Income inequality and economic complexity in Africa: the moderating role of governance quality

Abstract

This study examines the relationship between economic complexity and income inequality in 24 selected African countries. Specifically, it investigates the moderating role of governance quality in the link between income inequality and economic complexity. The study applied the systems GMM panel estimation method to data spanning from 2000 to 2018, sourced from the World Bank, Harvard Dataverse, and Baker Library databases. The findings indicate that economic complexity significantly worsens income distribution in the study area. However, when the composite governance indicator is used as a moderator, the coefficient of the interaction term becomes negative and significant, implying that governance factors play a crucial role in mitigating the adverse effects of economic complexity on income inequality. Similar results were arrived in the disaggregated analysis of governance quality indicators, where each governance factor offsets the negative impact of complexity on income inequality. The findings suggest policy measures aimed at enhancing the quality of governance system to counteract the deteriorating effects of economic complexity on income inequality.

Keywords:

• Economic complexity
• GMM
• governance quality
• income inequality

Reviewing Editor:

• George Mudimu, Development Sciences, Marondera University of Agricultural Sciences and Technology, Marondera, Zimbabwe

Subject:

• Econometrics
• Development Economics
• Economics
• Industrial Economics

1. Introduction

Economic complexity, a relatively recent addition to economic literature, provides novel opportunities for a comprehensive exploration of countries’ economic development processes. This concept aims to grasp the productive structures of economies through establishing the Economic Complexity Index (ECI), where the index serves as a metric capturing knowledge-based productive capabilities and the potential for economic diversification in countries (Allen Whitehead & Bhorat, 2021; Mealy et al., 2019; Mealy & Teytelboym, 2020).

Economic complexity is an important framework for understanding the complex situations that countries encounter in their development and economic diversification. It employs techniques of dimensionality reduction to create an economic sophistication index. This index condenses interconnected outcomes of economic growth, technological advancements, inequality, regional disparities, and resilience (Balland et al., 2022; Balsalobre-Lorente et al., 2022; Bhorat et al., 2019; Hidalgo, 2021). The primary aim of measuring the index is to loosen these complex interactions and analyze their impact on the social and economic development of nations. By identifying the productive knowledge and economic diversification of countries, the index offers valuable insights into the evolution of their productive structures and trade partners. This enables a more comprehensive analysis of extensive datasets and a deeper understanding of decision-making processes (Sciarra et al., 2020; Simoes & Hidalgo, 2011).

Before investigating into the effect of economic complexity on income inequality, let’s provide a concise overview of the factors influencing economic complexity. Many recent empirical studies have scrutinized these determinants. Gnangnon (2021), for instance, underscores the crucial role of a country’s productive capacity in shaping its economic complexity. Other studies, including those by Yalta and Yalta (2021), Caous and Huarng (2020), and Sadeghi et al. (2020), have probed into the impact of human capital on economic complexity, affirming that the accumulation of human capital significantly drives economic complexity. Notably, Mishra et al. (2020) and Sepehrdoust et al. (2019) have identified a positive correlation between financial globalization, trade liberalization, and economic complexity. Furthermore, Di Clemente et al. (2021) have tried to establish a link between economic complexity and urbanization. Collectively, these studies support the idea that a country’s productive capacity, technological knowledge, human capital accumulation, trade openness, and urbanization process are the key determinants of its economic complexity level.

In addition to exploring the determinants of economic complexity, there is a growing interest in understanding the relationship between economic complexity and the economic growth of countries. Numerous empirical studies, conducted in both advanced and developing nations, have investigated the impact of economic complexity on countries’ economic growth. These studies reveal that economic complexity exerts a significant positive influence on the economic growth of countries (Hoeriyah et al., 2022; Lybbert & Xu, 2022; Mao & An, 2021; Sayehmiri et al., 2021). Specifically, the findings indicate that countries with higher levels of economic complexity tend to experience greater prosperity, characterized by high incomes and complex production structures that contribute to increased economic gains. Furthermore, the research highlights the fact that differences in economic complexity can explain variations in growth rates among countries. For instance, Mealy et al. (2019) argue that economic complexity provides valuable insights into the observed disparities in economic growth across different countries.

Examining the impact of economic complexity on income inequality with governance quality as a moderating factor holds significance for several reasons. First, economic complexity is a recently introduced concept in economic literature, particularly relevant for understanding its effects on income inequality, especially in developing nations marked by elevated income disparities. Second, governance quality stands as a well-established factor with demonstrated significant impacts on income inequality. By scrutinizing the moderating influence of governance quality, we gain insights into how economic complexity shapes income inequality across diverse contexts. Lastly, the outcomes of this research carry substantial policy implications. This is because enhancing our comprehension of how governance quality moderates the relationship between economic complexity and income inequality provides a basis for formulating more effective policies aimed at reducing income disparities.

Empirical studies examining the relationship between economic complexity and income inequality are scarce, and existing research yields mixed results. Notably, studies by Chu and Hoang (2020), Lee and Wang (2021), Sepehrdoust et al. (2022), and Chutimunt (2019) indicate that an intensified level of economic complexity correlates with a decrease in income inequality. These authors explain the mechanism by which economic complexity influences income distribution. Firstly, an increased level of economic complexity suggests a sophisticated economic structure that could offer a broader array of occupational choices with enhanced remuneration, fostering a more equitable system. Secondly, complex economies producing sophisticated goods have a higher likelihood of maintaining a balanced occupational structure, allowing workers to accrue higher lifetime incomes. Lastly, complex economies, with complex productive structures, involve workers in production activities characterized by high level of productivity and increasing returns to scale.

On the other hand, additional studies (Bandeira et al., 2021; Hasanvand et al., 2022; Lee & Vu, 2020) revealed a positive relationship between economic complexity and income inequality. These outcomes imply that in developing countries marked by less complicated economic systems, individuals face fewer opportunities to secure highly productive occupations. Consequently, the task of reducing income inequality becomes more challenging in such settings due to the limited available opportunities.

Other empirical studies have further enhanced our understanding of the relationship between income inequality and institutional factors, demonstrating that enhancements in institutional quality can contribute to a reduction in income inequality (Blancheton & Chhorn, 2021; Coccia, 2021; Huynh & Tran, 2022; Zehra et al., 2021). However, Asamoah’s study yielded divergent results for various indicators of institutions, namely the World Governance and International Country Risk Guide indicators. Interestingly, a Kuznets inverted U-shaped relationship was distinguished between these indicators and income inequality, observed across both advanced and developing countries. This signifies a complex relationship between institutions and income inequality (Asamoah, 2021).

The empirical studies identified above have largely examined the distinct effects of institutional factors and economic complexity on income inequality, revealing that enhancements in these factors lead to a reduction in inequality (Chu & Hoang, 2020; Prasetiya, 2021). However, the influence of economic complexity on income inequality, considering governance as a moderating factor, remains unexplored. Thus, this study considers the interaction effect of the institutional factors on the relationship between income inequality and economic complexity in Sub-Sharan Africa during 2000 to 2018. Investigating this interaction could offer a more comprehensive understanding of the dynamics involved in income inequality and the role of governance in shaping this relationship. Such an investigation has the potential to deliver valuable insights for policymakers and stakeholders working to address income inequality and foster inclusive growth.

The remaining sections of the study are structured as follows. Section two presents a comprehensive review of existing literature on the relationship between income inequality and economic complexity, covering both theoretical and empirical perspectives. Section three is dedicated to describing the study’s utilized data and outlining the empirical estimation strategies applied. Moving forward, section four exposes the primary findings of the study and involves in an in-depth discussion of these outcomes. Lastly, the concluding section of the study concludes the research and provides policy implications based on the obtained results.

2. Review of literature

Earlier development theories focused on the trade-off between income inequality and long-term economic growth in developing countries. Kuznets’s inverted U-shape hypothesis, widely accepted, served as a framework for comprehending shifts in income inequality resulting from economic growth (Kuznets, 1955). However, subsequent empirical studies challenged this hypothesis and came out with inconclusive results concerning the impact of growth on inequality. Factors like reverse causality and variations in growth patterns among countries etc. contributed to this ambiguity (Kanbur & Sumner, 2012). This suggests that growth and income distribution are not mutually exclusive, necessitating policymakers to address both simultaneously.

In the late 1990s, the substantial increase in poverty rates in developing countries underscored the significance of personal income distribution and its impact on poverty reduction. Consequently, researchers and policymakers recognized the imperative of prioritizing both accelerated economic growth and more equitable income distribution for development. This acknowledgment led to the formulation of a pro-poor growth framework, aiming to parallel address poverty reduction, equitable income distribution, and accelerated economic growth. The framework suggested that by attaining rapid economic growth and greater income equality, effective poverty reduction could be achieved (Essama-Nssah, 2005; Kakwani et al., 2003; Kakwani & Pernia, 2000).

However, since the early 2000s, the emphasis on pro-poor growth has diminished, giving rise to an increasing focus on inclusive growth in the development sphere (Doumbia, 2019). The concept of inclusive growth contends that it is imperative to involve a broader segment of the population in sharing the benefits of growth, especially considering that a significant portion of the population resides at the lower end of the income distribution (Anand et al., 2013; Ranieri & Almeida, 2013). This move has brought two critical issues to the forefront for policymakers: the relationship between inequality and growth, as well as that of the link between inequality and poverty.

The increase in income inequality in developing and emerging economies since the 1980s has prompted intensified focus on the matter of inequality and redistribution policy in academic and political dialogues. These discussions revolve around comprehending the origins and consequences of inequality, along with assessing the effectiveness of policies targeting these concerns. Consequently, there has been an expanded exploration of the factors contributing to income inequality, encompassing trade globalization, financial liberalization, technological change, macroeconomic policies, labor market policies, and governance issues (Dabla-Norris et al., 2015). In the subsequent section, we will conduct a thorough review of recent empirical literature exploring into the relationship between economic complexity and income inequality. Furthermore, we will explore the role of governance quality in mitigating

In their study, Khanzadi et al. (2022) investigated the impact of economic complexity on income inequality in Iran. The study defines economic complexity as the level of production sophistication achieved through enhanced knowledge and skills within a country’s workforce, which according to the researchers, leads to the production of diverse and unique products, including more advanced and sophisticated ones. The improved specialization and product sophistication resulting from economic complexity ultimately contribute to increased income in society. Consequently, they argue that this helps alleviate income inequality by enhancing productivity and generating products with high rates of return. In summary, their findings revealed a significant negative effect of economic complexity on income inequality.

In Iran, Hasanvand et al. (2022) conducted a study examining the relationship between economic complexity and income inequality, utilizing a simultaneous equation modeling approach. The researchers postulated that economic complexity is an essential concept reflecting the role of knowledge in production and economic growth, playing a crucial role in addressing income inequality. The findings suggested that an increase in economic complexity leads to a reduction in inequality. The researchers recommended that enhancing knowledge and skills in production through education and industrial development can foster economic complexity, promote knowledge-based production, increase GDP, and effectively address income inequality. Consequently, the study advocates for a focus on improving the quality of education and fostering industrial development to fully realize the benefits of economic complexity.

In their study, Sepehrdoust et al. (2022) explored the impact of macroeconomic variables on income inequality in developing countries using data spanning from 2000 to 2019. Employing the panel-VAR model, they revealed that income inequality diminishes when economic complexity surpasses a specific threshold. The study underscores the crucial role played by economic complexity in determining the level of economic growth and development in the countries under investigation. The authors emphasized the significance of investing in knowledge and human capital to enhance the capabilities of economic agents and foster complex economies. They recommended that developing countries should aim to achieve economic complexity by leveraging technical knowledge and producing technology-based exports. Such an approach, they argued, would lead to more diversified economies, higher per capita income, and a more equitable distribution of income.

In a study, Bandeira et al. (2021) investigated the impact of economic complexity on income inequality in Brazil. Utilizing panel data for Brazilian states, they examined the relationship between regional economic complexities and income inequality. The study arrived at an interesting pattern whereby; as economic complexity increases, income inequality initially worsens but then improves, forming an inverted U-shape. Additionally, the researchers identified the significant role of urbanization in enhancing regional economic structure, subsequently reducing income inequality. This implies that a diverse range of industries and occupations within regional productive structures fosters income growth and diminishes inequality. The study further highlighted that the issue of inequality is particularly pronounced in less developed regional states with a more rural character. Overall, the study underscores the crucial role of economic complexity in facilitating product and occupational diversification, ultimately leading to enhancements in income distributions.

Similarly, in their study, Lee and Wang (2021) explored the relationship between economic complexity, income inequality, and country risk in 43 countries spanning from 1991 to 2016, utilizing a two-group finite mixture method. The analysis uncovered that the complexity-inequality nexus is influenced by country risk. In countries with low risk, an augmentation in economic complexity correlates with a more equal income distribution. However, in economies characterized by high economic risk, an increase in economic complexity does not necessarily result in reduced inequality.

In their study, Lee and Vu (2020) inspected the interplay between income inequality, economic complexity, and human capital across various countries. Utilizing OLS regression and system GMM techniques for data analysis, the findings indicated that countries with higher complex economic structure tend to exhibit lower levels of inequality. Furthermore, the study identified that the presence of human capital enhances this relationship, although with an indirect effect. However, according to their study the dynamic GMM analysis revealed a contrasting trend, that is, as economic complexity increases, income inequality also tends to rise.

In their study, Chu and Hoang (2020) analyzed the impact of economic complexity on income inequality across eighty-eight countries from 2002 to 2017. The researchers investigated whether this relationship is influenced by other economic and social factors. The study showed a significant positive effect of economic complexity on income inequality, worsening the unequal distribution of income. Additionally, the researchers explored the conditional nature of this relationship by examining factors such as education level, government spending, and trade openness. The findings suggested that when these factors reach a certain threshold, they have the potential to mitigate the negative impact of economic complexity on income inequality. However, in situations characterized by lower education levels, inefficient government spending, and limited economic openness, economic complexity fails to decrease income inequality. The study concludes by recommending that policymakers concentrate on improving education, enhancing government spending efficiency, and promoting economic openness to effectively address income inequality.

In her 2019 study, Chutimunt investigated the link between income inequality and economic complexity in various countries. The study examined sectoral data on wage inequality, comparing it to the level of economic complexity. The findings indicated that advanced countries with a higher per capita GDP tend to exhibit lower rates of wage income inequality across sectors. These results lend support to the notion that a more complex economy can contribute to the reduction of income inequality.

The subsequent section provides a thorough review of recent empirical literature exploring into the relationship between institutional factors and income inequality.

In his 2021 study, Asamoah employed a dynamic panel threshold model to investigate the impact of institutional quality on income distribution, utilizing data from 1995 to 2017 encompassing both advanced and developing countries. The findings revealed that the relationship between institutional quality and income inequality reduction varies across countries. In advanced countries, the indicators of governance quality exhibited a quadratic effect, while in developing countries, there was a consistent negative effect. The institutional indicators derived from the International Country Risk Guide also pointed to an inverted U-shape relationship between income inequality and institutional quality for both groups of countries. Notably, the threshold values for developing countries were observed to be higher than those for advanced countries.

Zehra et al. (2021) conducted a study examining the role of institutional quality in addressing inequality and promoting income distribution. The researchers gathered panel data from 114 countries spanning the period from 1984 to 2018, covering various political regimes. The analysis encompassed several institutional measures, including government stability, corruption index, bureaucratic quality, law and order, and democratic accountability. The study also employed the ‘governance index’ and its different dimensions as indicators of institutional quality. Utilizing the two-stage least square pooled regression and systems GMM methods, the researchers found that enhancing institutional quality generally leads to a reduction in income inequality. However, the impact varies depending on the specific indicator used. The governance index indicators, except for voice and accountability, were found to have a significant negative effect. Additionally, the study revealed that income inequality is inversely related to strong democratic institutions but positively associated with the autocratic nature of democratic institutions.

Coccia (2021) conducted a study involving 191 countries to analyze how institutional quality, measured by governance indicators, influences income inequality and poverty reduction. The findings revealed that countries with better governance parameters experienced greater reductions in poverty and income inequality. Additionally, stable and advanced countries achieved higher levels of poverty and inequality reduction compared to emerging and fragile states. To address these disparities, Coccia recommended that countries prioritize institutional improvements in governance effectiveness and the rule of law.

In a similar vein, Lee et al. (2021) explored the impact of political institutions on income inequality by examining citizens’ perceptions of distributional fairness and their ability to engage in politics effectively. Using data from the Asian Barometer Survey’s fourth wave (2014-2016), their findings indicated that political effectiveness is closely tied to how individuals perceive income distribution fairness and the role they expect the government to play in redistributive policies and programs.

In their study, Kunawotor et al. (2020) investigated the factors driving income inequality in Africa and examined the role of institutional quality in reducing it. Using a dynamic GMM technique with robust standard error, they analyzed a panel dataset from 1990 to 2017. The results revealed that aggregate institutional quality parameters did not have a significant impact on income inequality. However, specific governance indicators, such as corruption control and enforcement of the rule of law, were found to reduce income inequality. On the other hand, indicators like government effectiveness, voice and accountability, regulatory quality, and political stability did not show statistical significance in influencing income inequality. The authors recommended a stronger focus on controlling corruption and upholding the rule of law, while also emphasizing the need for overall institutional development in Africa, where weak institutions prevail.

3. Methodology

This section outlines the methodology employed in this study, providing an overview of the types and sources of data, an explanation of the variables incorporated in the analysis, and a description of the econometric models utilized.

3.1. Data sources and description

As discussed above, this study investigates the moderating role of governance quality in the link between economic complexity and income inequality. Data on income inequality, economic complexity, governance quality, and relevant control variables were gathered from official sources for African countries spanning the period 2000 to 2018. Due to data limitations, the analysis is focused on a restricted sample of 24 African countries. Table 1 in the annex section provides a comprehensive list of the countries considered in the study, along with their average values for economic complexity, income inequality, economic growth, and the aggregate governance index.

Table 1. (Annex Table): List of 24 African countries (with years average value of main variables).

Sunol	Country	Inequality (Gini %)	Economic Complexity	GDP Growth	Composite Governance Index
1.	Algeria	34.9895	−1.2453	3.4105	−0.3180
2.	Angola	49.7105	−1.6393	5.8105	−1.5862
3.	Botswana	58.3437	−0.4817	4.2105	2.0015
4.	Burkina Faso	43.0754	−0.7724	5.6632	0.3948
5.	Cameroon	44.5532	−1.1455	4.2579	−1.2119
6.	Côte d’Ivoire	53.5105	−0.9178	3.3316	−0.6971
7.	Egypt	40.0100	−0.2221	4.3737	−0.2777
8.	Ethiopia	32.9263	−1.0329	9.0895	−0.2388
9.	Gabon	40.7895	−1.3211	2.2158	−0.6471
10.	Ghana	42.7105	−0.9226	5.9842	0.5683
11.	Kenya	46.0211	−0.4402	4.6000	−0.9434
12.	Malawi	45.7895	−1.0225	4.2158	−0.2448
13.	Mauritania	38.7737	−1.4806	3.4263	−0.3052
14.	Mauritius	37.9368	−0.1941	4.0947	1.4933
15.	Morocco	41.1421	−0.5115	4.1947	0.3767
16.	Mozambique	45.8632	−1.2001	6.7474	−0.2374
17.	Namibia	65.5421	−0.3862	3.6895	1.4379
18.	Nigeria	42.9211	−1.9044	6.2789	−1.2591
19.	Senegal	41.3737	−0.5462	4.1105	0.4722
20.	South Africa	62.7737	0.1466	2.7737	1.2286
21.	Tanzania	44.4842	−0.8186	6.5368	−0.2364
22.	Tunisia	39.7053	0.2049	3.2263	0.6446
23.	Uganda	44.1368	−0.7337	6.4947	−0.8465
24.	Zambia	54.9789	−0.7328	6.0842	−0.0843

The Gini index is used as the dependent variable, representing income inequality, which ranges from 0 to 100 percent (or 0 to 1). A Gini index score of zero signifies perfect income equality, while 100% (or 1) indicates absolute inequality. Data for the Gini index is sourced from the Standardized World Income Inequality Database (SWIID V.9.2), sourced from Solt (2020). This comprehensive database offers comparable Gini coefficient indices for 198 countries since 1960. However, our analysis is constrained to the years 2000 to 2018 in developing countries due to data limitations. Additionally, linear interpolation techniques were applied to estimate few missing Gini index data for specific countries in the study.

The Economic Complexity Index (ECI) is an important factor in this study, serving as a crucial variable. This index gauges a nation’s productive output composition, representing the collective knowledge embedded within its economic structures. This knowledge, cultivated, disseminated, and preserved through networks of individuals and organizations, plays a vital role in productive endeavors (Hausmann et al., 2014; Hidalgo, 2021). The ECI serves as an indicator of a country’s sophistication and capabilities within its productive structure. Its calculation draws on the diversity and ubiquitousness of a country’s products. Sophisticated economies, characterized by higher ECI values, harness extensive networks and a rich knowledge reservoir to manufacture a diverse array of knowledge-intensive products. Conversely, less complex economies, with lower ECI values, possess a narrower knowledge base, yielding fewer and less intricate products, necessitating smaller networks. Economic complexity assumes critical importance as it explains variations in income levels among countries and forecasts their economic growth trajectories.

The data concerning economic complexity is obtained from Harvard’s Growth Lab’s Observatory Economic Complexity (OPEC), as outlined by Simoes and Hidalgo (2011). As of 2020, the countries topping the list in terms of the complexity index were Japan, Switzerland, Chinese Taipei, Germany, and South Korea. Japan claimed the foremost position with a score of 2.19, while Nigeria exhibited the lowest economic complexity index globally at -1.86. Within the African context, Tunisia and South Africa revealed positive and higher levels of economic complexity, showing respective indices of 0.2049 and 0.1466 (see Table 1 for further details).

Governance indicators, as defined by Kaufmann et al. (2009) and applied by various scholars in different contexts, are briefly summarized by Beyene (2022) and Seppo (2020). The data pertaining to governance quality indicators are sourced from the World Bank’s Governance Indicator database, encompassing six key indicators widely employed for assessing governance quality. These include:

Voice and Accountability gauges citizens’ perceptions of participation in voting, access to free media, freedom of association, and freedom of the press.

Political Stability is indicator which measures the likelihood of government destabilization arising from unconstitutional acts of violence and terrorism.

Government Effectiveness captures people’s perception of the quality, capacity, and independence of public services, as well as values related to policy formulation and implementation.

Regulatory Quality indicator assesses the state’s ability to design and implement policies and regulations promoting private sector development.

Rule of Law reflects citizens’ confidence in accepting and following societal rules, the enforcement of property rights, the reliability of security forces, and the risk of criminal activity.

Corruption Control is a metric which measures the extent to which government power is used for private advantage, encompassing both minor and major forms of corruption.

The governance scores for these indicators range from -2.5 to 2.5, with zero as the midpoint, where higher positive values signify better governance quality (the highest being 2.5). In addition to individual governance indicators, the Composite Governance Index (CGI) was computed using Principal Component Analysis (PCA) and utilized in the model analysis.

3.2. Econometric model specification

To explore the impact of economic complexity and governance quality on income inequality, we employed a robust two-step system dynamic GMM estimator. This methodology was chosen to alleviate potential endogeneity concerns and reinforce the credibility of our results. The basic model is presented below: (1) $${\mathit{\text{Gini}}}_{\mathit{\text{it}}}=\mathit{\alpha E}{C}_{\mathit{\text{it}}}+\mathit{\beta G}{Q}_{\mathit{\text{it}}}+{\delta }_i{Z}_{\mathit{\text{it}}}+\mathit{\mu i}\hspace{0.17em}+\hspace{0.17em}\mathit{\epsilon i}$$(1)

In this equation, i represents the cross-sectional units, t the time dimension, ${\mathit{\text{Gini}}}_{\mathit{\text{it}}}$ is Gin coefficient representing income inequality, EC_it is the economic complexity index, GQ represents factors of governance quality, Z_it is a vector of control variables. The coefficients α, β and δ are parameters to be estimated. µ_i is a country-specific unobserved effect, and ε_ii is the usual error term.

According to Arellano and Bond (1991), the dynamic GMM model addresses country-specific unobserved effects by taking the first difference. This allows us to re-write equation (1)(1) $${\mathit{\text{Gini}}}_{\mathit{\text{it}}}=\mathit{\alpha E}{C}_{\mathit{\text{it}}}+\mathit{\beta G}{Q}_{\mathit{\text{it}}}+{\delta }_i{Z}_{\mathit{\text{it}}}+\mathit{\mu i}\hspace{0.17em}+\hspace{0.17em}\mathit{\epsilon i}$$(1) as follows: (2) $$\Delta {\mathit{\text{Gini}}}_{i,t}=\mathit{\alpha \Delta E}{C}_{i,t}+\mathit{\beta \Delta G}{Q}_{i,t}+\mathit{\delta \Delta }{Z}_{i,t}+\Delta {\mu }_i+\Delta {\epsilon }_I$$(2) where Δ is the first difference operator.

The examination of equation (2)(2) $$\Delta {\mathit{\text{Gini}}}_{i,t}=\mathit{\alpha \Delta E}{C}_{i,t}+\mathit{\beta \Delta G}{Q}_{i,t}+\mathit{\delta \Delta }{Z}_{i,t}+\Delta {\mu }_i+\Delta {\epsilon }_I$$(2) offers insights into how economic complexity affects income inequality. However, to include the interaction effect of governance quality in the link between inequality and economic complexity, we modify equation (2)(2) $$\Delta {\mathit{\text{Gini}}}_{i,t}=\mathit{\alpha \Delta E}{C}_{i,t}+\mathit{\beta \Delta G}{Q}_{i,t}+\mathit{\delta \Delta }{Z}_{i,t}+\Delta {\mu }_i+\Delta {\epsilon }_I$$(2) according to the approach by Lee et al. (2020) and Gazdar and Cherif (2015). This adjustment enables a thorough analysis of the factors influencing income inequality. The following is the modified model by bring interaction effect. (3) $${\mathit{\text{Gini}}}_{i,t}=\mathrm{\phi }{\mathit{\text{Gini}}}_{\mathit{\text{it}}-1}+\alpha 0\mathrm{E}\mathrm{C}i,t+\alpha 1\left(\mathit{\text{ECit}}\mathrm{*}\mathit{\text{GQit}}\right)+{\sigma }_i.\mathit{\text{GQit}}+{\delta }_{\mathrm{i}}.\mathrm{Z}\mathit{\text{it}}+{\mu }_i+{\epsilon }_i\mathrm{ }$$(3)

Equation (3)(3) $${\mathit{\text{Gini}}}_{i,t}=\mathrm{\phi }{\mathit{\text{Gini}}}_{\mathit{\text{it}}-1}+\alpha 0\mathrm{E}\mathrm{C}i,t+\alpha 1\left(\mathit{\text{ECit}}\mathrm{*}\mathit{\text{GQit}}\right)+{\sigma }_i.\mathit{\text{GQit}}+{\delta }_{\mathrm{i}}.\mathrm{Z}\mathit{\text{it}}+{\mu }_i+{\epsilon }_i\mathrm{ }$$(3) allows us to analyze the effect of economic complexity on income inequality and evaluate how the inclusion of governance quality indicators alters the relationship between economic complexity and inequality. To introduce dynamism, the present lagged value of inequality is integrated into the model. The parameters 𝛼₀ and 𝛼₁ signify the direct and conditional effects of economic complexity, respectively. These parameters enable the estimation of four aspects of the hypothesis.

• First, if α₀ > 0 and α₁ > 0, economic complexity will augment income inequality, and governance indicators will exacerbate this effect. This implies that economic complexity contributes to increased inequality, and the presence of governance quality parameters will intensify this impact.

• Second, if α₀ > 0 and α₁ < 0, economic complexity will increase income inequality, but the governance factors will weaken this effect. In other words, economic complexity increases inequality, but the presence of governance factors will lessen this.

• Third, if α₀ < 0 and α₁ > 0, economic complexity reduces income inequality, and the existence of governance factors will mitigate this effect. Economic complexity decreases income inequality, and governance factors will further decrease this.

• Finally, if α₀ < 0 and α₁ < 0, economic complexity negatively affects income inequality and governance factors, further exacerbate the adverse effect. In other words, economic complexity improves income distribution, and this effect is stronger in the presence of governance factors.

Two specification tests were performed to assess the general validity of the instruments for the GMM estimator: second-order serial correlation of disturbances and the Sargan over-identifying restrictions tests (Arellano & Bond, 1991). A failure to reject the null of the Sargan test indicates the validity of instruments and, consequently, the correct specification of the model. Similarly, to substantiate the absence of second-order serial correlation in disturbances, we should confirm the absence of second-order serial correlation (AR2).

4. Findings and discussion

4.1. Descriptive statistics

Table 2 presents a descriptive analysis of the average data over 19 years for the key variables examined in this study. The data indicates a slight improvement in the income inequality index for the region, decreasing from 45.56 in 2000 to 45.35 in 2018. The dataset also identifies countries with the highest and lowest Gini coefficients. Namibia holds the highest Gini coefficient at 0.661 (66.1 percent), closely followed by South Africa at 0.628 (62.8 percent). In contrast, Ethiopia stands out with the lowest Gini index of 0.329 (32.9 percent), suggesting a relatively more equal distribution of income compared to other African countries.

Table 2. Summary statistics for main variables used in the model.

Variable	Obs	Mean	Std. Dev	Min	Max
Gini Coefficient	456	45.502	8.088	32.6	66.1
Economic Complexity	456	−0.805	0.564	−2.337	0.513
Composite Governance Index	456	2.70e-09	1.000	−1.931	2.50
GDP Growth	456	4.784	3.209	−7.70	18.30
Inflation	456	8.722	18.807	−8.238	324.996
Population Growth	456	2.338	0.781	0.055	3.789
Public Expenditure	456	23.760	8.118	9.760	55.801

In the region, the economic complexity index is -0.805, with Nigeria recording the lowest average value at -2.34 and Tunisia achieving the highest average value at 0.513. The composite governance index, derived from six governance factors using Principal Component Analysis (PCA), ranges from -1.93 (minimum) to 2.50 (maximum), where higher values signify countries with more robust governance systems. The table further provides the average values of additional control variables as shown in the table.

4.2. Empirical findings and discussion

A critical concern in panel data regression is addressing the issue of country-specific unobserved heterogeneity. To tackle this, we adopt a two-step GMM estimation approach that utilizes lagged values of the dependent and predetermined variables as instruments. We conducted Sargent and Hansen tests, and the results are presented in two separate tables (refer to Tables 3 and 4). The findings reported in these tables reveal that the Arellano-Bond test for first-order autocorrelation (AR (1)) is statistically significant, confirming the appropriateness of dynamic panel model in our study. Moreover, the Arellano-Bond test for second-order autocorrelation (AR (2)) is not significant, upholding the validity of the instruments. Additionally, both Sargent and Hansen tests produce insignificant results, and the number of instruments listed in the tables are fewer than the number of countries. Thus, the test outcomes confirm the validity of the instruments used in our analysis.

Table 3. Empirical results with and without moderating effect.

Dependent variable (Gini)	Model 1(with ECI only)		Model 2 (with CGI only)		Model 3 (ECI*CGI)
	Coef	Std. Err	Coef	Std. Err	Coef	Std. Err
Gini Coefficient (-1)	1.0092***	.003097	1.0062***	.0022938	1.0007***	.0030039
Economic Complexity Index	.22801***	.0419858
Composite Governance Index			.05919*	.0353643
ECI*CGI					−.15755***	.0259503
GDP growth	.00267	.0050412	−.00038	.0047452	.04167***	.0126303
Inflation Rate	.00179**	.0007135	.00129 *	.0007323	.00379***	.0008344
Financial Institution Index	−1.3245***	.2015946	−.99874***	.1572571	−1.2587***	.1944986
Population Growth Rate	−.06561***	.0203965	−.08192***	.0237168	−.0678***	.0190942
Public Expenditure	.00045	.0011531	−.00186	.0011315	−.00295**	.0011567
_cons	−.23324***	.073971	−.25121***	.0647566	−.62933***	.1281773
Arellano-Bond test AR (1)	P-value = 0.000		P-value = 0.000		P-value = 0.000
Arellano-Bond test AR (2)	P-value = 0.319		P-value = 0.405		P-value = 0.487
Sargan test (P-value)	(0.125)		(0.291)		(0.282)
Hansen test (P-value)	(0.282)		(0.350)		(0.313)
Instruments	22		22		22

Table 4. Disaggregated effect of governance quality on income inequality.

Variable	Model 1 (with CC)	Model 2 (with GE)	Model 3 (with PS)	Model 4 (with RQ)	Model 5 (with RL)	Model 6 (with VA)
	Coeff	Coeff	Coeff	Coeff	Coeff	Coeff
Gini (-1)	1.0077***	1.0175***	1.0059***	1.0067***	1.0205***	1.096***
GDP Growth	.00167	.00256	−.0028	.0019	.0056	.0061
Inflation	.00263***	.0026***	.0018**	.0030***	.0027***	.0026***
ECI*CC	−.1667***
ECI*GE		−.18433***
ECI*PS			−.0512***
ECI*RQ				−.1660***
ECI*RL					−.1655***
ECI*VA						−.2462***
Financial Institution Index	−1.1186***	−1.0964***	−.9075***	−1.0507***	−1.267***	−1.333***
Population Growth	−.0694***	−.0643***	−.1022***	−.0746***	−.0909***	−.0748***
Public Expenditure	−.0007	−.0005	−.0021*	.0006	.0005	.0024*
Constant	−.2787***	−.2867***	−.1779**	−.2866***	−.3683***	−.3651***
Arellano-Bond test AR (1)	Pv-0.000	Pv-0.000	Pv-0.000	Pv-0.000	Pv-0.000	Pv-0.000
Arellano-Bond test AR (2)	Pv-0.441	Pv-0.418	Pv-0.396	Pv-0.520	Pv-0.423	Pv-0.707
Sargan test (P-value)	(0.150)	(0.190)	(0.484)	(0.391)	(0.164)	(0.123)
Hansen test (P-value)	(0.678)	(0.648)	(0.899)	(0.898)	(0.787)	(1.000)

Pv is the p-value.

Table 3 specifically presents the econometric results of three models in such a way to avoid multicollinearity issues during estimation. The first model is estimated using solely the economic complexity indicator (including control variables but excluding CGI), the second model employs only the composite governance index (excluding ECI), and the final model incorporates the interaction to capture the moderating role of the composite (aggregate) governance indicator.

The estimated results from models 1 and 2 (Table 3) indicate a positive relationship between economic complexity, governance quality, and income inequality. In simpler terms, both ECI and CGI contribute to a worsening of income distribution in the study area. The presence of such positive relationships can be attributed to several factors. Firstly, it may be linked to the current state of economic complexity and governance qualities in these countries, both of which are notably weak. The negative values in governance indices across most countries indicate poor governance systems, potentially contributing to the observed positive association. Secondly, issues such as the country’s level of risk, political instability, and institutional shortcomings, including labor market conditions, may further contribute to lower levels of employment and income distribution disparities, fostering inequitable income distribution. Lastly, the relationship may hinge on a country’s efforts and capabilities to generate employment in manufacturing and sophisticated service sectors. Given that many African nations struggle to produce knowledge-based and diverse commodities, their economies tend to exhibit a low level of complexity, consequently leading to these positive relationships. This finding aligns with the findings of empirical studies by Hasanvand et al. (2022), Bandeira et al. (2021), and Lee and Vu (2020), but contrasts with the results reported by Khanzadi et al. (2022) and Hasanvand et al. (2022).

Model 3 has included governance factors as a moderating role in the same table (Table 3). The coefficient of the interaction term is negative and statistically significant. This finding supports our second hypothesis, which states that economic complexity has a positive effect on inequality, but when combined with governance factors, it reverses the relationship. In other words, economic complexity alone results in an increase in inequality (Model 1), yet this rise is weakened when governance factors are considered (Model 3). Hence, it is imperative to account for the interplay between economic complexity and governance factors to realize a reduction in inequality.

In summary, the influence of governance quality and economic complexity on income distribution according to this study reveals that when examined in isolation, both factors contribute to an aggravation of income inequality. However, when the governance factor is allowed to moderate the relationship, the positive impact of economic complexity becomes less pronounced. These findings have significant policy implications, suggesting that to mitigate the adverse impact of economic complexity on income inequality, policy measures should be accompanied by improvements in governance factors rather than pursued in isolation.

Concerning the influence of control variables, the table indicates that the development of the financial system has a negative and statistically significant impact at the 1% level. This is consistent with established theories, as a well-functioning financial sector facilitates economic growth by efficiently allocating financial resources to productive investments which consequently, promotes job creation and contributes to a decrease in income inequality. On the other hand, inflation and GDP growth, especially in the interaction model (Model 3), significantly exacerbate income distribution issues.

Table 4 presents the findings of the interaction effects involving the disaggregated governance indicators. The results reported in this table consistently follow that of composite governance index (Table 3). Particularly, all models demonstrate that the interaction terms are both negative and significant. This indicates that when economic complexity is coupled with each governance factor, it results in a decrease in income inequality. This suggests that governance quality factors play a role in alleviating the adverse effects of economic complexity on income inequality. Moreover, the financial development index and population growth exhibit a negative association with income inequality, while GDP growth and public expenditure are found to be statistically insignificant.

5. Conclusions

The main objective of this study was to explore the relationship between economic complexity and income inequality in 24 selected African countries. Specifically, the study examined the moderating role of governance quality in the link between economic complexity and income inequality using the systems GMM panel estimation method with data from 2000 to 2018.

The findings of this study hold significant implications in the study area, revealing that economic complexity exacerbates income inequality in these nations. However, the crucial insight emerges when considering the moderating role of governance quality. Accordingly, the study demonstrates that, when the composite governance indicator is employed as a moderator, it reduces the adverse effect of economic complexity on income inequality. This implies that effective governance system plays a critical role in mitigating the negative impact of economic complexity.

In the context of policy implications, the study underlines the importance of prioritizing measures to enhance the quality of governance systems. By focusing on governance improvements, policymakers in these countries can effectively counteract the deteriorating effects of economic complexity on income inequality. This understanding of the interplay between economic complexity, governance quality, and income inequality provides a tailored framework for policymakers to pursue more equitable and sustainable development strategies.

Disclosure statement

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

Availability of data and materials

The data used for the current study are available from the following sources and can be openly accessed.

1. World Bank Database (World Development Indicators)

   https://datacatalog.worldbank.org/dataset/world-development-indicators.

2. World Bank Database (Worldwide Governance Indicators database)

   https://databank.worldbank.org/source/worldwide-governance-indicators.

3. Standardized World Income Inequality (Gini coefficient)

   https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/LM4OWF.

4. Observatory of economic complexity

   https://www.Library.hbs.edu/citations/the-observatory-of-economic complexity.

Additional information

Notes on contributors

Amsalu Bedemo Beyene

Dr. Amsalu Bedemo is a experienced educator and researcher in economics within Ethiopia’s higher education sector. With over seventeen years of experience in teaching and research, he brings a wealth of expertise to his roles. Dr. Bedemo pursued all of his academic degrees in Economics, ultimately specializing in Agricultural Economics for his terminal degree. Currently, he serves as an Associate Professor of Economics at the Department of Policy Studies at the Ethiopian Civil Service University. Dr. Bedemo’s research interests encompass a wide range of topics, including political economy analysis, poverty, inequality, and macroeconomic research. His dedication to addressing these pressing issues reflects his commitment to contributing meaningfully to both academia and the broader community.