Financial Development, Inequality and Poverty
Some International Evidence
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

Contributor Notes

Authors’ E-Mail Addresses: sbennaceur@imf.org, zrx0212@gmail.com,

This paper provides evidence on the link between financial development and income distribution. Several dimensions of financial development are considered: financial access, efficiency, stability, and liberalization. Each aspect is represented by two indicators: one related to financial institutions, and the other to financial markets. Using a sample of 143 countries from 1961 to 2011, the paper finds that four of the five dimensions of financial development can significantly reduce income inequality and poverty, except financial liberalization, which tends to exacerbate them. Also, banking sector development tends to provide a more significant impact on changing income distribution than stock market development. Together, these findings are consistent with the view that macroeconomic stability and reforms that strengthen creditor rights, contract enforcement, and financial institution regulation are needed to ensure that financial development and liberalization fully support the reduction of poverty and income equality.

Abstract

This paper provides evidence on the link between financial development and income distribution. Several dimensions of financial development are considered: financial access, efficiency, stability, and liberalization. Each aspect is represented by two indicators: one related to financial institutions, and the other to financial markets. Using a sample of 143 countries from 1961 to 2011, the paper finds that four of the five dimensions of financial development can significantly reduce income inequality and poverty, except financial liberalization, which tends to exacerbate them. Also, banking sector development tends to provide a more significant impact on changing income distribution than stock market development. Together, these findings are consistent with the view that macroeconomic stability and reforms that strengthen creditor rights, contract enforcement, and financial institution regulation are needed to ensure that financial development and liberalization fully support the reduction of poverty and income equality.

I. Introduction

The beneficial role of financial development in economic growth has been well documented; however, the literature on the nexus of financial development and income distribution is still nascent. Theories on the effect of financial development on income distribution offer conflicting predictions: one strand of the literature proposes an inverted-U relationship between finance and income inequality, while the other predicts a linear relationship.

Greenwood and Jovanovic (1990) predict a nonlinear relationship between finance and inequality, wherein the distributional effect of financial development depends on the level of economic development. At early stages of development, only the rich can access financial services because of the fixed cost of joining the financial coalition, resulting in wider income inequality. As the economy develops, the financial system becomes more accessible and affordable to the poor because human capital replaces physical capital as the main driver of growth. Galor and Zeira (1993) and Galor and Moav (2004) posit a linear relationship between financial development and income distribution. They suggest that financial deepening eases credit constraints, which benefits low-income groups through the channels of human capital and capital accumulation.

Although theory provides conflicting conclusions on the finance-inequality nexus, empirical works suggest that financial development contributes to reducing poverty and inequality. Cross-country evidence from Beck and others (2004), Beck and others (2007), Honohan (2004), Li and others (1998), and Rajan and Zingales (2003) suggests that expanding private credit can stimulate income growth for the poorest quintiles and reduce income inequality, strongly refuting the position of Greenwood and Jovanovic (1990). One similarity of these empirical works is the use of the ratio of private credit to GDP as a measure of financial development. Such an indicator covers only one dimension of financial development: financial system depth while overlooking access, efficiency, and stability.

More recent papers attempted to include other dimensions of financial development. For example, Claessens and Perotti (2007) and Demirguc-Kunt and others (2008) found evidence supporting the importance of access to finance in reducing poverty and inequality. Jeanneney and Kpodar (2011) establish that financial instability worsens poverty; and Kunieda and others (2011) find that financial integration aggravates income inequality by benefiting the most privileged. Similarly, Furceri and Loungani (2015) study the impact of capital account openness on inequality and find that liberalizing domestic financial systems can aggravate income inequality, both in the short and medium run.

This paper will help distinguish between the conflicting views on the link between finance and income distribution, by assessing the impact of the different dimensions of financial development on both the level of income inequality and the level of poverty, using a large sample of 143 countries taken from the period 1961 to 2011. There are three key interrelated findings on the global sample. Strengthening financial access, depth, stability, and efficiency contributes to reducing income inequality and poverty, which is robust to different datasets and measurements. The results suggest also that financial sector liberalization, particularly capital account liberalization, widens inequality and the poverty gap. Financial institution development exerts a stronger impact on income distribution and poverty than financial market development.

The remainder of the paper is structured as follows: Section 2 describes the data and methodology. Section 3 presents the empirical results as well as the robustness checks, and Section 4 offers a conclusion.

II. Data and Methodology

A. The sample and variables

Our sample contains data on 143 countries, both developing and developed. Though not all the estimations cover the same time period, the sample spans from 1961 to 2011. We use measures of inequality, poverty, and financial sector development that have previously been used in the literature. Income inequality and poverty indicators are included as dependent variables and data are from the World Bank’s inequality and poverty database. The proxy for income inequality is measured by the commonly used Gini coefficient, which is a relative ratio of the areas on the Lorenz curve diagram. It is scaled from 0 to the 100th percentile in our dataset. Zero percent represents a perfectly equal outcome, in which each individual receives the same level of income; 100 percent represents extremely unequal distribution, where one person takes all the income in the economy. Though the Gini coefficient, to some degree, reflects the distribution of income, it is unable to show the welfare of the low-income group: the poverty level may be reduced with or without the Gini coefficient decreasing. To understand the role of financial development in combating poverty, the poverty gap index is also used in our estimations. The poverty gap measures the average income shortfall of the poor individual from the poverty line ($1.25 a day).2

To cover the five dimensions of financial sector development (access, depth, efficiency, stability, and liberalization), 10 variables from the Global Financial Development Database (GFDD) are used. Each dimension is measured by two variables, one related to financial institution development and the other to financial market development.

For access to financial services, we choose bank accounts per 1,000 adults and value traded of the top 10 trading companies to total value traded. The former measures bank access, while the latter reflects access to financial markets.

Financial depth is measured by two indicators: banks’ private credit to GDP and the stock market’s total value traded to GDP, which are the most widely used indicators for financial deepening. Higher values suggest deeper financial institutions and stock markets.

We select the net interest margin and the stock market turnover ratio as measures of financial efficiency. High net interest margins suggest low bank operating efficiency, while high turnover ratio (stock traded/capitalization) reflects an efficient financial market.

The stability of the financial system is measured by the ratio of regulatory capital to risk-weighted assets and the volatility of the stock price index. A high level of regulatory capital implies that banks have a lower probability of default, whereas a higher volatility of stock prices is indicative of a more unstable financial market.

For financial liberalization, two proxies are used: domestic liberalization and external liberalization. Using the financial reform database from Abiad, Detragiache, and Tressel (2008), we aggregate the index of credit control, interest rate control, entry barriers, and privatization to proxy for domestic financial reform. To measure external financial liberalization, we use the ratio of consolidated foreign claims of BIS-reporting banks to GDP, with larger values suggesting a more liberalized financial system.

Finally, we control for several other variables that have been previously used as determinants of poverty and inequality: real GDP per capita, government expenditures to GDP, trade openness, and the inflation rate. Real GDP per capita is included to control for the economic growth effect, as the literature suggests a strong relationship between income distribution and economic development. The coefficient on GDP per capita is expected to be negative, because lower inequality and poverty are associated with a higher income level. Similarly, negative signs are expected on the coefficients of government expenditure to GDP and trade openness, which are included to capture the benefits of public spending and openness to foreign trade. According to Easterly and Fischer (2001), the coefficient on the inflation rate is expected to be positive, because inflation hurts the poor more than it hurts the rich.

B. Methodology

We follow the basic regression specification from the income distribution and financial development literature:

Ginii,t=α+βFDi,t+γ1Yi,t+γ2Infli,t+γ3Tradei,t+γ4Govi,t+ɛi,t(1)
Povgapi,t=α+βFDi,t+γ1Yi,t+γ2Infli,t+γ3Tradei,t+γ4Govi,t+ɛi,t(2)

In this equation, Ginii,t and Povgapi,t, represent the Gini coefficient and the poverty gap, respectively. FDi,t is the key explanatory vector that we are interested in; it covers the 10 indicators of financial development previously described: β is expected to be negative3, which implies that higher financial development can lower inequality and poverty. Yi,t is the log of GDP per capita used to control for the wealth effect, and we expect γ1 to be negative. Infl, Trade, and Gov are a set of control variables representing inflation, trade openness, and government size. Following the literature, γ2 is expected to be positive; γ3 and γ4 are expected to be negative.

The relationship between financial development and income inequality and poverty might be a case of reverse causation. For example, a lower level of poverty implies that financial services are already more affordable and accessible to the poor, and thus, stimulating the development of the financial sector. Similarly, a narrower poverty gap or less income inequality might also promote economic growth according to the inverted-U pattern for the impact of income distribution on economic growth. Therefore, controlling for the possible reverse causation and simultaneity bias is essential for studying the impact of finance on inequality and poverty.

To control for endogeneity and reverse causation, we use instrumental variable (IV) regressions. Two types of instruments are used. The first set includes the lagged values of the endogenous variables (second lags and higher are used to avoid autocorrelation with the current error term). In the second set, we use instruments based on the theoretical and empirical finance and growth literature, such as ethnic fractionalization, linguistics, religious composition, and legal systems.

To check the validity of these instruments, we use Hansen’s J-test of the over-identifying restrictions. Under the null hypothesis, the instruments are uncorrelated with the error term, and the excluded instruments are valid instruments.4 A rejection of the null hypothesis invalidates the instruments. We use also the LM under-identification test to check whether the excluded instruments are correlated with the endogenous independent variables. A rejection of the null indicates that the model is identified5.

Furthermore, this paper examines the financial development effects with consideration to the income levels of each country. Countries here are divided into three groups: low-income, middle-income, and high-income. Following equations (3) and (4), the regression results can tell how different the financial effects are in the three groups.

Ginii,t=α+β1FDi,t+β2FDi,t×Dlow+β3FDi,t×Dmid+Dlow+Dmid+γ1Yi,t+γ2Infli,t+γ3Tradei,t+γ4Govi,t+ɛi,t(3)
Povgapi,t=α+β1FDi,t+β2FDi,t×Dlow+β3FDi,t×Dmid+Dlow+Dmid+γ1Yi,t+γ2Infli,t+γ3Tradei,t+γ4Govi,t+ɛi,t(4)

Dlow and Dmid are dummies for the low-income group and middle-income group. If they both equal zero, β1 is the coefficients of finance on income distribution of high-income countries. Analogously, β1 + β2 shows the finance effect of low income countries, whereas β1 + β3 is the effect in middle-income countries.

To study how the quality of institutions can affect the finance-inequality-poverty nexus, we employ the rule of law as the indicator. Using the same method, equations (5) and (6) are created to test the impacts of governance on finance and income distribution. If β2 and β1 are both negative, a better quality of institution tends to amplify the reducing-inequality (poverty) effect of financial development. Conversely, if they show different signs, the quality of institution may reduce the financial effect on income distribution.

Ginii,t=α+β1FDi,t+β2FDi,t×Rule+Rule+γ1Yi,t+γ2Infli,t+γ3Tradei,t+γ4Govi,t+ɛi,t(5)
Povgapi,t=α+β1FDi,t+β2FDi,t×Rule+Rule+γ1Yi,t+γ2Infli,t+γ3Tradei,t+γ4Govi,t+ɛi,t(6)

III. Empirical Results

Tables 1 and 2 report the descriptive statistics and correlation coefficients. Most of the variables are negatively correlated with the Gini coefficient and the poverty gap, reflecting possibly favorable effects of finance on income distribution;6 however, the correlation of income distribution, with inflation and domestic/external financial liberalization, suggests a widening of income inequality and therefore more poverty.

Table 1:

Summary Statistics

article image
Table 2:

Correlations

article image

Tables 3-7 list the effects on the Gini coefficient and poverty gap from all five financial aspects. This paper performs both OLS and IV regressions for each financial variable. In the IV results, suggested by the Hansen J-Statistics, we are unable to reject the null hypothesis that our instruments are uncorrelated with the error terms for all regressions, meaning that the instruments are appropriate.

Table 3:

Effects of Financial Access on Income Inequality

article image
Table 4:

Effects of Financial Deepening on Income Inequality

article image
* p<0.10, ** p<0.05, *** p<0.01Note: parentheses report standard errors
Table 5:

Effects of Financial Efficiency on Income Inequality

article image
* p<0.10, ** p<0.05, *** p<0.01Note: parentheses report standard errors
Table 6:

Effects of Financial Stability on Income Inequality

article image
* p<0.10, ** p<0.05, *** p<0.01Note: parentheses report standard errors
Table 7:

Effects of Financial Liberalization on Income Inequality

article image
* p<0.10, ** p<0.05, *** p<0.01Note: parentheses report standard errors

The impact of financial access on income inequality and poverty is shown in Table 3. Results for the global sample using OLS (Column 1) and IV (Column 2) estimators suggest that increasing the number of bank accounts per 1,000 adults can reduce income inequality. Columns (3) and (4) report results on value traded in the top 10 trading companies as a market access variable. As larger values indicate less access to the financial market, the results show that value traded in the top 10 trading companies is less likely to affect income inequality. Columns (5) to (8) report similar effects of financial access on poverty. Both OLS and IV results support the poverty-alleviating effect of improving banking access. On average, an additional banking account opened per 1000 people tends to reduce the poverty gap by a percentage point of 0.007. In contrast to the results on access to financial institutions, none of the coefficients on the market access variable are shown to be statistically significant in reducing poverty.

Table 4 reports results on the impact of financial deepening on inequality. Coefficients on the ratio of private credit to GDP are negative and highly significant at the 1 percent level in both inequality (Columns 1 and 2) and poverty (Columns 5 and 6) regressions, reflecting a beneficial effect of financial deepening in the global sample. A 1 percentage point increase in private credit to GDP ratio tends to reduce the Gini coefficient by more than 0.041%; and reduce the poverty gap by a percentage point of 0.019. Similarly, all coefficients on the ratio of stock market total value traded to GDP are negative and significant. These findings are consistent with Clark and others (2006), Beck and others (2004), and Beck and others (2007). They support the inequality-reducing effect of financial deepening from both financial institutions and stock markets.

Table 5 reports the results on the relationship between financial efficiency, income inequality, and poverty. The coefficients on net interest margin are all positive, but only significant in Gini coefficient regressions. As large net interest margins suggest less efficiency, the positive coefficients reflect inequality-reducing effects through enhanced efficiency in financial institutions. The results indicate that a reduction of 1 percentage point in the net interest margin can reduce inequality by a percentage point of 0.44. The results on the measure of stock market efficiency are negative and significant in all the regressions, which is consistent with the beneficial effect of increasing stock market efficiency for inequality reduction. In quantitative terms, the estimation results imply that a 1 percent increase in the stock market turnover ratio can reduce the Gini coefficient by a percentage point of 0.055 (Column 4), and reduce the poverty gap by a percentage point of 0.016.

Table 6 summarizes the findings on the impact of financial institutions and market stability on income inequality and poverty. Similarly to Jeanneney and Kpodar (2011), our results confirm the worsening effects of financial instability on income distribution. The coefficients show that only the stability of financial institutions, measured by the ratio of regulatory capital to risk-weighted assets, helps to reduce income inequality. On average, a 1 percent increase in the ratio of regulatory capital to risk-weighted assets can lower the Gini coefficient and poverty gap by percentage points of 0.375 and 0.342, respectively.

In contrast to the favorable impact of financial depth, access, efficiency, and stability on income distribution, Table 7 shows that domestic and external financial liberalization tend to widen income inequality and poverty. Positive coefficients on domestic liberalization suggest that the aggregation of indices of credit control, interest rate control, entry barriers, and privatization can significantly worsen the Gini coefficient, but not on the poverty gap. Turning to external liberalization, both inequality and poverty could increase as the consolidated foreign claims of BIS-reporting banks to GDP (%) increase.

Tables 8 and 9 investigate the role of country income levels on the relationship between inequality, poverty, and finance. In Table 8, qualitatively looking at the coefficients of interaction and finance terms, the three income groups in most cases have the same finance-inequality relationships. However, in Columns (2), (3), (7), and (8), the low-income countries have different finance-inequality effects compared to rich countries. Most noticeably, a 1 percentage point increase in private credit tends to reduce the Gini coefficient by a percentage point of 0.059 in the high-income group; while this effect increases the Gini coefficient by a percentage point of 0.029 (0.088-0.059=0.029) in low-income countries. This result shows financial depth affects inequality differently according to a country’s income level. Similarly in Table 9, the poverty reduction effects of some financial development indicators are different in the three income groups. However, the differences are not consistent to conclude that the income level of a country can affect the finance-poverty relationship.

Table 8:

Inequality-FD add Country income level interaction

article image
Note: parentheses report standard errors* p<0.10, ** p<0.05, *** p<0.01
Table 9:

Poverty-FD add Country income level interaction

article image
Note: parentheses report standard errors* p<0.10, ** p<0.05, *** p<0.01

Tables 10 and 11 test whether institutional quality affects the linkage of inequality, poverty, and finance. When regressed with the Gini coefficient (Table 10), the interaction and finance terms are both positive and significant in Column (5), which studies the net interest margin effect. These two positive signs imply that lower interest margins (better financial efficiency) can reduce income inequality and that this effect is larger when a country has a stronger rule of law. Columns (9) and (10) show diverse signs in both interaction and finance terms. However, since financial liberalization is shown to aggravate inequality, the negative sign in the interaction terms means improving the rule of law can reduce this worsening effect. Table 11 provides similar results in the regressing-with-poverty gap. Only Columns (6) and (7) show significant and consistent signs, suggesting that a better rule of law tends to intensify the favorable effects of market efficiency and bank stability.

Table 10:

Inequality-FD add quality of institution interaction

article image
Note: parentheses report standard errors* p<0.10, ** p<0.05, *** p<0.01
Table 11:

Poverty-FD add quality of institution interaction

article image
Note: parentheses report standard errors* p<0.10, ** p<0.05, *** p<0.01

The estimated coefficients on the control variables turn out to be as expected. GDP per capita, government consumption, and trade openness are negatively correlated with inequality and poverty, which means that higher income, government spending, and trade openness contribute to narrowing inequality and alleviating poverty. The coefficient on inflation is positive, reflecting a worsening distribution effect.

To find out the robustness of our results to the business cycle effect, we re-estimate all the regressions using non-overlapping five-year average data. The initial annual results are mostly confirmed and displayed in Tables 12 and 13.

Table 12:

Gini coefficient 5 year average sample

article image
Note: parentheses report standard errors* p<0.10, ** p<0.05, *** p<0.01
Table 13:

Poverty Gap 5 year average sample

article image
Note: parentheses report standard errors* p<0.10, ** p<0.05, *** p<0.01