even if the merger augments the market power of the combined firm with respect to output, there need not be an adverse effect on wages.

Monopsony Power in the Labor Market

What if, instead of creating market power in the sale of corn, the merger created monopsony power with respect to farm workers? What happens to the output market? Again, one should be aware that merging firms’ output can change when market output does not. By the definition of monopsony power, such a merger would reduce the output (i.e., corn) of the merging firms. This effect arises because the merging firms (by assumption) face an upward-sloping labor supply curve and pay a single wage (per position)—producing more corn becomes incrementally costlier from a labor standpoint. The reduction in output also has two effects. First, it moves the firm down the labor supply curve, thus lowering the wages paid to workers. Second, the firm internalizes the effect of changes in wages on inframarginal workers because, if the firm were to hire an additional worker, it would cause wages to rise for all workers. Thus, notwithstanding the fact that the merger lowers wages, it increases the marginal cost of labor that the firm realizes and thus may put upward pressure on corn prices (downward pressure on the merging firms’ corn output), depending on the production function (i.e., the monopsonized labor might not be a marginal input into downstream output; for example, research and development labor in the industry might be monopsonized, but such labor might not be a direct marginal input into actual production). This is a counterintuitive result, but an important one to be aware of. Nonetheless, if the downstream market is competitive and the merging firms lack market power in the downstream market, then, by definition, downstream prices will not go up (or not significantly) and downstream output will not go down.

The relationships between the upstream and downstream markets, and the overall effects on wages, downstream prices, and consumer welfare, can be affected by other considerations, which will depend on the specific facts of the case. We consider several scenarios below.

Non-Labor Efficiencies

Suppose that the merger of Acre Farm and Barn Farm generates some production efficiencies. For example, as a merged firm, Corn Co. can more efficiently load its trucks so that the cost of shipping a bushel of corn is reduced. That marginal cost reduction means that the merged firm can more profitably produce corn. Corn output increases, the price of corn decreases, and consumers are better off.

The output increase could also be beneficial to farm workers. If Corn Co. can profitably produce more corn, it may need more workers to produce the increased quantity. If the upstream labor market in which Corn Co.’s workers participate is highly competitive, this may make no difference to wages. However, if Corn Co. has some monopsony power, then the increased demand for labor from the reduction in marginal cost will have at least some offsetting impact on the monopsony, and could result in a net increase in labor used. If monopsony power stems from the merger, then the overall impact of the merger on labor can be ambiguous and require a close analysis of the countervailing pressures.

Labor-Specific Efficiencies

Even mergers that create cost savings by reducing the marginal reliance of the merged firm on labor can—on net—benefit workers. For instance, suppose Acre Farm and Barn Farm collectively use 20 worker hours to produce 1000 bushels of corn. By merging, Corn Co. can use just 15 worker hours to produce 1000 bushels of corn. As a result, the firm has a lower marginal cost of production. There is, therefore, both resulting downward pressure on wages from the reduced demand for labor, but also upward pressure on wages because the firm has an incentive to increase its output of corn (and it requires more labor to produce the incremental output), and the marginal productivity of the workers is higher. The net effect on wages could be negative or positive, depending on the relative degree of impact. And if the market for labor in which Corn Co.’s employees participated were perfectly competitive, it would have no impact at all since Corn Co.’s choice of how many workers to hire would not affect wages.

However, a critical point here is that reduced labor usage at a specific firm due to increased efficiency of production does not adversely affect competition. The economic concern about monopsony is inefficiency. Producing products more efficiently—using fewer resources for a given level of production—is what innovation and competition are all about. As an example, the shift from crafting guilds to assembly lines adversely affected the crafting guilds, but did not reduce competition for consumers or labor overall—nor was it an exercise of monopsony power.

Competition in the Output Market from Firms with Alternative Production Technologies

Consider a case in which the merging firms have monopsony power in the relevant labor markets but lack market power in the output markets. Such a case may arise when the merging firms compete with firms that operate either in different labor markets (e.g., if the product market is national, but the labor market is local) or with a different form of production (e.g., one production technology relying primarily on labor, and one relying primarily on capital). In such a case, the merger could put downward pressure on the merging firm’s wages and upward pressure on the merging firm’s marginal costs of labor (for the reasons described above) but have no effect on downstream prices or consumer welfare.

Where Monopsony Conditions are Not Met

An important assumption in the discussions above is that the merging firm must set a single wage for each class of workers. It is this condition that causes wages to fall, but the marginal cost of labor to rise as a result of the merger, because the employer cannot engage in wage discrimination between employees. Similarly, on the monopoly side, inefficiency arises because the monopolist cannot price discriminate between consumers. If a monopolist could charge different consumers different prices, each based on their willingness to pay, then there may actually be no output reduction from the monopoly. The same economic logic applies for monopsony. For example, consider a firm that can pay different wages to new versus existing workers. Such a firm could potentially pay a higher wage to attract a new marginal worker without having to raise the wages of its existing workers. The ability to price (wage) discriminate has complex welfare implications, but can break the connection between wages and the marginal costs of labor. In that case, there can be a transfer of rents to the monopolist (monopsonist) from consumers (workers) but no inefficient reduction in output (employment), i.e., consumers can be harmed because of paying higher prices but no economic inefficiency is created.

Beyond Monopsony: Merger Effects on Labor from Shifting Bargaining Power

In order for a merger to substantially lessen competition for labor, there must be a mechanism by which the merger creates or augments monopsony power. However, in recent years, concerns about the impact of mergers related to relative bargaining power (even in the absence of monopsony power) have been raised. Bargaining effects are distinguished from monopsony effects because monopsony is associated with a reduction in the amount of labor, whereas bargaining is not. This is an important point worth unpacking.

While the specifics vary, theories related to bargaining often rely on how changes in the options available to employees affect negotiations with employers. This sort of bargaining framework, roughly speaking, addresses how gains are split between employees and employers—that is, given a pie of a certain size, how is the pie divided between the negotiating parties? Bargaining outcomes can be affected by the “threat points” of the two negotiating parties—what are their best alternatives if they fail to reach a deal. Anything that improves, or worsens, a party’s threat point will improve, or worsen, their bargaining position. In the labor context, for example, an employee who quits might know that a firm across the street is hiring. If, however, that outside firm disappears, the employee’s next best option might be inferior, worsening his or her bargaining position. This could happen in a merger if, for example, the merger eliminated an attractive alternative source of employment for some employees.

There is a key distinction between monopsony and bargaining theories in that bargaining theories are not associated with a reduction in output (and may actually lead to increased downstream output). The HMG notes, “[r]eduction in prices paid by the merging firms [for an input such as labor] not arising from the enhancement of market power can be significant in the evaluation of efficiencies from a merger. . . .” Monopsony, like monopoly, involves an economically inefficient reduction in output. A change in bargaining power does not.

Complex welfare issues can arise if the merger reduces bargaining power for workers, with no other efficiencies. In such a case, a plausible outcome is that the reduced worker bargaining power reduces wages, with the reduced labor costs being passed through, in part, to customers in the form of lower prices and/or better products. This is because, unlike in the monopsony case, the employer would face a lower marginal cost of labor in this situation. Therefore, downstream prices might be lower and firm profits higher, but worker wages lower. Such a case would require regulators and practitioners to weigh the (possibly dispersed) benefits to consumers against the (possibly concentrated) harms to workers from lower wages.

In general, antitrust seeks to protect competition overall rather than the interests of individuals. For example, a merger that is substantially procompetitive overall (e.g., because it leads to large efficiencies and does not materially harm competition) would not generally be blocked because an individual or group of individuals (e.g., someone with a strong idiosyncratic affinity for a brand that is eliminated by a merger) is harmed. However, given the increasing focus on bargaining-type theories of harm, and potentially complex interacting effects, practitioners should consider them when assessing mergers in addition to traditional monopsony theories.

Applications and Issue Spotting

Bearing in mind the caveat that the results of a product-market analysis may have little or no relation to any potential concerns on the labor side, when should a practitioner or enforcer raise concerns about the potential labor effects of a merger? Furthermore, in some mergers there are countervailing forces to any potential price increases, such as how large, sophisticated buyers may be able to resist price increases. Are there equivalents on the labor side?

Here, the empirical literature is limited, but there is a leading paper on the topic that provides at least some guidance, suggesting that, although monopsony concerns can arise, the circumstances are fairly limited, and there are countervailing forces that can exist on the labor side as well. For example, some research, such as the article discussed next, has suggested that unionization may help insulate workers from potential anticompetitive effects from a merger.

The leading paper on this topic is authored by Prager & Schmitt (2021). The authors studied hospital mergers and provide useful guidance as to when and where issues may arise in relevant labor markets.

First, the paper looked at multiple groups of employees, distinguishing between unskilled workers and skilled workers (particularly those with skills tied to the industry, i.e., nurses). They found no merger effects on unskilled workers whatsoever. That is consistent with the discussion above about unskilled labor likely being part of a broader labor market, such that any given merger is unlikely to have any meaningful impact on a labor market for workers without industry-specific skills.

Second, they looked at different changes in concentration in markets comprised of all labor at the hospitals. They found effects only for mergers in the top 25% of changes in concentration (i.e., the top quartile of changes in HHI). The highest quartile—the only one where effects were found—had an average pre-merger HHI of 4,580 and an average change in HHI of 2,764. The next-to-highest quartile had an average pre-merger HHI of 3,205 and an average change in HHI of 618. The authors found no effect for this quartile. Even within the top quartile, the authors only found effects on skilled workers, not on unskilled workers. This suggests that, although labor issues can arise as a result of firm consolidation, as a practical matter they are most likely to occur only with substantial changes in concentration, and only for labor markets that are relatively narrow, i.e., for labor with specialized skills and limited employment alternatives outside of the immediate industry.

Third, even when looking at mergers with the largest changes in concentration and labor with skills focused on a particular industry, the authors found countervailing effects. In particular, the authors only found harm to workers in right-to-work states (states that prohibit mandatory union membership in some manner). As the authors note, unions are thought to have less power in wage negotiations with employers in right-to-work states. The effects disappeared entirely in states without right-to-work laws. The authors also looked directly at unionization levels and found statistically significant reductions in the estimated wage effects where unionization levels were higher. The authors concluded that “high levels of unionization appear to meaningfully attenuate the estimated post-merger reductions in wage growth.”

Given the great interest in this topic, additional empirical analyses are likely forthcoming and practitioners may be well advised to keep abreast of this developing area.

Conclusion

Antitrust enforcers are focusing more on labor issues than in prior years, and practitioners may expect to be asked about such issues when presenting to regulators. For an initial review, practitioners should consider whether (1) there are distinct groups of employees with skills tied specifically to the industry of the merging parties, (2) there would be a significant change in concentration as a result of the merger in terms of employment opportunities (as opposed to within the product market), and (3) whether there are collective bargaining arrangements present, or the likelihood of them becoming present, that might impact negotiations. In addition, analyses should not end without considering the overall effect of the merger on demand for labor. Efficiencies, for instance, could mean that a more efficient firm actually increases demand for labor even while the firm makes more efficient use of labor. Practitioners may also wish to examine the merger-specificity of any theory of harm (i.e., is any observed correlation between concentration and wages in an industry plausibly related to competition concerns, as opposed to other causes such as a declining industry or a shift from labor to capital) and to keep abreast of new empirical research on the effects of merger on wages.

Finally, the interaction between labor market effects and overall welfare effects can be complex and is relatively underexplored in the economics literature. Depending on the conditions, labor and product market effects can be reinforcing (both going in the same direction), offsetting (going in opposite directions), or isolated (there is an effect in one market but not the other). The offsetting case is likely to be the most challenging as it raises novel, and difficult, questions about the standard for evaluating overall welfare effects that balance the welfare of consumers and workers—two groups who might be fully, partially, or not at all overlapping—and the tools that are widely used today for evaluating mergers are not designed for that case.