The hidden source of inequality: market power

BRIEFING NOTE
Market Power and Inequality

Source:
Ennis, S. Gonzaga, P. and Pike, C (2019) “Inequality: A hidden cost of market power,” Oxford Review of Economic Policy, 35(3): 518–549. https://doi.org/10.1093/oxrep/grz017

Executive Summary

This research paper examines the impact of market power on wealth and income inequality in eight OECD countries. By developing a new economic model and analyzing empirical data, the authors suggest that a lack of competition tends to increase the wealth of the wealthy and decrease the income of the poorest. The approach suggests that the top 10% may derive 10-21% of their wealth from market power. The study quantifies this redistributive effect and explores the implications for competition policy, highlighting that consumer welfare norms may have fairer distributive consequences than total welfare norms. The findings indicate that illegitimate market power may be a significant driver of economic disparities, warranting increased policy attention. The proportion of effect that comes from market power that is legitimate (like patents) as opposed to illegitimate (like cartels) remains unknown.

Key words:

Wealth inequality
Income inequality
Market power
Monopoly
Competition

What is the topic?
• This paper investigates the relationship between market power and income/wealth inequality across several OECD countries. The central argument is that market power, manifested in the form of increased mark-ups, acts as a hidden mechanism that exacerbates wealth and income disparities due to differential ownership of assets.
Approach
• The authors develop a theoretical model and calibrate it with cross-country data to estimate the redistributive effects of market power. They distinguish between the current state with observed market power and a counterfactual hypothetically competitive steady state with reduced mark-ups. Their model provides an original and elegant view of the relationship between these two. The authors argue that market power can also lead to inequality. They develop a novel model for showing and quantifying such effects, and apply the model to data.
• The model has the advantage of parsimony, intuitive relationships, and reliance on widely available data.
Why does this matter?
Explaining the extent of potential wealth or income effects from market power helps to link the markups literature to the broader literature on inequality.
• Rising Mark-ups and Rising Inequality: The paper acknowledges recent research indicating a substantial increase in mark-ups (price over average cost) in the US and Europe over recent decades. Studies by De Loecker and Eeckhout (2017), Weche and Wambach (2018), and Gutiérrez and Phillippon (2016) are cited as evidence.
• Recent re-search, such as De Loecker and Eeckhout (2017), shows that mark-ups have increased substantially in the US between 1950 and 2014, and suggests that this can be attributed to market power.
Concepts
• Legitimate vs. Illegitimate Market Power: The authors clarify that not all market power is undesirable. Some degree of market power, arising from patents, trademarks, and brand differentiation, provides incentives for innovation and investment. TO the extent wealth and income inequality arise from legitimate market power, this may yield substantial overall social benefits. In contrast, when originating from illegitimate market power, the nature of the problem is different. Competition laws typically target the "abuse" of market power rather than its existence.
• Significance of Illegitimate Market Power: Despite the hidden nature of much anti-competitive behavior (e.g., cartels, anticompetitive regulation), the paper argues that its economic impact on inequality is likely significant, with cartels alone perhaps accounting for £200-600 billion per year.
Model and Methodology
The paper employs a comparative static analysis, contrasting the current state with market power to a hypothetical competitive steady state. The model assumes that the primary impact of monopolies is to raise prices, while real output remains constant in this comparison. Key variables in the model include mark-ups (µ), labour income share (αL), and saving propensities (average s̄ and marginal s′).
For this purpose, the authors make a comparative static analysis between two alternative steady states: the first is the current state where business owners have market power and the second is a hypothetical steady-state where markets are competitive.
To isolate the redistributive effect of market power, the authors consider that the only impact of monopolies is to raise prices, while real output remains constant.
The paper builds up from fundamental macroeconomic identities, with a variable for market power added (µ, the excess markup). It then notes how variables in these identities differ in the presence of ‘excess’ markups.
• Fm = µFc, Ym = µYc, Cm = µCc.
Redistributive Effects of Market Power: The model suggests that market power leads to a transfer of wealth and income from consumers to business owners. Population groups with a higher share of wealth than income (typically the wealthier) benefit from market power, while those with a higher share of income than wealth (including much of the middle and lower classes) experience a net loss.
Equation (6) states that the difference in income shares for group i, with and without market power, is an increasing function of the difference between the wealth share and income share of the group, and declines in the labour share (multiplied by the mark-up) times the wealth differential.
Intuitively, market power generates a transfer effect from consumers (in proportion to the income they earn) to business owners (in proportion to the wealth or capital they hold).
Impact on Wealth Shares: The paper's simulations across eight OECD countries (Canada, France, Germany, Japan, Korea, Spain, UK, US) indicate that reducing excess mark-ups would lead to a decrease in the wealth share of the top percentiles and an increase in the wealth share of lower percentiles. The magnitude of this effect varies across countries depending on the level of excess mark-ups and the initial wealth distribution, as well as the extent to which the saving rate on extra income created from market power (s’) is treated as marginal income or inframarginal, s̄.
For example, in the US, with s′ = 1.5s̄, eliminating excess market power is estimated to reduce the wealth share of the top 1% from 36.85% to 34.18%.
This paper provides evidence that market power can contribute substantially to wealth inequality, augmenting wealth of the richest 10 per cent of the population on average by 12–21 per cent.
Impact on Income Shares: Similarly, reducing market power is projected to increase the income shares of lower income groups and decrease the income shares of higher income groups.
• The poorest 20% are estimated to experience an increase in their income share by 11 per cent or more."
• For instance, in the UK, a 1% reduction in mark-ups is estimated to increase the income of the poorest 20% by 0.12 percentage points.
Estimation of Excess Mark-ups: The study uses sector-specific mark-up data from Høj et al. (2007) for 17 countries (excluding Luxembourg). To estimate "excess" market power, they compare each country's sector mark-up to the lowest observed mark-up in that sector across all countries. This "minimum mark-up" is considered a proxy for a more competitive environment. The difference is the "excess mark-up." This approach recognises that some degree of markup is normal, and that this varies across industries.
• "The estimates of excess mark-up by country-sector are calculated, for each of the 18 sectors, as the difference between the actual mark-up and the lowest observed mark-up across all countries in the sample for each sector."
• The average excess mark-up across the studied countries ranged from 3.9% (UK) to 10.6% (Germany).
Importance of Saving Behavior: The model's results are sensitive to assumptions about saving behavior, particularly the marginal propensity to save (s′) relative to the average saving rate (s̄). Different ratios of s′/s̄ are considered to illustrate the range of potential effects.
• The equations show that a reduction of market power will increase (decrease) the income and wealth of population groups whose income share exceeds (is less than) the corresponding wealth share.
Policy Implications
The findings potentially support a greater focus on consumer welfare standards in policy decision-making, as total welfare standards might overlook significant distributional impacts that disproportionately benefit a small, wealthy segment of the population.
• The paper suggests that addressing illegitimate sources of market power could be a lever for reducing inequality without necessarily hindering innovation.
• Arguably, these findings may inform the case for focusing on consumer welfare standards for policy decision-making, as total welfare standards can yield substantially different policy outcomes, and distributional impacts, for the benefit of only a relatively small share of the population.
Further Considerations
• The model makes simplifying assumptions, such as constant real output when prices change and a constant marginal propensity to save across wealth groups (though this is noted as a conservative assumption).
• The mark-up data used are historical (1975-2002), which might not fully reflect current market power dynamics. However, the authors argue these historical mark-ups played a role in shaping current wealth distributions.
• The study focuses on the price effects of market power and does not explicitly model efficiency losses (deadweight loss) and their impact on inequality.
This paper provides a compelling theoretical and empirical argument for considering market power as a significant, and often overlooked, driver of income and wealth inequality in developed economies. The findings suggest that policies aimed at fostering greater competition could have meaningful redistributive effects.

Quotes
• "This paper provides evidence that market power can contribute substantially to wealth inequality, augmenting wealth of the richest 10 per cent of the population on average by 12–21 per cent... and reducing the income of the poorest 20 per cent of the population by 11 per cent or more."
• "Much anti-competitive behaviour by firms is hidden, because it is illegal. However, the business activities of discovered cartels already show a broad range of affected sectors. Accounting for both prosecuted and undiscovered cartels... Ennis (2014) suggests cartel commerce may amount to about US$2 trillion per year, and that anti-competitive price increases from cartels are in the range of US$200–600 billion per year."
• "The groups of the population who are typically harmed by market power are the 0 to 80th percentiles, and interestingly the harm appears to be particularly accentuated on a middle class comprised somewhere between the 20th to 60th percentiles."
• “Some degree of product market power is desirable to provide sustained incentives to innovate. Patents, trademarks, and brand differentiation, for example, often involve creation of market power accompanied by a positive impact on incentive to innovate."


Paper Summary Initial Draft By NotebookLM