Examining the impact of personal freedom on income inequality: Evidence from Sub-Saharan Africa and Western European regions

Benard Ohene Kwatia; Godfred Amewu; Christopher Boachie

doi:10.1371/journal.pone.0302730

. 2024 Apr 29;19(4):e0302730. doi: 10.1371/journal.pone.0302730

Examining the impact of personal freedom on income inequality: Evidence from Sub-Saharan Africa and Western European regions

Benard Ohene Kwatia ^1,^*, Godfred Amewu ^2,^#, Christopher Boachie ^3,^#

Editor: Ghulam Rasool Madni⁴

PMCID: PMC11057724 PMID: 38683838

Abstract

Rising income inequality has recently garnered intensive attention owing to its significance in theory and practice. This has rendered the mechanism through which inequality can be understood, even from a microscopic perspective, to institute proper policies that curtail it important. We contribute to the literature by examining the effect of freedom on income inequality using data from 34 least-free and 18 most-free countries from 2000 to 2020 in Sub-Saharan Africa and Western Europe, respectively. A novel fixed-effects panel quantile regression econometric estimator was employed, and the findings showed a non-linear relationship between personal freedom and income inequality for both regions. Again, both regions reported an inverse relationship between personal freedom and income inequality within the medium-run quantiles. However, a positive relationship was reported in the short-run and long-run quantiles of Sub-Saharan Africa and Western Europe, respectively, whereas the inverse was the same. Finally, the significant adverse relationship was more dominant in Western Europe, implying that high levels of personal freedom in the region explain the low levels of income inequality compared to Sub-Saharan Africa, which has lower levels of personal freedom and higher income inequality. Concerning the findings, it is recommended that policymakers and governments of least- and most-free regions institutionalise personal liberties that support human capital development and establish mechanisms to implement enacted freedoms.

1. Introduction

Over the past decades, the global rise in income inequalities has widened within countries [1,2]. To address this challenge, the UN’s 2030 agenda for sustainable development goals (SDG) has focused on ‘leaving no one behind’ [3]; This has directed the discussions on many global policies to confront all forms of inequalities to foster equal access to opportunities, making poverty a primary focus of research in development economics, especially in Africa [4,5]. However, Africa has recorded resilient economic growth among many countries and has enjoyed a recovery since the mid-1990s yet, poverty and human development indicators have not developed as anticipated, and the World Bank report on the attainment of the Millennium Development Goals (MDGs) showed that, poverty has been declining across all regions of the world except Africa [5–8]. Income inequality has empirically been shown to impede growth [9–13]. As asserted by the United Nations Development Programme [14–16], high- and rising-income inequality impedes and halts efforts to eradicate poverty. Thus, had income disparity been lower, African countries would have made considerably greater progress in growth and poverty reduction than they have done over the years [7]. This raises concerns for scholars, governments, and policymakers about understanding the determinants and trends of income disparity within and among countries [17].

Many studies have explored the determinants of income inequalities from several aspects of economic variables, namely; financial development [18], globalization [19], market reforms and private capital [20], technology [21,22], educated labour [23,24], financialization, trade union, labour and welfare [25], remittances and financial development [26], labour institutions [27,28], tourism and governance quality [29], taxes [30] and inflation, export, population, labour force, and unemployment [31]. However, very few empirical studies focus on income inequality and institutions [7,32–35] Besides, these few studies also focus on formal institutional measures that capture political, legal, and economic institutions such as government effectiveness, government spending, rule of law, democracy, regulatory quality, voice and accountability, economic freedom, inflation, external debt, and trade ignoring the informal institutions-income inequality nexus. Given that, informal institutions have been established as a crucial driver of economic relevant behaviour such as financial development [36], social capital and economic transition [37], development [38], bank performance [39] and economic development [40]; it is startling that, empirical studies examining the impact of informal institution on income inequality are still in the minority.

To fill the highlighted existing gap in the literature, this study contributes to the literature as follows: First, the study provides further insight into the determinants of income inequality by extending the informal institutions-inequality nexus by proxying informal institutions with personal freedom. The study opts for personal freedom over other informal institutional measures because, personal freedom captures the fundamental human rights of individuals to exercise autonomy, pursue their own goals, express opinions and conduct themselves as one desire, without coercion and undue interference, so far as others are not harmed [41,42]. To this end, the study advances recent studies on informal institutions and income inequality [43,44], which focus on trust and attitudes and informal rule systems, respectively, to determine the relationship between informal institutions and income inequality. This study argues that informal rules, attitudes, and trust are built due to patterns of behaviour based on the internalisation of values, norms and faith in enforcing unwritten rules [45]–primarily influenced by personal freedom. Thus, personal freedom creates an environment of empowerment and autonomy that fosters a sense of confidence and trust [46]. Consequently, the study opines that, without personal freedom, one may not have any trust in social relationships and institutions to spur the needed innovations and competition essential to promoting inclusive policies that address income disparities [47,48].

Second, to help draw meaningful conclusions and better understand whether or not Africa’s persistent income inequality flows from personal freedom, the study employs an advanced form of personal freedom variable by Fraser Institute that captures freedom of religion, freedom of relationships, freedom of association, assembly and civil society, freedom of expression and information, and freedom of movement, all within the context of legal protection and security. Third, the persistence of income inequality and low levels of freedom in Africa is alarming. In Africa, the average earnings of the top 10% are about 30 times that of the bottom 50%, while in Europe, the difference is about 11 percentage points. This makes Africa record the utmost gap between the average earnings of the top 10% and incomes of the bottom 50% above other regions, even in contrast with high-inequality regions, such as the Middle East, Brazil, and India [49,50]. Again, on personal freedom, Fraser Institute’s 2022 report on the human freedom index identifies sub-Saharan Africa, herein SSA, as the least-free region, while Western Europe, herein West, is identified as the most-free region [51]. For this reason, this study seeks to understand how the dynamics of environments characterised by high and low levels of personal freedom impact income inequality by comparing the two regions to draw empirical conclusions. Fourth, as argued by [49,52,53], synthetic income inequality measures such as the Gini coefficient are unable to accurately account for the structure of inequality, especially for heavy-tailed distributions [54] hence, call for the need to go yonder. To help better understand the structure of persistent income inequality in Africa, this study employs the income share of the top 10%, which has been described as key in explaining overall inequality [55], as a proxy for income inequality. Finally, the literature on income inequality has mostly examined the relationship with other economic variables by employing a conditional mean model with a fixed effect as seen in studies such as; [22,29,56]; however, this study goes beyond by employing a panel quantile regression estimator [57] to test for non-linearity and understand the heterogeneity in the effect of personal freedom on income inequality at different quantiles.

The remainder of the paper is structured as follows. Section 2 provides a brief review of the literature. Section 3 focuses on the data description and their sources, methodology and analysis approach employed in the study. The empirical results and discussions are presented in Section 4, and Section 5 concludes the study with policy recommendations.

2. Literature review

2.1 Theoretical review

The initial studies on inequality date back to [58], who associated aggregate demand with income distribution. However, the seminal work of [59] provided an approach to studying income inequality and economic development. The study investigated the relationship between inequality and development. It argued that income inequality tends to increase during the initial stages of economic development while inequality decreases in the progressive stages of development. Nonetheless, as economic development peaks, inequality again increases–a phenomenon commonly described as the Kuznets inverted U-shaped curve and, thereafter, the Kuznets curve. Theoretically, the relationship between personal freedom and inequality emerged years ago. [60] posits that high levels of personal freedom that give individuals the liberty to participate in free trade while upholding their interests will inevitably result in greater equity in the distribution of wealth and opportunities. [61] also argued that expanding people’s capabilities and opportunities–what he termed as the "capabilities approach", can foster narrowing inequality. Thus, equitable distribution of opportunities and resources is enhanced within societies where individuals can pursue their goals and make choices. More recently, [46] distinguished negative freedom–the absence of interference, from positive freedom–the ability to act and attain goals autonomously; and argued that enhancing positive freedom by empowering individuals to overcome barriers that restrict their prospects on economic, societal and political fronts reduces inequality and creates more equitable distribution. As asserted by [62], the humanly devised constraints that structure interactions among groups and individuals, comprising patterns of organization embodied in laws, habits, and customs, and formerly constituted establishments; thus, the rules of the game in society have been described as institutions [63]. Conversely, [63] postulate that a market economy’s efficient and proper functioning depends on suitable institutions. Formal institutions are the explicit set of written and codified laws, contracts, constitutions, official rules, regulations and structures [63–65], while informal institutions emerge gradually and are often deeply entrenched in the beliefs, traditions, culture, trust, norms, and values that shape behaviour and guide social interactions within a community [63,66–68].

2.2 Formal institution–income inequality nexus

Prior studies have shown that institutions majorly contribute to rising income inequalities among and within countries [34]. However, the effect documented in the literature has not reached a consensus. [69] investigated how labour market institutions affect inequality, income volatility and labour productivity in Brazil. They found that, in general, Brazil’s social protection system is redistributive. However, it may improve its progressive aspects, notably concerning the pension system, and systemic changes can increase efficiency and reduce inequality. [70] explored the impact of foreign direct investment (FDI) on income inequality in Vietnam using a generalised method of moment (GMM) estimation on a panel of 63 provinces from 2012 to 2018 under the moderating effects of institutions and education. The authors found evidence of non-linearity between FDI and income inequality. Furthermore, FDI increases income inequality in Vietnam, but the effect differs depending on the level of education and institutional quality of the province. [71] used the 2010 China Family Panel Studies (CFPS 2010) to assess income inequality in China. They found that the rise in income inequality in China is driven by structural factors attributable to the Chinese political system. In the studies of [11], income inequality in Jordan was explored using expenditure inequality as a proxy to determine the drivers of income disparity between rural and urban areas and between male-headed households (MHHs) and female-headed households (FHHs). The unconditional quantile regression (UQR) model was employed for household expenditure and income surveys (HEIS, 2017/2018), and the results showed that the education of household heads is a crucial driver of income inequality in Jordan, as well as the geographical location of households. [55] used microdata from the Household Finance and Consumption Survey (HFCS) from 2010 to 2016 to study income and wealth disparities in Malta. The study employed decomposition methods and binary response models and found that income and wealth distribution changed in favour of households in the upper parts of the distributions and persons with tertiary education. Additionally, the decomposition of income inequality revealed labour income as a key driver of income inequality; however, pension income reduces the inequality gap. [72] explored income inequality and institutions in 20th-century America, comparing the current situation to those from the early postwar years. The authors argued that during the early postwar years, government efforts to distribute income were instituted through minimum wages, negotiating frameworks, unions, and progressive taxes. However, in recent years, there has been a reversal in all these institutional patterns. The study concludes that income inequality in the United States can be explained by international trade and skill-based technical change, functioning within a larger institutional narrative. Moreover, the income distribution was found to be strongly influenced by economic institutions.

In their empirical analysis, [73] employed a system-GMM estimator to test the dynamic relationship between institutions, proxied by an extensive array of governance measures, and inequality proxied by the Gini index, to determine if there is reinforcement as predicted by the model. Again, the panel VAR estimator was employed to test the direction of causality between the two through linear feedback analysis on a global sample of 121 countries from 1960 to 2000. The study found that institutions cause income inequality while inequality causes institutions; however, the direction of causality appears to be dominant from inequality to institutions. [7] focused on 52 African countries to understand the dynamics or prevalence of income inequality in the region using macroeconomic and institutional data–the rule of law and transparency. Using panel data models such as ordinary least squares (OLS) and two-stage least squares (TSLS) over the period 1980–2017, they found that, in Africa, income inequality decreases in the least developed and low-income economies while in high-income economies, it increases. Again, the study found that, across convergence clubs, the role of institutions and macroeconomic factors in explaining income inequality is limited and differs. [74] explored a panel of 68 countries comprising 35 developed and 33 developing countries from 2000 to 2019 using the generalised method of moments (GMM) to establish the impact of institutional capacity and fiscal policy on income inequality. The authors found that income tax decreases income inequality in developing countries but not in the total sample. Government size and health and education expenditures were found to be negatively linked to income disparities in developing countries. Institutional capacity measured by corruption and government effectiveness was found not to affect income inequality in the global sample; however, significantly, the coefficients of the interactions between institutional capacity and fiscal policy bear the expected signs, although insignificant. In the work of [75], a first attempt was made to address the question of whether institutions and freedom of the press affect income distribution in BRICS-T (Brazil, Russia, India, China, South Africa, and Turkey) countries from 1993 to 2016, and found that institutional quality and freedom of the press reduce income inequality. An augmented mean group (AMG) was employed to estimate the long-run relationship between institutions, freedom of the press and income inequality. [76] examined the effects of legal, economic, and political institutions on income inequality. Data on 21 OECD advanced countries over two decades were employed in the analysis, and it was found that the impact of society’s utilisation on income inequality is more pronounced than that of unionisation. In the empirical examinations of [77], the authors employed panel data to assess the impact of financial reforms on income inequality in 62 countries from 1973 to 2005. They found that removing subsidised directed credit, improving the securities market policy, decreasing excessively high reserve requirements, and fostering financial reforms foster equal income distributions. They also provide evidence of a non-linear relationship between income inequality and per capita income (Kuznets curve), showing that the size of government is critical to reducing income inequality while trade enhances unequal income distribution.

2.3 Informal institution–income inequality nexus

On informal institutions and inequality, [44] compared two highly income-unequal Latin American cities, Santiago in Chile and Medellin in Colombia, to examine how informal institutions impact income distribution. Valenzuela-Levi found that informal rule systems influence income distribution within public transport. [78] examined the link between informality and income inequality using time series and panel data from African countries. The authors estimated a fixed effect model on the panel from 1990 to 2015 and found that informality contributed to decreasing income disparities in Africa. In the empirical examination of [43], the authors employed unbalanced panel data from 1981 to 2014 to examine the impact of trust and attitudes towards income inequality in a global sample. They found that mistrust increases the desire for income inequality in societies where individualism is strongly valued. [79] empirical study was on the relationship between inequality and trust in 22 European countries, covering 216 regions from 2010 to 2014. The study employed multilevel modelling in its analysis and found that differences in income inequality across countries, rather than within countries, account for most of the variation in social trust. [80] investigated the mediating role of the fairness gap (the gap between perceived and preferred inequality) in political trust and income inequality from a cross-country perspective. The author found that a high perceived fairness gap relates to low political trust and moderates the political trust-income inequality nexus. Additionally, when the fairness gap is disaggregated, it becomes clearer that political trust is more closely correlated with changes in inequality perception than with changes in inequality preference. The findings were obtained using pooled ordinary least square regressions and mediation analysis. [81] employed a structural equation model to examine the direction of causality between trust, welfare state and income inequality in 104 countries and found that trust supports welfare state policies that decrease net income inequality. However, the reduced inequality does not increase trust. [56] examine the impact of national culture on income inequality using OLS and 2SLS instrumental variable estimators in a cross-country analysis. The authors found that countries with high levels of collectivist culture tolerate high-income disparity compared with individualistic countries. In support of the above, [82] studied the relationship between culture and income inequality, concentrating on individualistic and collectivistic cultural dimensions. The author found that individualistic countries have significantly lower after-tax income inequality and higher income reallocation than collectivist countries. The study also found that the individualism-redistribution nexus is driven by countries with higher incomes, which seem to be affected by distinct strains of individualism.

3. Methodology and data

3.1 Data

The study uses data from different sources to investigate a panel of 52 countries comprised of 34 sub-Saharan African countries and 18 Western European countries from 2000 to 2020. The choice of study period and countries employed in the study were dictated by the period for which reliable data were available for the study’s comparative analysis. The regressand in this study is income inequality. As revealed by [49], the pan-African income distribution characterized by extreme levels of inequality with highly high incomes at the top and relatively low incomes at the bottom, calls for the need to go beyond synthetic inequality measures such as the Gini coefficient which is unable to accurately account for the structure of income inequality especially in heavy-tailed distribution as seen in Africa. As such, the study proxy’s income inequality as the pre-tax income share held by the top 10% of the population sourced from the World Inequality Database (WID).

The main regressor is personal freedom, which is sourced from Fraser Institute’s 2022 Human Freedom Index. To help establish whether the causal effect of personal freedom on income inequality is reliable, control variables–trade openness, inflation, general government expenditure, gross domestic product (GDP) growth, and age dependency ratio–deemed significant for the investigation of income inequality are included in the analysis. Trade openness captures the summation of exports and imports normalized by GDP and is expected to reduce inequality [83]. Inflation measures the annual changes in the cost to the average consumer of acquiring a basket of goods and services; at a higher level, it is expected to increase inequality [84]. High levels of inequality call for resource allocation to fund the redistribution of income, thus increasing total government expenditure; as such, inequality is expected to narrow at high levels of government expenditure [85]. The level of economic growth is captured by GDP growth; as growth increases, income inequality is likely to decrease [86]. Age dependency is the sum of the working-age population relative to the sum of the non-working population (those below the age of 15 and above the age of 65). The higher the dependency ratio, the wider the income inequality is expected to widen [87]. All control variables were sourced from World Development Indicators (WDI). Details of variables employed in the study, their coded description, data sources and priori expectation have been described in Table 1 below.

Table 1. Variable description and data sources.

Variable	Code	Source	Expected Sign
Income Inequality	INI	World Inequality Database
Personal Freedom	PEF	Fraser Institute	-
Trade Openness	TOP	World Development Indicators	-
Inflation	INN	World Development Indicators	+
General Government Expenditure	GOV	World Development Indicators	-
Gross Domestic Product	GPG	World Development Indicators	-
Age Dependency	AGD	World Development Indicators	+

Open in a new tab

3.2 Empirical estimation

Most prior studies examining the factors that impact income inequality have often employed a conditional average model with a fixed effect to explore the relationship [7,22,29,80,88].

{Q (y}_{i t} (τ x_{i t}, \partial_{i}) = + x_{i t}^{'} β (τ) + \partial_{i},

(1)

Where y_it is the logarithm of Income Inequality for country i at a year t, $x_{i t}^{'} \dots \dots x_{i t p}^{'}$ is the px1 vector of the independent variable and ∂_i is the unobserved country fixed effect, which controls for the time-variant of unobserved heterogeneity. However, this estimation considers only the conditional expectation of the regressand and does not adequately capture the conditional distribution’s overall picture [26]. As the variables employed in this study do not fulfil the assumption of no outliers and normality, we adopt fixed effects panel quantile regression [57] to understand the heterogeneity in the effect of personal freedom on the conditional distribution of income inequality. These heterogeneous effects have been demonstrated to offer valuable information that conventional regression techniques, such as ordinary least squares (OLS), cannot deliver [89]. In contrast to conventional regression, such as OLS, panel quantile regression is more robust in addressing problems of unobserved heterogeneity, heteroscedasticity and outliers in the variables [57,90]. Thus, it uses non-linearity and asymmetry to deal with heterogeneity and endogeneity concurrently [91]. This makes it more appealing for policymakers to do panel analysis, which stratifies the distributive effect of the variables in different quantile distributions [92].

The empirical model is set following Koenker (2004) [57]:

{Q y}_{i t} (τ x_{i t}) = \partial_{i} + x_{i t}^{'} β (τ) + μ_{i j} j = 1, \dots ., m_{i}, \dots i = 1, \dots n

(2)

where Qy_it(τx_it) represents the τth quantile of the dependent variable x_it depicts the vector of explanatory variables, ∂_i denotes the individual effect, τ represents the quantile, β(τ) is the regression parameter of the τth quantile and can be estimated as follows:

\min_{(\partial, β)} \sum_{k = 1}^{q} \sum_{j = 1}^{n} \sum_{i = 1}^{m_{i}} w_{k} ρ_{τ k} (y_{i t} - \partial_{i} - x_{i t}^{'} β (τ))

(3)

where $ρ_{τ (u)} = u (τ - 1 (u < 0)),$ represents the piecewise linear quantile loss function [93]. Weight w_k controls the relative influence of the quantile on the estimation of the ∂_i which is defined as:

If y_{i t} - \partial_{i} - x_{i t}^{'} β (τ) < 0, then

(4)

w_{i t} = τ

y_{i t} - \partial_{i} - x_{i t}^{'} β (τ) < 0,

(5)

w_{i t} = 1 - τ

From Eqs (4) and (5), we note that quantile regression is a type of weighted regression by setting 1&τ positive and negative residuals, respectively [94], which ignores unobserved country heterogeneity. However, owing to the limited yearly observations in our study, we follow [57] estimation approach, in which the unobservable individual effect ∂_i acts as one of the regression parameters. This is defined as follows;

\min_{(\partial, β)} \sum_{k = 1}^{q} \sum_{j = 1}^{n} \sum_{i = 1}^{m_{i}} w_{k} ρ_{τ k} (y_{i t} - γ_{1 τ} {P E R}_{i t} {+ γ_{2 τ} T O P}_{i t} + γ_{3 τ} {I N N}_{i t} + {γ_{4 τ} G O V}_{i t} {+ γ_{5 τ} G P G}_{i t} + γ_{6 τ} {A G D}_{i t} - \partial_{i}) + λ \sum_{i = 1}^{n} | \partial_{i} |

(6)

Where λ is the turning parameter of the individual effect, w_k is the weight the of k^th quantile, m is the index of the quantile PER, TOP, INN, GOV, GPG and AGD of a nation i, time t and income inequality y.

4. Results and discussions

4.1 Descriptive statistics

The characteristics of the variables of interest are presented in Tables 2 and 3 for sub-Saharan Africa and Western Europe, respectively. Noticeably, the variance of the variables is relatively large, with TOP, INN and AGD dominating for SSA and TOP, GOV and AGD dominating for West. The summary statistics show that the series are leptokurtic distributions, meaning that the distribution has a fatter tail. This supports [95] argument that normality tests are significantly non-normally distributed.

Table 2. Summary statistics for income inequality and personal freedom in sub-Saharan Africa.

Summary statistics	INI	PEF	TOP	INN	GOV	GPG	AGD
Mean	0.520	6.646	65.114	8.374	14.398	4.089	84.459
Median	0.503	6.720	57.497	4.690	13.825	4.470	87.221
Maximum	0.716	8.520	163.619	557.202	43.484	33.629	111.477
Minimum	0.405	3.880	16.352	-16.860	0.952	-36.392	40.421
Std. Dev.	0.067	0.970	28.427	34.995	6.608	4.624	14.449
Skewness	0.872	-0.164	1.003	13.170	1.474	-1.136	-0.989
Kurtosis	3.123	2.312	3.541	186.748	6.295	17.186	4.021
Jarque-Bera	79.931	15.196	112.686	900192.500	510.654	5392.145	129.525
Probability	0.000	0.001	0.000	0.000	0.000	0.000	0.000

Open in a new tab

Note: INI = Income inequality, PEF = Personal freedom, TOP = Trade openness, INN = Inflation, GOV = General government expenditure, GPG = Gross domestic product growth, and, AGD = Age dependency ratio.

Table 3. Summary statistics for income inequality and personal freedom in Western Europe.

Summary statistics	INI	PEF	TOP	INN	GOV	GPG	AGD
Mean	0.335	9.375	114.457	1.820	20.364	1.756	51.167
Median	0.333	9.400	85.143	1.731	19.985	1.947	51.111
Maximum	0.426	9.690	377.843	12.694	28.148	24.370	62.765
Minimum	0.268	8.140	45.419	-4.478	10.424	-11.325	43.397
Std. Dev.	0.029	0.250	74.456	1.514	3.848	3.327	3.785
Skewness	0.261	-1.113	1.749	1.873	-0.542	0.472	0.422
Kurtosis	2.759	4.979	5.239	15.117	2.930	12.099	3.130
Jarque-Bera	5.217	139.680	271.685	2533.592	18.561	1317.866	11.508
Probability	0.074	0.000	0.000	0.000	0.000	0.000	0.003

Open in a new tab

The stationary properties of all variables were checked, and the [96,97], ADF-Fisher Chi-square [98], and PP-Fisher Chi-square [99] results are detailed in Tables 4 and 5 for both regions. Most series are I(1) except for AGD and INN in SSA and AGD, GPG and INN for West that were I(0). Given that cross-sectional dependence is present in our model, the first-generation panel unit root test–which often assumes cross-sectional dependence–may not be adequate to address it [92].