What Has Competition to Do with Inequality?

1. Introduction

Inequality has been one of the most debated issues. By far the leading explanations of inequality revolve around technological progress and globalization of trade (Alvaredo, et al., 2013), but these are not the only contributing factors. Recent empirical evidence uncovered that market power and dominance of few firms in many industries has substantial effect on work and labor share (Syverson, 2019; Eeckout, 2021a).   

Over the past 40 years, market power has transcended its traditional microeconomic domain and become a globally spread phenomenon, which does not merely affect the prices of goods, but has an impact on labour (Eeckhout, 2021). It has been suggested that market power is a unifying cause of several trends in recent decades such as declining labour share, rising margins, growing concentration, reduced (Syverson, 2019). There are multiple forces effecting market power itself where competition and competition policy plays a prominent role.

The article does not suggest that other forces such as technological progress and globalization of trade are unimportant. On the contrary, it is more likely that multiple factors simultaneously play a role. It is argued, however, that the role of technological progress and globalization is different in that they primarily contribute to the building and preservation of market power.

2. Rising inequality

There are two broad concepts of inequality: functional and personal income distribution (Dünhaupt, 2013). Functional income distribution refers to income received by factors of production such as capital and labour. Personal income distribution refers to the distribution of income between individuals regardless of the type of income (Stockhammer, 2011). Over the past 40 years, inequality has worsened by both measures (Eeckhout, 2021a; Philippon, 2019).

During most of the 20th century, in western capitalist countries labour share and capital share[i] were stable and represented around 66% and 34% of GDP respectively (Philippon, 2019). Kaldor was one of the first to identify this empirical regularity. Kaldor’s findings held up for many years and labour and capital shares maintained their respective rates of two-thirds and one-third (Eeckhout, 2021). From the 1980s onwards, and in particular since the early 1990s, there has been a fundamental shift and most developed and developing countries have seen a decline in labour share. 

Globally, labour share declined from 65% in the 1970s to 59% in 2017. The trend is most visible in advanced economies. Dao et al. (2017) show that labour share throughout 1991-2014 has declined in the 29 major economies, which collectively account for 2/3 of global GDP. According to OECD (2015), emerging and developing countries experienced even greater declines in the labour share than advanced economies, with Asia and North Africa experiencing larger declines and Latin America experiencing a more stable but still declining labour share.  Along with labour share, income inequality (personal income distribution) has worsened. From 1980 -2016, the income share of the wealthiest 10% of society increased in most parts of the world (Chancel et al., 2022, Wolf, 2016).[ii]

3. Market power

Market power is broadly understood as the ability of a firm to affect the price at which its product is sold. The stronger the capacity of the firm to raise prices above cost and generate excess profit, the higher its market power (Diez, Leigh & Tambunlertchai, 2018).

Market power can be durable once it has been acquired as a result of various economic forces and strategic behavior of firms. For example, it has been shown that high markups are maintained in part by sunk costs (Berry, Gaynor & Morton, 2019). There are three broad measures of market power: concentration level, markups and profitability.[iii] Empirical investigations have found that all three measures have gone up since the 1980s.

3.1. Concentration

When an industry is dominated by a few large firms, it is more likely that they will have market power. Concentration is measured as the market share of top firms (concentration ratio) or Herfindahl-Hirschamn Index (HHI).[iv] There is a considerable rise in concentration level over the past decades.

In a study of over 20,000 antitrust market investigations between 1995 and 2014, Affeldt et al. (2021) show that in worldwide markets, HHI increased from 2,500 in the mid-1990s to more than 3,000 in 2014. The authors highlight that existing literature understates the actual concentration level in antitrust markets, which is significantly higher and increased over time. 

Grullon, Larkin & Michaely (2019) show that between 1997-2014, HHI increased in the US in over 75% of industries, with an average concentration level of 90% higher. During the same period, in most industries (including manufacturing, retail, financial, and service sectors), the four largest public and private firms have increased their market share significantly, and the size of public firms, the biggest players in the economy, has increased threefold.  Koltay, Lorincz & Valletti (2021) studied the share of the top four companies in the five largest European economies (the UK, Italy, Spain, Germany and France) from 1998-2019, showing that concentration (share of the four largest firms) increased in 73% of European industries and more low-concentration industries shifted towards oligopoly.

The negative relationship between labour share and concentration is well-documented. Autor, et al. (2019) examine the period between 2001-2011 in 14 economically advanced countries and conclude a negative relationship between increased concentration and diminishing labour share in 12 out of 14 of the countries. The study finds that globally, labour shares have fallen in broadly similar industries, and those with the most concentration have seen the greatest declines in labour shares. These declines are predominantly accounted for by the reallocation of sales between firms rather than decreases within firms. Similarly, in OECD countries, the industries with the largest concentration growth had the steepest decrease in labour share. According to Barkai (2020), in the US, the industries that experienced larger increases in concentration also experienced larger declines in labour share.

3.2. Markups and profitability

Markups (price a firm charges over the unit cost) are the commonly used measures of market power. It is based on variable costs and do not account for fixed costs (De Loecker & Eeckhout, 2021). This means that a firm may have high markups to recover the fixed costs, rather than because it has market power associated with high profits (Diez, Leigh & Tambunlertchai, 2018). Therefore, to measure market power it is necessary to assess profitability along with markups.

De Loecker & Eeckhout (2021) studied the evolution of markups[v] globally for over 70,000 firms spanning 134 countries. The authors conclude that the global markups rose from 1.1 in 1980 to 1.6 in 2016. Companies are now selling 60% more over their marginal cost. 

Average markup in the US began to rise abruptly, from 1.21 in 1980 to 1.54 in 2019 (Eeckhout, 2021). Markups have increased from roughly 1-1.2 to 1.5-1.6 in Europe, North America, and Oceania during the same period. In developing countries of South America and Africa, markups have increased moderately, but they had higher initial levels. The rise in markups is not confined to any specific sector and the increase is visible in all industries from textile to retail and tech (De Loecker & Eeckhout, 2021). Importantly, for many firms, markups did not change. For the most part, the dominant firms at the technological forefront, which account for a large share of economic activity, enjoy high markups (OECD, 2018).

Along with markups, the global profit rate sharply increased from 2% of sales in 1980 to 8% in 2016 (Eeckhout, 2021). The trend is visible in Europe, the US and developing economies. A study by Diez, Leigh & Tambunlertchai (2018), covering 74 economies from 1980-2016, finds  that markups are strongly correlated with profitability at the firm level, indicating that the surge in market power  relates to increasing markups.[vi] An increase in profitability as a result of rising markups is also confirmed by IMF’s (2021) recent study on market power. The findings establish that bigger markups result in higher earnings and this is not due to higher overhead costs (De Loecker & Eeckhout, 2021).

There is a negative relationship between increasing profitability, markups and labour share. According to Barkai (2020, p.2424) in the US, “a decline in competition and an increase in pure profits have played a significant role in the decline in the labour share.” Similarly, according to Cairo & Sim (2020, p.1), “the profit share is negatively correlated with labour share, and the degree of correlation is strong: -0.91 over the 1980-2018 period.”

3.3. Market power and labour share

There are two main channels by which market power affects labour share: firms hire fewer workers and they pay them less (Eeckhout, 2021).

The crucial part of the market power story is that industries are dominated by a handful of firms (De Loecker, Eeckhout & Mongey, 2021). This enables them to sell their products at high markups and reduce the quantity sold. Firm optimization decision dictates that if firms can produce and sell fewer products at a higher price they will require few inputs, primarily labour (Eeckhout, 2021b). Because market power is widespread, economy-wide demand for labour falls and the effect is a decline in the labour share. According to Eeckhout (2021), this phenomenon is visible in the decline in labour force participation in the US where, in the case of men, the inactivity rate (labour force participation) increased from 3% in the 1960s to 11% in 2020.

The indirect effect of market power is the decline in wages (the general equilibrium effect). Despite part of the workers becoming inactive, the majority cannot remain idle and have to accept jobs. Because the firms with market power demand less labour and market power is widespread, the wages decline. Due to this reason, in the US for example, the median worker’s wages have stagnated since the 1980s (De Loecker et al., 2021; Eeckhout, 2021).

3.4. Other explanations

Other prominent explanations for declining labour share are technological improvements and the globalization of trade.[vii]

According to Karabarbounis & Neiman (2013), technological advances, and in particular improvement in information and communications technology (ICT), affect labour share and capital share by reducing the relative price of the investment goods, resulting in decreased capital cost and prompting the firms to substitute labour with capital. This explanation has been questioned by Barkai (2020), who studied the evolution of capital share, labour share and pure profits during 1984-2014 and concludes that, from 1984-2014, capital share of gross value added fell from 32% to 25%. Despite decreasing capital costs, firms did not respond with increasing spending on capital goods. Instead, the rising concentration and markups have increased profits at the expense of both labour and productive capital.

The globalization of trade argument, another factor explaining the declining labour share, is based on the Stolper-Samuelson theorem, which predicts that abundant factors will benefit from the trade. In advanced economies which specialize in capital-intensive industries, globalization will benefit capital, while in developing countries, with labour-intensive sectors, globalization should boost labour share (Stockhammer, 2011). Empirical evidence does not, however, support this prediction. In addition to advanced economies, labour share declined in relatively labour-abundant countries such as India, Mexico and China, among others (Karabarbounis & Neiman, 2013). Besides, the decline in labour share is visible in most sectors, including utilities, retail and wholesale trade, where the impact of international trade is more limited (Autor et al., 2019).

Despite some criticism, both technological advancements and globalization played significant a role in the declining labour share. However, as will be shown, they have an impact primarily on market forces (such as scale economies), which help firms build and maintain market power.

4. Drivers of the market power

A key finding of macro market power literature is that many industries are shifting towards oligopoly. Typically, the dominant firms are at the technological forefront and use relatively little labour. The relocation of economic activity to these firms is one of the underlying causes of declining labour share (Autor et al., 2019). The question is what forces are behind market power itself. Most explanations focus on two factors: decline in competition and scale-biased technological advantage.

4.1. Insufficient competition

Increasing mergers and lax antitrust enforcement over the past decades have contributed to the emergence of dominant firms. Mergers and acquisitions are much less likely to be challenged by the authorities in advanced economies, which has further consolidated market power. For example, some economists claim that in the US, Facebook should not have been allowed to buy Instagram and WhatsApp (Eeckout, 2021b) and similarly, European beer giants Annheuser Busch, InBev, and Miller should never have merged to become the dominant company in the market.

From 1985- 2016, global M&A activity surged to more than 10 times its initial level, and markups increased by 30 percentage points during this period (De Loecker & Eeckhout, 2021). Even though this does not prove causality, the relationship is noteworthy.

In the US, excessive tolerance of mergers in manufacturing resulted in the worsening of competition (Baker, 2017). Philippon (2019) explicitly attributes this to declining labour share. Similarly, Koltay, Lorincz & Valletti (2021) demonstrate that given the concentration trend in many European industries, it is hard to conclude that enforcement is becoming tougher. They find that, for some years, decreasing enforcement has been associated with rising concentration.

Killer acquisitions (incumbent firms buying potential entrants to prevent competition) have also been blamed for the worsening of competition. The practice is documented in the tech and pharmaceutical industry (Berry, Gaynor & Morton, 2019).

M&A activity is also related to increased common ownership. Fichtner, Heemskerk & Garcia-Bernardo (2017) discover that asset management funds such as Vanguard and BlackRock are the biggest shareholders in around 90% of the most prominent publicly listed US corporations. Backus, Conlon & Sinkinson (2017) connect the common ownership of firms to increased markups and find that in the US common ownership influenced the increase of markups from 1.2 to 1.56 from 1980-2017.

Given the global rise in market power, there has been a renewed focus on antitrust. Competition law can be viewed from two perspectives: on the one hand, lenient antitrust enforcement is blamed for contributing to the rise in market power (Grullon, Larkin & Michaely, 2019). Inorganic growth via mergers creating a dominant position is one of the causes of market power. At the same time and regardless of a separate explanatory power of antitrust enforcement, it remains one of the most effective tools, which can be deployed to address the problem. This holds true even if the dominant cause lies elsewhere.

4.2. Dominance by technology

Technological innovation has several effects: it increases the efficiency and productivity of firms; however, it also contributes to the formation of market power by creating barriers to entry and exploiting economies of scale.

One of the important drivers of market power is high initial sunk costs (Eeckhout, 2021b). Today’s large corporations routinely require a massive up-front investment in customized software. Where servers and hardware may be of a general character, the IT systems of these corporations are based on proprietary software and corresponding organizational and technical skills, which are not easily replicable (Sanders, 2016). Bessen (2020) provides the example of Walmart, which made considerable investments in technology to manage and monitor warehouses and inventory. Walmart’s IT system is substantially different from general, “off-the-shelf” systems, and because it is not available to competitors, is hard to replicate and provide a sustainable competitive advantage. Similarly, Inditex (the parent company of Zara) incurred substantial costs to develop information technology for logistics management (Eeckhout, 2021b). These companies have grown organically through investments in technologies and not primarily via mergers. Technological superiority by large upfront fixed costs is not feasible for small competitors. This creates barriers to entry granting market power to the firms.

Another source of dominance is economies of scale. Autor et al., (2019) point out that scale-biased technological change leads to a “winner-take-most” scenario. Large investments in information technology allow for economies of scale, which results in lower unit costs and which, in turn, strengthens market power. Technological progress increasingly favours first-mover advantage, resulting in the quick consolidation of dominance (Eeckhout, 2021b). The first company to set up a logistical warehouse, a distribution network or a digital platform gains a decisive advantage.

The importance of technology in production is not only characteristic of firms in advanced economies. Large manufacturing firms in developing countries like Tanzania and Ethiopia are surprisingly capital/technology intensive too. Diao et al. (2020) explain this finding by globalization, which had the effect of homogenizing the adoption of technology globally.

5. Conclusion

According to three broad measures - concentration, markups and profitability – there is some evidence indicating that since 1980s market power has increased, which is connected to the declining labour share.

The article suggested that two forces drive market power: firstly, insufficient enforcement of competition has enabled firms to grow in size by merging without challenge, acquiring nascent competitors and being common owners in major companies; secondly, technological advancement has made firms more efficient, but at the same time, erected barriers to entry, amplified the effect of economies of scale and created “winner-take-most” dynamics, while globalization has homogenized the uptake of technology around the globe irrespective of the level of development of a country.

The rise in market power has important implications for competition policy, which is increasingly viewed as an important tool to moderate the occasional failures of the market. This is not to suggest that competition law should bear more than it has been designed to do. It is not the author’s view that antitrust should be burdened with goals and purposes, which is inimical to its raison d'etre. At the same time, more targeted, prudent regulation, which will support the market, legitimate business interests and competition on the merits, may be a better course of action. This is already happening in the EU, which recently adopted new legislation (Digital Markets Act and Digital Services Act) to effectively address the market power in the tech sector.