Financialization and income inequality: An empirical analysis

Abstract

This study investigates the effect of financialization and financial rent on income inequality across countries, including advanced and emerging markets and developing countries, in the 1998–2017 period. Employing a new measurement and dynamic panel estimation, we find that financialization measured by financial rent increases income inequality. We also find that the more monopolized banking sector is associated with higher income inequality. It suggests that market concentration in the banking sector creates ample opportunities for banks to generate excess profit and income, which leads to higher top income concentration and overall inequality. Commonly used variables for financialization such as bank income, the stock trading value, stock market capitalization, and the banking sector asset compared to the economy are also harmful to inequality.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 More recent studies that use these measures include Pariboni and Tridico (2019), De vita and Luo (2020). Huber et al. (2020) and Alexiou et al. (2021).

2 More precisely, one gets rent if he earns an income higher than the minimum that person would have accepted, defined as the income in the next opportunity (Khan, 2000).

3 The unit cost is more precisely a measure of the efficiency of the financial sector. Philippon (2012) finds that the annual cost of financial intermediation has increased in the U.S. along with the development of the financial sector over the past 30 years. It suggests that the efficiency of the finance fell because of rising rent. The unit cost is measured by value added in the financial sector as a share of GDP divided by total intermediated assets.

4 Though the deposit rate is deducted as the cost of the fund during the final accounting of ROC, we may consider deposit money as input resources of intermediation in the banking sector and subtract the deposit rate as a proxy of the next possible returns from other financial assets.

5 There is a debate regarding the causality between financialization and lower profitability. Although the decline in the profit rate could promote financialization, as some Marxist political economists argue (Lapavitsas, 2011), empirical evidence suggests that higher dividend and interest payment have an adverse effect on profitability and capital accumulation (Stockhammer, 2004; Tori and Onaran, 2018). There could be interactions between them as financialization lowered real investment, aggregate demand, and profitability, which further deepened financialization.

6 It should be noted that an increase in tax at top income does not result in a decrease in the growth rate, and thus its source is mainly rent. If these incomes were a result of their efforts, those at the top would respond by working less hard, with adverse effects on GDP growth (Philippon and Reshef, 2012).

7 There are also country-level studies using various methodologies that find that financialization increases income inequality. They include Lin and Tomaskovic-Devey (2013) on the US, Alvarez (2015) on France, and Stockhammer (2004) on the U.S., UK, France, and Germany.

8 We derive the Gini coefficient data from version 8.1 of the SWIID dataset, which covers 187 countries for the 1981-2017 period.

9 The H-statistic is a measure of the degree of competition in the banking industry. It measures the elasticity of banks' revenues relative to input prices. Under perfect competition, the H-statistic equals 1. Under a monopoly, H-statistic is less than or equal to 0. When H-statistic is between 0 and 1, the system operates under monopolistic competition (World Bank, 2019)

10 Our sample covers a country-year observation unit. The number of countries ranges from 43 to 128, and the time varies between 1980–2017 and 1998–2017, depending on model specifications.

11 The SWIID presents the Gini data, based on an estimation of the relationship between LIS GINI and all other GINI data available for the same country and year that are not included in LIS but available in other sources. These sources are income distribution data from the OECD, the socio-economic database for Latin America and the Caribbean, Eurostat, PovcalNet, and national statistical offices around the World (Solt, 2020).

12 Kus (2012) makes a financialization index using the z-score of these variables. Considering data constraints, we take banking sector assets as one component of the composite index by replacing banks' securities.

13 In the regression analysis, we remove very few extreme values or outliers of deposit rates to reduce measurement error, using the interquartile range (IQR) method. However, the regression results without excluding these do not change qualitatively.

14 We have considered a possible conditional relationship and have tested several interactions of financial rent and condition variables such as GDP per capita and others. However, we do not find any significant result of the interaction variable.

15 We also find that bank income and stock trading value are significant and positive to disposable income Gini in our benchmarking specification including the Kuznets curve.

16 The effect of the Lerner index remains positive and significant when we use this index as an alternative proxy of rent-seeking in the alternative specification without the Kuznets curve relationship, similar to table 4.