The impact of labour market institutions on income inequality: evidence from OECD countries

ABSTRACT

Using data from 35 OECD countries for the 1993–2017 period, this paper shows that stronger Labour Market Institutions (LMI), such as trade unions and bargaining coverage, contribute to a decrease in income inequality measured by the Gini coefficient. Additionally, there seems to be a positive interaction effect between unions and bargaining coverage. The marginal impact of trade unions (bargaining coverage) is enhanced by the presence of high levels of bargaining coverage (trade unions). Notwithstanding, this impact seems to be higher for bargaining coverage. On average, an increase in unions’ density (bargaining coverage) by 1% leads to an estimated decrease in inequality by 0.30% (0.35%), ceteris paribus.

KEYWORDS:

• Inequality
• labour market
• panel data
• OECD

JEL CLASSIFICATION:

• C23
• D63
• J50

Acknowledgments

We would like to thank Andrea Garnero for useful comments and suggestions and Sandra Lacerda Graça Oliveira for editorial support. The research of the first author has been financed by Portuguese public funds through FCT - Fundação para a Ciência e a Tecnologia, I.P., in the framework of the project UID/ECO/04105/2019.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Additionally, understanding the inequality phenomenon itself is also challenging: ‘Increasing inequality over the past three decades highlights a major puzzle in political economy’ (http://www.people.fas.harvard.edu/~iversen/PDFfiles/Alt&Iversen_2016.pdf, p. 21).

² Group 1 is composed of Iceland, Sweden, Denmark, Finland, Belgian, Norway, Luxembourg, Italy, Slovenia and Austria.

³ Group 2 is composed of Greece, Australia, Germany, Portugal, Netherlands, Spain, and France.

⁴ Group 3 is composed of Canada, Switzerland, Chile, Czech Republic, Estonia, United Kingdom, Hungary, Japan, South Korea, Mexico, New Zealand, Poland, Slovakia, Turkey and the United States.

⁵ Our indicator consists on the average between the ratio calculated using administrative and survey data.

⁶ From the 37 OECD countries, there are no data available for Colombia and Latvia. The panel is unbalanced because there is more information available for some countries than for the others.

⁷ The GMM model was estimated with Stata’s xtabond2 command proposed by Roodman (2009).

Additional information

Funding

This work was supported by the Fundação para a Ciência e a Tecnologia [UID/ECO/04105/2019].