The Effect of Capital Share on Income Inequality: Identifying the Time Patterns ∗

Oğuzhan Akgün, Ezgi Özsöğüt · 2025

Read The Effect of Capital Share on Income Inequality: Identifying the Time Patterns ∗ by Oğuzhan Akgün, Ezgi Özsöğüt (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-02371.txt.

Introduction

Oğuzhan Akgün and Ezgi Özsöğüt wrote a paper from 2025 on how the capital share of income feeds inequality, and how that link shifts over time. The pair built matching measures for fifty-six countries over four decades. Capital share came from national accounts, and top income shares from a global inequality database built on the same frame. Their method lets the pass-through strength vary by country and year.

On average, a one-point rise in the capital share lifted the top five percent's income share by about a fifth of a point. Rich countries showed a steady pass-through while both capital and labor inequality rose. Emerging economies showed a growing pass-through with rising capital inequality. A third group showed a fading pass-through and rising labor inequality. Across all, shifts in the capital share explain about half the rise in inequality over the period.

Catalog record

CREATOROğuzhan Akgün, Ezgi Özsöğüt
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE75 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-02371

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The Effect of Capital Share on Income Inequality: Identifying the Time Patterns ∗ Oğuzhan Akgün, Ezgi Özsöğüt Abstract This study explores the link between the capital share and income inequality over the past four decades across 56 countries. Calculating the capital share from national accounts alongside top income share data from the World Inequality Database, which is based on the Distributional National Accounts methodology, we ensure the consistency in the theory and measurement. Employing a structural econometric approach, we account for heterogeneous and time-varying transmission coefficients from the capital share to personal income inequality. Our findings reveal that a one percentage point (pp) increase in the capital share raises the income share of the top 5% by 0.17 pp on average. Advanced economies show a stable transmission coefficient with rising capital and labor income inequality, while emerging economies experience an increasing transmission coefficient alongside growing capital income inequality. In contrast, a third group exhibits a declining transmission coefficient and rising labor income inequality. Overall, changes in the capital share account for approximately 50% of the rise in income inequality, underscoring its pivotal role over the last four decades. Keywords : Functional income distribution, Grouped time-varying coefficients, KMeans clustering, Personal income distribution, Shapley decomposition, Top income shares. JEL classification : C23, D31, D33. ∗ This paper developed as a chapter of the Ph.D. project of Ezgi Özsöğüt. We thank Ezgi Özsöğüt’s Ph.D. advisors, Florin Bilbiie and Romain Rancière, for their feedback on earlier versions of this paper, as well as jury members Luca Fornaro, Tobias Broer, and Mirko Wiederholt for their comments. Special thanks to Caroline Betts for her careful review and valuable suggestions as the main reviewer. We also thank the seminar participants at Galatasaray University Economic Research Center (October, 2024). The usual disclaimer applies. Email: [email protected] (Corresponding author, O. Akgün) 1 Introduction In a dual-class society, where the population is divided into wage-earning workers and wealthy capitalists who own capital, the link between personal income distribution and the capital-labor split in national income is straightforward. However, such societies arguably no longer exist. Although the capital remains concentrated in the hands of high income earners, in modern capitalist economies, different income groups earn both capital and labor incomes ( Milanovic, 2017a ) . This weakens the assumption of an obvious positive and strong relationship between the capital share and income inequality prompting us to ask some previously unremarkable questions: how does the capital share impact income inequality in different countries, and how does this relationship evolve over time? Once the theoretical relationship linking the capital share to income inequality is established (e.g., Atkinson & Bourguignon, 2000a ; Atkinson, 2009a ; Milanovic, 2017a ) , the most direct way to quantify this link is by applying an inequality decomposition using microdata sources ( Lerman & Yitzhaki, 1985a ; Milanovic & Yitzhaki, 2002a ; Amarante, 2016a ; Nolan et al., 2021a ) . These sources include surveys and what we refer to as fiscal data throughout the analysis, constructed using micro-level tax records or tax tabulations. Relying on country-level data sources, some studies have successfully applied these methods (e.g., Giangregorio & Villani, 2024a ; Aaberge et al., 2018a ; Atkinson & Lakner, 2021a ) , but extending the analysis to a cross-country framework is challenging due to the lack of high quality microdata in most countries. Even when the data are available, there are usually significant methodological gaps both in fiscal data and surveys, complicating cross-country comparisons. Another strand of the literature applies decompositions using harmonized surveys from the Luxembourg Income Study (LIS) (e.g., García-Peñalosa & Orgiazzi, 2013a ; Francese & Mulas-Granados, 2015a ) , but despite substantial efforts in standardizing definitions across countries, LIS data still faces important limitations in measuring income inequality accurately. Flores (2021a) shows that both surveys and fiscal data systematically underestimate capital income, and hence income inequality, due to measurement errors and conceptual differences between microdata and macrodata sources. According to the author’s calculations, surveys and fiscal data in the United States capture only 20% and 30% of capital incomes on average, respectively, while they capture up to 70% and 80% of labor incomes. Such underestimation of capital incomes, even in a country with exceptionally high quality individual data, leads to biased estimates of income inequality, making decomposition-based analyses unreliable for exploring the effect of the capital share on income inequality. The rest of the empirical literature, on the other hand, largely relies on reduced form approaches to estimate the effect of the capital share on income inequality ( Daudey & García-Peñalosa, 2007a ; Francese & Mulas-Granados, 2015a ; Schlenker & Schmid, 2015a ; Bengtsson & Waldenström, 2018a ; Erauskin, 2020a ) . While these models offer flexibility, they often fail to capture the true structural relationship between the two variables. This limits their capacity to explain the underlying mechanisms and complicates the interpretation of their findings ( Flores, 2021a ) . Furthermore, in most cases, the measurement inconsistency persists because the dependent variable, income inequality, is still calculated using microdata (i.e., surveys or fiscal data) and the right-hand side variable, the capital share, is derived from macrodata (i.e., national accounts). Bengtsson & Waldenström (2018a) , for instance, which is arguably the study most closely related to ours, relies on top income shares calculated based on fiscal income . Although these were the longest series covering a relatively large number of countries at the time of the study, they do not fully align with the income definitions used to calculate the capital shares from national accounts. As the authors acknowledge, this inconsistency possibly bias their estimates but this limitation stems from the study’s focus on long-term relationships and the construction of historical series, which it successfully accomplishes. On the empirical side, the study assumes a log-linear relationship between the capital share and inequality, which cannot fully capture the theoretical link between the two variables. Dividing the sample into sub-periods provides further insights on parameter shifts over time but remains limited in addressing broader time-varying effects. The main contribution of the present paper is building an empirical framework addressing these shortcomings of the previous empirical studies and closing the gap between the theoretical and the empirical literature. Specifically, we directly estimate the structural relationship between the capital share and income inequality while ensuring the consistency in the measurement of both variables. To measure inequality, we use the top income share series from the World Inequality Database (WID) calculated using the Distributional National Accounts (DINA) methodology proposed by Blanchet et al. (2024a) . When rigorously implemented, the DINA methodology ensures that all types of income recorded in national accounts are properly accounted for and distributed to individuals, providing consistency in the income definitions used to construct capital shares and top income shares. The data set that we construct covers the last four decades of up to 56 emerging and advanced countries. This effort of building a large and internally consistent data set of capital shares and income inequality represents a simple yet significant improvement in the measurement over the previous studies. In terms of the econometric methodology, our contribution is the estimation of a structural equation, while accounting for the heterogeneous and time-varying nature of what we refer to

Read the complete text