OIT - Labour market concentration and wage inequality - A cross-country descriptive analysis

X Labour market concentration and wage inequality: A cross-country descriptive analysis Authors / Clemente Pignatti, Sévane Ananian

March / 2026 ILO Working Paper 167© International Labour Organization 2026 Attribution 4.0 International (CC BY 4.0) This work is licensed under the Creative Commons Attribution 4.0 International. See: https:// creativecommons.org/licenses/by/4.0/. The user is allowed to reuse, share (copy and redistribute), adapt (remix, transform and build upon the original work) as detailed in the licence. The user must clearly credit the ILO as the source of the material and indicate if changes were made to the original content. Use of the emblem, name and logo of the ILO is not permitted in connection with translations, adaptations or other derivative works. Attribution – The user must indicate if changes were made and must cite the work as follows: Pignatti, C., Ananian, S. Labour market concentration and wage inequality:: A cross-country descriptive analysis. ILO Working Paper 167. Geneva: International Labour Office, 2026.© ILO. Translations – In case of a translation of this work, the following disclaimer must be added along with the attribution: This is a translation of a copyrighted work of the International Labour Organization (ILO). This translation has not been prepared, reviewed or endorsed by the ILO and should not be considered an official ILO translation. The ILO disclaims all responsibility for its content and accuracy. Responsibility rests solely with the author(s) of the translation. Adaptations – In case of an adaptation of this work, the following disclaimer must be added along with the attribution: This is an adaptation of a copyrighted work of the International Labour Organization (ILO). This adaptation has not been prepared, reviewed or endorsed by the ILO and should

not be considered an official ILO adaptation. The ILO disclaims all responsibility for its content and accuracy. Responsibility rests solely with the author(s) of the adaptation. Third-party materials – This Creative Commons licence does not apply to non-ILO copyright materials included in this publication. If the material is attributed to a third party, the user of such material is solely responsible for clearing the rights with the rights holder and for any claims of infringement. Any dispute arising under this licence that cannot be settled amicably shall be referred to arbitration in accordance with the Arbitration Rules of the United Nations Commission on International Trade Law (UNCITRAL). The parties shall be bound by any arbitration award rendered as a result of such arbitration as the final adjudication of such a dispute. For details on rights and licensing, contact: rights@ilo.org. For details on ILO publications and digital products, visit: www.ilo.org/publns.

ISBN 9789220432778 (print), ISBN 9789220432785 (web PDF), ISBN 9789220432792 (epub), ISBN 9789220432808 (html). ISSN 2708-3438 (print), ISSN 2708-3446 (digital) https://doi.org/10.54394/00033457

The designations employed in ILO publications, which are in conformity with United Nations practice, and the presentation of material therein do not imply the expression of any opinion whatsoever on the part of the ILO concerning the legal status of any country, area or territoryor of its authorities, or concerning the delimitation of its frontiers or boundaries. See: www.ilo. org/disclaimer. The opinions and views expressed in this publication are those of the author(s) and do not necessarily reflect the opinions, views or policies of the ILO.

Reference to names of firms and commercial products and processes does not imply their endorsement by the ILO, and any failure to mention a particular firm, commercial product or process is not a sign of disapproval. Information on ILO publications and digital products can be found at: www.ilo.org/researchand-publications ILO Working Papers summarize the results of ILO research in progress, and seek to stimulate discussion of a range of issues related to the world of work. Comments on this ILO Working Paper are welcome and can be sent to research@ilo.org.

Authorization for publication: Caroline Fredrickson, Director, Research Department ILO Working Papers can be found at: www.ilo.org/research-and-publications/working-papers Suggested citation: Pignatti, C., Ananian, S. 2026. Labour market concentration and wage inequality:: A cross-country descriptive analysis, ILO Working Paper 167 (Geneva, ILO). https://doi.org/10.54394/0003345701 ILO Working Paper 167

Abstract This paper examines the relationship between labour market concentration and wage inequality using global survey data for the period 2006–2022. The results show that higher labour market concentration is associated with higher wage inequality, especially in the top half of the wage distribution. However, labour market institutions such as trade unions, collective bargaining and minimum wages appear to mitigate this trend. Finally, the study finds that the association between labour market concentration and wage inequality is stronger in developing countries than in developed countries.

Keywords: Enterprises; Mergers; Remuneration; Inequality; Developing countries; Collective bargaining; Minimum wage.

About the authors Clemente Pignatti is Assistant Professor of Economics at the University of Milan. Previously, he worked at Bocconi University and in the Research Department of the International Labour Organization. His research focuses on how labour market policies, social protection schemes and public institutions affect individuals' socio-economic decisions and public welfare. Sévane Ananian is an economist in the Research Department of the International Labour Organization (ILO), where he has worked since September 2021. His work focuses on wages

Organization. His research focuses on how labour market policies, social protection schemes and public institutions affect individuals' socio-economic decisions and public welfare. Sévane Ananian is an economist in the Research Department of the International Labour Organization (ILO), where he has worked since September 2021. His work focuses on wages and working conditions, inequality, and the transition to a green economy.02 ILO Working Paper 167 Abstract 01 About the authors 01 X Introduction 04 X 1 Literature review 06 The effects of labour market concentration 06 Quasi-experimental evidence 08 X 2 Data sources, definitions and descriptive statistics 09 World Bank Enterprise Survey 09 ILO Microdata collection 11 Additional data sources 13 X 3 Methodology 14 Main specification 14 Databases matching 15 X 4 Results 17 Baseline results 17 The role of labour market institutions 19 Heterogeneous analysis across countries 22 X Conclusion 24 Annex 25 References 26 Acknowledgements 29 Table of contents03 ILO Working Paper 167 List of Tables Table 1: Descriptive statistics of the HHI 11 Table 2: Descriptive statistics on measures of wage inequality 12 Table 3: List of countries and years included in the baseline specification 15 Table 4: Baseline results of the effects of labour market concentration on wage inequality 18 Table 5: Baseline results at the country level using also imputed values of HHI 19 Table 6: The mediating role of trade union membership 20 Table 7: The mediating role of collective bargaining 21 Table 8: The mediating role of minimum wages 22 Table 9: Differences in treatments effects by development status 23 Table A1: Differences in treatments effects by level of GDP per capita 2504 ILO Working Paper 167 X Introduction In its simplest definition, a labour market is considered concentrated when few firms have large market shares. 1 The few large firms operating in the market have significant bargaining power

Table A1: Differences in treatments effects by level of GDP per capita 2504 ILO Working Paper 167 X Introduction In its simplest definition, a labour market is considered concentrated when few firms have large market shares. 1 The few large firms operating in the market have significant bargaining power over individuals in the labour market due to the scarcity of outside employment options (Azar et al., 2019). This can lead to a number of distortions, ranging from sub-optimal levels of employment to wages that are kept below productivity. If new firms cannot enter the market, this situation will not be resolved without some form of public intervention. Public interventions to reduce labour market concentration can include measures targeted at firms and measures aimed at correcting market power imbalances between firms and workers, for example through the introduction of a minimum wage or support for collective bargaining. Consistent with these theoretical predictions, a series of studies have found negative effects of labour market concentration on economic and social outcomes. In particular, studies have found a negative relationship between the degree of labour market concentration and average wages (Amodio et al., forth.; Bassanini et al., 2024; Benmelech et al., 2022; Qiu and Sojourner, 2023; Rinz, 2022). The negative wage effect is particularly large for new hires (Bassanini et al., 2023), for whom there is also a negative effect on employment probability (Marinescu et al., 2021). Similar results are found in studies that have exploited exogenous changes in the degree of labour market concentration, such as those that occur after mergers and acquisitions or after policy reforms (Arnold, 2019; Prager and Schmitt, 2021; Thoresson, 2024). Studies have also found that higher labour market concentration reduces job security (Bassanini et al., 2024) and increases income inequality (Rinz, 2022). Some studies find supportive evidence for the fact that the presence of trade unions mitigates the negative effects of labour market concentration (Benmelech et al.,

labour market concentration reduces job security (Bassanini et al., 2024) and increases income inequality (Rinz, 2022). Some studies find supportive evidence for the fact that the presence of trade unions mitigates the negative effects of labour market concentration (Benmelech et al., 2022; Qiu and Sojourner, 2023). The contribution of the present paper is to provide cross-country descriptive evidence on the relationship between labour market concentration and wage inequality for a sample of more than 40 countries, with data between 2006 and 2022. In doing so, the paper aims to make the following contributions to the existing literature. First, this study examines the relationship between labour market concentration and wages in a cross-country framework that includes developing countries. With the exception of Amodio et al. (forth.), all previous studies have analysed this issue in the context of developed economies, due to the greater availability of good quality data. However, the effect of labour market concentration can be greater in developing countries. This could for instance be due to high informality and lower compliance with the employment protection legislation, which may exacerbate the negative effect of labour market concentration on wages and employment. 2 Second, the paper examines the effect of labour market concentration on wage inequality, while most of the available papers have examined the effect on wages. Indeed, the study of labour market concentration should consider the possibility that labour market concentration has different effects at different points of the wage distribution. For instance, labour market concentration could be relatively beneficial for the wages of managers and executives, while lowering 1 These market shares have been defined over different dimensions, including employment (Bassanini et al., 2023), hires (Marinescu et al., 2021), vacancies (Azar et al., 2022) or payrolls (Berger et al. 2022). 2 Of course, expanding the country coverage comes at some costs, in terms of data quality and methodology, which is presented in greater details below in the paper. For these reasons, the present analysis should be read as providing descriptive, rather than caus2 Of course, expanding the country coverage comes at some costs, in terms of data quality and methodology, which is presented in greater details below in the paper. For these reasons, the present analysis should be read as providing descriptive, rather than causal, evidence.05 ILO Working Paper 167 the wages of low-skilled workers. However, the presence of minimum wages, or other wage rigidities at the bottom of the distribution, could also mean that the wages of high-skilled workers are more affected by a reduction in workers’ labour market power. Third, the paper explores the potential role of labour market institutions in shaping the wage effect of labour market concentration. While some papers have looked at the differential effect in unionized and non-unionized sectors (Benmelech et al., 2022; Qiu and Sojourner, 2023), the present contribution looks more comprehensively at the potential mediating role of (i) trade union density, (ii) collective bargaining coverage, and (iii) minimum wages. While the literature on these labour market institutions has traditionally examined the potential distortions they can create in the market, here the hypothesis is that these institutions can help address inefficiencies arising from firms’ market power. Methodologically, we follow previous contributions and capture labour market concentration with the Herfindahl–Hirschman Index (HHI), which measures the size of firms in relation to the size of the industry. 3 For each available country, we measure HHI indexes at the sectoral level using the World Bank Enterprise Surveys (WBES). 4 This is a firm-level survey conducted by the World Bank in more than 150 countries. We then match this measure of labour market concentration to a set of wage inequality indicators computed from surveys available from the ILO Microdata collection. Finally, we obtain information on labour market institutions (i.e. trade union density, collective bargaining coverage and the level of the minimum wage) from complementary sources. 3 This means that we measure concentration in terms of the share of employment held by firms, rather than in terms of the shares

collection. Finally, we obtain information on labour market institutions (i.e. trade union density, collective bargaining coverage and the level of the minimum wage) from complementary sources. 3 This means that we measure concentration in terms of the share of employment held by firms, rather than in terms of the shares of hiring by these same firms (or other measures used in the literature, as reviewed above). See section 2 below for the reasons and implications of this choice. 4 The obtained measure of the HHI is rather imprecise and subject to bias, as discussed in greater detail in section 2. Therefore, it should be interpreted with caution, particularly when compared to other existing HHI measures.06 ILO Working Paper 167 X 1 Literature review

The effects of labour market concentration A number of studies have estimated the level of labour market concentration in a particular market and studied the effects of labour market concentration on wages and other working conditions. While the methodological details of these studies differ, most of them use administrative data sources to compute an index of labour market concentration that is specific to (i) a given geographical area, and (ii) a given industry or occupation. The main rationale behind this approach is that jobseekers in a country are not exposed to the same level of labour market concentration, but rather to a specific one that is jointly determined by their location and their occupation and/or industry of employment. Most of the papers that follow this definition of labour market concentration aim to estimate its effects on wages and other working conditions. The main intuition behind this approach is that higher levels of labour market concentration could lead to lower wages by giving employers more bargaining power. These papers are generally based on extremely rich data with precise information on both the location of individuals and firms and the detailed occupation of employment or sector of activity. Using this information, these papers argue that the observed

correlation between the degree of labour market concentration and wages is causal in nature. Most of these studies measure labour market concentration using the HHI. The HHI is calculated by taking the square of each firm’s market share in the industry and summing the resulting shares over all existing firms. The market share of a given firm is usually computed using information on either its number of employees or the number of job vacancies that the firm is posting. As such, the HHI can range from 0 to 1, with higher values associated with a higher level of labour market concentration. Generally speaking, a labour market is considered highly concentrated if the HHI is above 0.18. 5 One of the first studies of this type is conducted by Azar et al. (2022), who use information on job postings on an employment website to calculate labour market concentration for over 8,000 US labour markets defined as the intersection of commuting zones and six-digit occupations. Based on this definition, they find that most labour markets are highly concentrated. In addition, they find that vacancies advertised in more concentrated labour markets tend to be associated with relatively lower wage offers, a result confirmed in both a standard ordinary least squares (OLS) analysis and in an instrumental variable (IV) approach. These results are confirmed by a number of studies in the United States (Azar et al. 2019; Azar et al., 2020; Azar et al., 2022). By contrast, the results from Portugal are partly different. Using a methodology similar to Azar et al. (2022), Martins and Melo (2024) confirm the existence of a link between higher labour market concentration and lower wages, but much weaker than in the United States. The author interprets this finding in the light of the potential role of labour market institutions, in particular collective bargaining coverage, in reducing the negative effect of employer market power. 5 Threshold set by the Federal Trade Commission in the United States, US. A HHI between 0.1 and 0.18 is generally considered moderately concentrated.07 ILO Working Paper 167

market institutions, in particular collective bargaining coverage, in reducing the negative effect of employer market power. 5 Threshold set by the Federal Trade Commission in the United States, US. A HHI between 0.1 and 0.18 is generally considered moderately concentrated.07 ILO Working Paper 167 Nevertheless, a number of studies across European countries confirm a negative relationship between firms’ market power and wages (Arquié and Bertin, 2021). In particular, Bassanini et al. (2023) and Marinescu et al. (2021) find that in France there is a negative relationship between concentration and wages, and that this is stronger for new hires compared to incumbents, possibly because the former are more likely to negotiate their wages. Bassanini et al. (2024) also conduct a comparative analysis of labour market concentration in some European countries (Denmark, France, Germany, Italy, Portugal and Spain). They find that, after controlling for product market concentration, the magnitude of the effect of labour market concentration on wages is remarkably similar across countries. They also find that higher labour market concentration reduces job security, although the effect varies across countries. Dodini et al. (2024) demonstrate that workers affected by mass lay-offs in more concentrated labour markets experience lower earnings subsequently compared to those in less concentrated markets. These findings point to the presence of employer market power stemming from the concentration of skill demand among firms. A few studies directly test for the mediating role of labour market institutions in shaping the relationship between concentration and wages. In the United States, Benmelech et al. (2022) find that labour market concentration reduces the positive relationship between productivity and wages, while the presence of strong trade unions mitigates this effect. Qiu and Sojourner (2023) find that the negative effect of labour market concentration on wages is weaker in labour markets characterized by a relatively high share of trade union coverage. In the United Kingdom, using establishment survey data from the private sector between 1998 and 2018, Abel, Tenreyro and

that the negative effect of labour market concentration on wages is weaker in labour markets characterized by a relatively high share of trade union coverage. In the United Kingdom, using establishment survey data from the private sector between 1998 and 2018, Abel, Tenreyro and Thwaites (2020) highlight an association between lower levels of pay and higher labour market concentration for workers not covered by a collective bargaining agreement (CBA). However, this negative correlation disappears for those who are covered by a CBA. Some studies examine the effect of labour market concentration on income inequality. Results in this area are still preliminary. Lipsius (2018) observes that average labour market concentration in the United States has declined over time, and that the negative relationship between labour market concentration and wages has also weakened. This implies that labour market concentration alone cannot explain the observed decline in the labour income share in recent decades. 6 However, Rinz (2022) finds that higher labour market concentration is associated with an increase in income inequality in the United States, suggesting that, while labour market concentration cannot explain the increases in income inequality over time, higher concentration at any point in time is associated with higher income inequality. To our knowledge, the only available evidence on labour market concentration from outside of developed economies comes from Amodio et al. (forth.). Using data from Peru, they document the existence of a wage premium in the economy. Specifically, individuals earn only 70 cents for every dollar that they produce, meaning that 30 cents go to companies. Moreover, the size of this markdown is larger in markets with higher levels of labour market concentration, consistent with the fact that firms have monopsony power and use it to offer lower wages. Finally, the study also finds that the markdown decreases with the rate of self-employment in the labour market. This could suggest that the possibility of working as a self-employed person reduces the market power of firms vis-à-vis workers. 6 The labour income share is the percentage of aggregate output that goes to compensating labour.08 ILO Working Paper 167 Quasi-experimental evidence Another stream of the literature examines the effects of changes in labour market concentramarket power of firms vis-à-vis workers. 6 The labour income share is the percentage of aggregate output that goes to compensating labour.08 ILO Working Paper 167 Quasi-experimental evidence Another stream of the literature examines the effects of changes in labour market concentration on employment, wages and other working conditions. While the research questions in these papers are often similar to those reviewed above, the main differences relate to (i) the focus on the effects of changes in the level of labour market concentration rather than on the effects of a given level of concentration, and (ii) the presentation and discussion of an explicit identification strategy that exploits changes in policy regulation, or mergers between competing firms in the same industry. The first group of studies in this area includes papers that have examined the effects of the entry of Walmart retail stores in the United States since the 1970s, as the entry of new and large retail outlets can indeed lead to the exit of smaller outlets and thus increase the degree of concentration in the labour market. The available evidence shows that the entry of Walmart into a given market is associated with a reduction in prices paid by consumers (Hausman and Leibtag, 2007). However, increased competition also causes smaller stores to exit the market (Basker, 2005). As a result, the opening of a Walmart superstore is associated with lower employment levels and lower wages (Dube et al., 2007; Neumark et al., 2008; Wiltshire, 2023), consistent with Walmart exercising monopsony power (Dube et al., 2022). A more recent wave of studies examines the effects of changes in the degree of labour market concentration in a given labour market. Arnold (2019) examines the effects of mergers and acquisitions on the level of labour market concentration and the employment outcomes of workers. He finds that mergers reduce wages for workers, with the range of individuals affected depending on the size of the merger relative to the overall market. Prager and Schmitt (2021) look at the effects of mergers involving hospitals, whose employees have more limited outside employment options. The authors find that employees’ wages decrease after a merger. However, the effect is not homogeneous across the population. In particular, unskilled workers see no wage effect from the merger, possibly because the existence of minimum wages prevents downward wage adjustments for this population. Instead, highly skilled workers experience a drop in wages, but only when the merger leads to a large increase in labour market concentration. The authors also find that the effects of the merger are mitigated when there is a high level of unionization or pro-labour regulation. Finally, Thoresson (2024) looks at the effect of a change in policy regulation that entailed a reduction in the level of labour market concentration in the pharmaceutical industry in Sweden. Wages rose after the reform, in line with the fact that employers lost part of their monopsony power. These positive effects were particularly large for relatively more mobile workers, such as young workers and immigrants.09 ILO Working Paper 167 X 2 Data sources, definitions and descriptive statistics

World Bank Enterprise Survey The WBES is a firm-level survey administered in a large number of countries. The survey is answered by business owners and/or top managers who report information on their firms’ performance, including information on the number of individuals employed in the firm, workers’ compositions, investments, costs of inputs and annual sales. The WBES are generally conducted in coordination with national authorities and/or with the support of the main employers’ organizations in the country. The survey samples private-sector companies with five or more employees in manufacturing and services sectors (see below for details on the covered industries). The number of observations varies between countries, ranging from around 1,800 respondents at each wave in large economies, to 150 interviews in small

support of the main employers’ organizations in the country. The survey samples private-sector companies with five or more employees in manufacturing and services sectors (see below for details on the covered industries). The number of observations varies between countries, ranging from around 1,800 respondents at each wave in large economies, to 150 interviews in small countries. The survey uses a stratified random sampling methodology, with strata corresponding to firms’ size (5–19, 20–99 or 100 and above employees), sector of activity (manufacturing, retail and other services) and geographical region (varying across countries), and includes weights to adjust the number of establishments in the sample to the actual population size of the strata. This study uses the WBES to estimate a measure of labour market concentration for each country and year for which the survey is available. This corresponds to a total number of 305 observations from 153 different countries. The first available surveys are from 2006, while the most recent surveys at the time of the analysis are from 2022. The main measure of labour market concentration constructed using the WBES is the HHI. This is calculated in three steps. First, for each enterprise in the database, the share of employees in the enterprise relative to the total market is calculated. Second, the squares of these shares are obtained. Finally, the HHI is derived by summing these squared market shares across all firms in the economy. By construction, the HHI can range between 0 and 1, with larger values being associated with higher levels of labour market concentration. For each country and year, the empirical analysis uses measures of the HHI for two broad sectors, namely services, and manufacturing and construction. 7 This is because the level of concentration that is relevant to a particular worker is likely to be that of the sector in which he or she is employed, rather than the

economy-wide level of the indicator. 8 Computing the HHI with the WBES raises two additional issues. The first is that without information on the exact location of the enterprise or the specific sector of economic activity, we can only compute an HHI which simply differs between two main sectors. Therefore, in this study, HHIs reflect sectoral employment concentration at the national level. However, not all jobs within a 7 The survey is representative of businesses included in the following sectors (using the ISIC rev. 4 classification): (i) construction, (ii) manufacturing, (iii) wholesale and retail trade, (iv) transportation and storage, and (v) accommodation and food services activities. Based on this list of sectors, we construct two broad sectors as (i) manufacturing and construction (sectors (i) and (ii) above), and (ii) services (with all remaining sectors listed above). The choice of this regrouping ensures suffic