Because of its importance for social justice, income inequality is now one of the most debated central issues in the social sciences (Zhou and Shi, 2022). Since the early 2000s, rising income inequality and structural transformation have become two of the most important phenomena in the world (Dabla-Norris et al. 2015). Because of their importance, these two concepts are seen as a problem that needs to be addressed by policymakers, economists, and other members of civil society in order to ensure better living conditions for current and future populations. When it comes to rich or developed countries such as the United States of America, income inequality has come to the forefront of the economic debate, with talk today of only 1% versus 90% (in the US) and the emergence of the middle class in Asian countries (China) (Yang and Greaney, 2016). These concerns about income inequality sparked the Occupy Wall Street movement in the US, and continue to raise intense questions around the world.
However, according to Marx's theory of marginal remuneration, a worker's income is directly linked to the added value he or she brings to the production process. According to Marx, workers are only paid for the time they spend working, while employers can make additional profits by exploiting workers' labor. This theory implies that each factor is remunerated at its marginal productivity, i.e. workers receive wages, capitalists profits and landowners rents. Consequently, reducing income inequality requires a fundamental transformation of production relations, where they exist. Roemer and Lee (1998) go a step further and view income inequality as a set of individual preferences and lifestyle choices. For them, workers make rational calculations, arbitrate between leisure and work, and will therefore not work for less than the equilibrium income. So, in short, they deny the existence of income inequality.
Friedman (1953) acknowledges the existence of income inequality, for whom it can act as a facilitator or incentive for investment. Income inequality encourages people to redouble their efforts and move up the income hierarchy. If the distribution of effort and efficiency were fair, it would be unfair to reward everyone equally. For Bosmans and Ozturk (2021), the reward principle tolerates and justifies income inequality as a result of the exercise of individual responsibility.
Today, income inequality is increasingly seen as the difference between the richest 1% and the poorest 50% (Piketty, 2015). This difference even lies in the definition of income. Today, we distinguish between two types of income: wage income and non-wage income. Wage income corresponds to what economic agents receive in exchange for their participation in the production of a good, commonly referred to as "wages". Non-wage income is made up of goods such as real estate (rents), transfers in kind and financial assets (investments). In most cases, economic agents have a mixed income, i.e. a wage and financial assets or housing. This new literature accepts the idea that capital income is growing faster than labor income, deepening the income inequalities seen in most developed countries (Piketty and Saez, 2013; Antonelli and Tubiana, 2020).
In the African context, although income inequality persists, economic growth has become the main concern of our time (Ibrahim, 2020). Since the early 2000s, however, this part of the world has been seen as a booming continent, with a decline in poverty, the emergence of a middle class, growing political stability and some trade liberalization (Busse, Erdogan and M¨uhleny, 2017).
However, questions remain about this growth spurt, which does not appear to be based on a solid mechanism that could ensure its long-term sustainability, such as "structural transformation." What's more, this sudden growth does not give the whole population access to productivity, as was the case in developed countries in the past. This subject is attracting a great deal of interest, as evidenced by the United Nations University (UNI-WIDER), which has launched a research program to study Africa's structural transformation and describes the conundrum of Africa's industrial underdevelopment as a situation in which Africans "work hard but are poor". Despite this, McMillan and Harttgen (2014) note the existence of structural transformation in Africa characterized by a decline in the agricultural workforce of almost 10 percent.
The idea of structural transformation has thus played an important role in most poor or emerging countries, especially in sub-Saharan Africa. For some pioneers, structural transformation is the transfer of resources (labor and goods) from agriculture to manufacturing and finally to services (Lewis, 1954; Kuznets, 1955; Munyi, 2020). In the broadest sense, it is defined as the redistribution of economic activity and labor to the main economic sectors, such as manufacturing and services, leading to the accumulation of economic growth (Marcolino, 2022). The traditional analysis of Lewis (1954), Kuznets (1955) and Robinson (1976) goes even further in this direction, dividing the economy into two "dual economy" sectors: (i) a traditional sector, mainly agriculture, characterized by low productivity, and (ii) a modern sector (industry) characterized by relatively high productivity. This structuralist vision has led some studies to define structural transformation as the share of employment in agriculture, industry, and services, or as the share of value added between these three sectors (Busse, Erdogan, and M¨uhleny, 2017; Baymul and Sen, 2018).
In contrast to these authors, the modern economic view decomposes total labor productivity into two components, namely intra- and inter industry productivity (McMillan and Rodrik, 2011; de Vries et al., 2012; de Vries, Timmer, and de Vries, 2015). Intra industry productivity, also known as the within-sector effect, is considered to be the variation in productivity within a sector. Inter-sector productivity, on the other hand, is the effect of changes in the sectoral allocation of labor, which we refer to as the reallocation or displacement effect. It is the latter that is taken as an indicator of structural transformation. This means that recent development economics literature has focused much more on the factors that drive productivity and growth within an economy.
However, economic theory suggests that productivity can be related to income inequality in different ways and that there are different approaches to explaining it. Productivity is an important determinant of income. For example, in a competitive economy, workers who produce more goods or services are more valuable in the marketplace and therefore tend to earn higher wages. Companies that can produce more efficiently also tend to make more profit, which can be translated into higher wages for their employees. For this reason, economists continue to debate how to increase productivity while reducing income inequality, and how economic policies can promote equitable and sustainable growth. To find answers to these questions, economists have redoubled their efforts to understand the relationship between income inequality and structural change. Numerous studies have examined this issue, both unidirectionally and bidirectionally.
a) Effects of income inequality on structural transformation
Income inequality is an indicator of the distribution of wealth in society (Djahini and Ayassou, 2016), since it is considered by some researchers to be a by-product of growth (marginal productivity). Thus, the study of income inequality has mobilized a large body of literature, with the aim of understanding how it arises, how it is maintained, and how it relates to the process of structural transformation.
Theoretically, much of the literature asserts that the effect of income inequality on economic development is positive. According to this view, greater income inequality encourages low-income earners to work harder to increase their incomes, which leads to development (Barro, 2000). Among capitalists, for example, rising income inequality promotes productivity by pushing people to outperform and compete ( Barro, 2000; Bosmans and Ozturk, 2021). The theory of endogenous growth, formulated by Romer in the 1980s, goes some way to explaining this view. For him, innovation and productivity are closely linked to worker compensation. According to this theory, investment in knowledge, technology and education is the key to boosting productivity and raising incomes. This approach emphasizes the importance of investing in human talent, research and technological development to stimulate the economy. Another important theory is that of the efficiency wage, proposed by Bentham in the 1970s. This approach suggests that firms can increase productivity by offering above-average wages. This can motivate workers to be more productive and work harder to keep their well-paid jobs. However, this approach can also lead to an increase in income inequality, as some workers are paid more than others, and there is little empirical support for this view of income inequality.
However, other studies conclude that income inequality could be detrimental to an economy. As a result, it would have negative effects on: (i) health and education (Wilkinson and Pickett, 2010; Luoa and Xie, 2020). Wilkinson and Pickett (2010) go a step further in their analysis of a sample of 23 countries and conclude that income inequality can have negative effects on several indicators of social development, including health, education, social mobility, trust... This conclusion is particularly consistent with the recent study by Ndiaye, Dagoudo and Mbengue (2021), who examine a sample of sub-Saharan African countries and find that income inequality does not explain economic development using a Solow model without human capital. Internationally, dependency theory shows that income inequality can lead to greater dependence on foreign capital and technology. Indeed, developed countries are often the "teachers" to which poor countries must conform. This conformity can limit the ability of developing countries to develop their own innovative industries and thus transform their economies.
On the other hand, income inequality could (ii) affect political stability in fragile countries, as certain ills such as insurgencies, civil wars and riots could spread, thereby reducing the process of structural transformation (Alesina and Rodrik, 1994; Alesina and Perotti, 1996). This led (Stiglitz, 2012) to question the classical assumption that income inequality could lead to disincentives such as corruption, nepotism, and favoritism. In response, Boyce (1994) introduced the approach that income inequality can lead to environmental degradation through CO2 emissions. Other studies conclude that income inequality has mixed effects on structural transformation. Considering the geographical dimension, the study by Fawaz, Rahnama, and Valcarcel (2014) shows that the effect of income inequality on structural transformation related to the service sector is statistically positive in developed countries, but negative in developing countries. From the above, it is clear that the debate on the relationship between income inequality and structural transformation is not yet settled.
b) Effects of structural transformation on income inequality
The study of the impact of structural transformation on income inequality is more in line with Kuznets' (1955) inverted U curve hypothesis. According to this hypothesis, income inequality rises in the early stages of development until it reaches a maximum and then begins to fall as labor shifts from the less productive to the more productive sector. According to this hypothesis, structural transformation has a negative effect on income inequality. Contrary to this result, others find that structural transformation has a negative effect on income inequality. For example, technological change distorts the labor market by retaining only skilled workers and penalizing unskilled workers (Antonelli and Tubiana, 2020). In fact, increased productivity often allows entire firms or industries to expand rapidly. Highly skilled workers and innovative firms that can keep pace with rising productivity are generally rewarded with higher income gains. Less skilled or innovative workers and firms, on the other hand, risk losing market share and seeing their incomes stagnate or decline further. On the other hand, higher productivity often allows firms to enter more competitive global markets. This can benefit higher-skilled workers who are able to compete with foreign competitors. Less skilled workers or those living in more closed economies are often at a competitive disadvantage, contributing to rising income inequality.
A look at the literature shows that the effects of structural change on income inequality can vary, as in the study by Baymul and Sen (2018), who find that structural changes related to the service sector increase income inequality in developed countries and reduce inequality in developing countries. Chambers (2010) reaches similar conclusions, taking into account the time dimension, and finds that structural changes increase income inequality in the short run and the opposite effect in the long run. Piketty's (2005) study concludes that structural change has a neutral effect on income inequality because he finds that the reduction in income inequality is mechanically linked to unexpected events such as wars, inflation, taxes, and natural disasters, which contradicts Kuznets' thesis.
As far as the first or second relationship is concerned, the result is that there is still no consensus on the link between income inequality and structural transformation. The causes of these contradictions may lie in the unidirectional analysis of this relationship or in the analytical data. For example, according to Lundberg and Squire (2003), income inequality and structural transformation may share the same determinants. Therefore, in this paper we analyze this relationship simultaneously. Compared to the existing literature, to the best of our knowledge, very few research studies address the bidirectional causality between income inequality and structural transformation. As a result, one of the main criticisms of early empirical work on income inequality and structural transformation is that causality could run in either direction.
What's more, despite the controversy over the empirical results observed, both groups of studies have never taken into account internal productivity and resource migration, which together form the basis of the structural transformation to explain income inequality. Compared to the existing literature, our work has the merit of linking income inequality to the two indicators of Overall labor productivity, namely (i) intra industry productivity (intra or internal effect) and (ii) inter industry productivity (reallocation or structural change effect), which have remained largely unexplored in current economic analysis (Escobar and Mühlen, 2019).
The aim of this article is to examine how income inequality interacts with structural transformation in Sub-Saharan Africa. Our motivation lies in the fact that, to our knowledge, very few studies have examined the relationship between income inequality and structural transformation in a bidirectional manner. Moreover, this study is the first to consider inter industry productivity as an explanation for income inequality. To this end we test the panel Granger causality proposed by Dumitrescu and Hurlin, (2012) on our main variables (income inequality after taxes and transfers, inter industry productivity, intra industry productivity and Overall labor productivity) and for robustness we subsequently apply a Granger non-causality test in heterogeneous panel data models, using the methodology developed by Juodis, Karavias, and Sarafidis, (2021). This method is valid in models with homogeneous or heterogeneous coefficients..
The rest of the article is organized as follows: section 2 presents the data and the evolution of income inequality and structural transformation in Sub-Saharan Africa; section 3 presents the analysis methodology. Section 4 analyzes and comments on the results, while section 5 concludes.