Is Inequality Slowing Down Africa’s Industrialization?

Africa has also experienced a decline in the level of industrialization for at least three decades. Examining the dynamics of industrialization, and its effect on inequality, therefore remains a strikingly topical issue. This paper assesses the effects of industrial transformation on inequality in Africa over the period 1980-2016. Using a sample of 48 African countries, we estimate a dynamic panel data model using the Generalized Method of Moments in System (GMM-S). Our results show that strong industrialization would reduce inequality in Africa. The robustness of the results is tested using a PSTR (Panel Smooth Transition Regression) model and a PTR (Panel Transition Regression) model. The study recommends that economic, social and environmental disparities be taken into account in the process of industrial transformation on the continent.


Published by SCHOLINK INC.
industrialization on inequalities. Moreover, the use of panel data makes it possible to trace the dynamics of behavior and their possible heterogeneity, and to reduce the risk of collinearity between explanatory variables (Sevestre, 2002). From a strategic point of view, the article approaches industrial transformation through the implementation of decisions aimed at translating the quality of life of populations into the long term. Logically, Africa cannot remain at the bottom of the industrial transformation ladder. However, sustainable industrialization matters for Africa's future. Thus, is there a threshold beyond which industrialization could have a negative or positive impact on inequality in Africa?
The objective of this paper is to assess the effects of industrialization on inequalities in Africa. Three types of inequalities are considered: income inequalities, housing inequalities and environmental inequalities. The application of the S-GMM and the robustness analysis by the PSTR and PTR confirm a U-shape between industrialization and inequality in Africa.
Following this introduction, the second section identifies some stylized facts. The third section presents a state of the art. The fourth section outlines the methodological choices. The fifth section presents and discusses the results. The sixth concludes and suggests policy recommendations.

Some Stylized Facts
Two stylized facts emerge from the observation of inequalities and industrialization in Africa.


Manufacturing value added remains very low in Africa Figure 1 shows that, unlike other regions of the world, Africa has been characterized by low manufacturing value added since 1995 due to poor governance, political instability, a weak institutional environment, poor business performance and, in particular, the lack of effective integration. In addition to these characteristics, there is a shortage of skilled labor, low technological competence, insufficient energy supply, poor infrastructure, and in particular a lack of diversification of the productive structure.

Socio-economic and environmental vulnerabilities remain a concern
Industrialization poses problems of housing, living conditions, and inequality. The problems of housing and living conditions suggest that slums are growing considerably in African cities. Sub-Saharan Africa has 199.5 million people living in slums. In some large African cities, nearly 80% of the people live in slums. Figure 2 shows that income inequalities are increasingly observed and are accompanied by housing inequalities, which, due to the lack of a sustainable waste management policy, promote pollution, deteriorate air quality and hinder the management of natural resources. The UN-Environment (2018) points out that Africa produces 62 million tons of urban waste per year or an average of 0.65 per person per day. Despite its low production compared to other regions of the world, Africa is expected to revise its urban waste production to 442 million tons per year by 2025. This high production is the result of the growth of the urban population, which has been poorly managed up to now.

Literature Review
The theoretical anchoring of the role of industrialization in the resurgence of inequalities can be attributed to Williamson (1965), who, starting from a spatial extension of Kuznets' analysis (1955), lays the foundations of the theory of development phases and illustrates regional disparities through the inverted U curve in three phases: (i) the phase of transition to an industrial economy, which is accompanied by an increase in regional disparities; (ii) the pre-mature phase, characterized by increased migration flows and the dynamics of market forces, leads to the maintenance of disparities (iii) the maturity phase, in which regional disparities show a relatively downward trend as per capita income increases. Later, Krugman (1994) and Krugman and Livas-Elizondo (1996), using a theoretical model, emphasized that industrialization generates disparities through the spillover effects associated with the modernization process of economies. Similarly, Ellison and Gleaser (1997)  finds that the process of industrialization generates an expansion of the tax base that contributes to accentuating income inequalities within the population. Davis and Henderson (2003) and Lee (1997) found that industrialization favors the emergence of urban polarization and increases income inequality. James and Mark (2000), using old data in the form of wage ratios and new data from captures of changes in the aggregate wage distribution, calculate Theil indices for production workers, find that 19th century industrialization did not increase income inequality. Lin et al. (2003), looking at urban employment in China, concluded that industrialization through sustainable urbanization of cities tends to reduce income inequality. Chuan (2008) using Simultaneous Equation Models, finds that industrial development increases farmers' incomes while reducing the income gap between urban and rural areas. Ocampo et al. (2009), Easterly (2003 in their studies of developed and developing countries, reported that industrialization has been accompanied by a lack of reduction in inequality and poverty, a lack of infrastructure and income-generating activities, and an expansion of the informal sector. Chen (2010) found through SVAR modeling that industrial development has contributed to increased income inequality in China. Hung andKucinsha (2011), Luo andZhu (2009) found that to generate the decline in inequality and poverty in China, the effect of the industrialization process is through economic growth. Lustig et al. (2013), Pamuk (2008 found that the industrialization process in the Newly Industrialized Countries has significantly promoted the decline of income inequality. Chong and Wu (2014) in order to provide a better perception of inequality in China, construct an empirical model circumscribed on 22 provinces during the period 1997-2010. At the end of the econometric investigations, they found that structural transformation and industrial modernization have contributed to mitigating income inequality. Wang et al. (2015), Wu (2014) in their studies on Chinese provinces, find that industrialization has a non-linear effect on income inequality. Antoci et al. (2014) interested in the environmental externalities generated by the industrialization process, used a Solow-type model. They found that the industrialization process has led to a reduction in inequality, poverty, as well as an improvement in people's well-being.

Methodological Strategy
The methodological strategy is presented in three successive steps: the empirical model, the estimation technique and the data.

The Empirical Model
The model is inspired by Wu and Rao (2016), Kanbur and Zhuang (2013), anchored on Robinson's (1976) demonstration that under the assumption of economic duality measured by two sectors, income inequality either increases or remains unchanged for a relatively long time during the industrialization process. The logarithmic equation of income inequality is described by the model (1) below: Where Y 1 and Y 2 and 1 2 et 2 2 are the logarithms of the means and variances in the two respective sectors. P 1 and P 2 are the respective population shares of the two sectors. If the aggregate income is distributed according to equation (2).
Then by substituting equations (2) and (3) into equation (1), we can write: by the Gini index. Following the specification of Anyanwu et al, (2016) which incorporates a range of explanatory variables and appears to be suitable for the African context, the compact version of the model to be estimated is specified below.
Where Inequalities it represents the inequality index of country i at date t captured by three measures.
Unlike previous studies, we postulate that industrialization affects inequality along three dimensions: (i) the income dimension that describes the income gap between the affluent who can afford better living conditions and those forced to live below the monetary poverty line; (ii) the housing dimension that differentiates those living in slums from those living in well-built houses; (iii) the environmental dimension that is the subject of particular renewed interest in development studies with respect to its unrationalized management, creates further inequalities. Inequalities it-1 is the index of lagged inequality of one period; U it is the matrix of variables of interest. X it is the matrix of control variables consisting of the log of GDP per capita, domestic investment, FDI inflows, youth unemployment rate, and institutional variables captured by the state of governance and democracy. The exploded models are specified below.
γ Unemployment +γ Gover +γ Democ +t +δ +ε (8) t t , δ i , and ε it are respectively the time fixed effects, the country fixed effects and, the rest of the perturbation.
Income inequalities are approximated by the Gini coefficient. Housing inequalities are measured by the proportion of the population living in slums or in non-decent housing. Environmental inequalities are captured by the number of people living on the margins of nature protection or living in insalubrious conditions marked by poor garbage management. The lagged inequality variable allows us to assess the memory effect of past inequalities on present inequalities.
The matrix of variables of interest consists of industrialization approximated by manufacturing value added relative to GDP, and industrialization squared. As shown by Kuznets (1955), these variables allow us to determine the breakpoint and thresholds of the transition variables. We postulate a U-shaped relationship between industrialization and inequality.
The explanatory variables include: (i) real GDP per capita which captures the standard of living of the country's population. We postulate that income disparities in Africa increase inequality in urban areas; (ii) domestic investment proxied by gross fixed capital formation. Lee et al, (2013)  Africa. Moreover, the "pollution haven" hypothesis will be verified in the context of the estimation of environmental inequalities; (iv) the unemployment rate, which is approximated by the number of unemployed young people in relation to the total number of young people. It naturally increases inequalities and its effect depends on the measure of inequalities. Indeed, unemployed young people are better at protecting the environment and thus improving the living environment; (v) local governance; (vi) democracy, whose indices are between -2.5 and 2.5.

The Estimation Technique
Assuming that the level of past inequality would influence the level of inequality in the current period, we opt for a dynamic panel whose Ordinary Least Squares ( The study period chosen is from 1980-2016, justified by data availability. The descriptive statistics contained in Table 1 show the absence of variation. It is generally accepted that when the data do not fluctuate, the results tend to converge. The correlation matrix (Appendix 2) shows weak interdependencies, which suggests an absence of multicollinearity between the dependent and explanatory variables.

Results
We present results from the baseline models and the robustness analysis.

The Results of the Basic Models
We estimate the model under three specifications ( of the industrialization process plays a decisive role in the exploitation of the productive forces of the cities, insofar as the sectors that provide jobs for young people abound and exploit the possibilities of the demographic dividend, thus helping to reduce income inequalities. These results corroborate those of Zhang (2017).
The estimation of model 2 also shows two main results overall. On the one hand, manufacturing value added, domestic investment, governance and FDI have positive and statistically significant effects on housing inequality. The underlying explanation is that industrialization in Africa is essentially unprincipled. As a result, unsuitable investments are favored, which, for lack of a housing policy, increase housing inequalities. In some countries, the establishment of FDI is done to the detriment of anarchic constructions, forcing some inhabitants to settle in slums. On the other hand, accelerated industrialization and GDP per capita contribute to a significant reduction in housing inequalities. Indeed, the alignment of an industrial policy forces companies forced to exploit the productive forces of cities to respond to urban demand by building housing in advance. In addition, compliance with the standards set by UN-Habitat creates sustainable industrialization, less well served and more compartmentalized, which, through residential compartmentalization, reduces housing inequalities. The results obtained corroborate those of Eeckhout et al. (2010).

Income inequalities Housing inequalities Environmental inequalities
Incom_ineq ( The estimation of model 3 shows two main results. On the one hand, manufacturing value added, unemployment and the level of democracy have positive and statistically significant effects on environmental inequality. The underlying explanation is that industrialization in Africa is essentially characterized by non-compliance with environmental standards. Consequently, this situation, due to climate change, contributes to the deterioration of air quality and generates moderate, often localized environmental impacts (odors, noise pollution, pollution) and increase the environmental and carbon footprints of cities due to car traffic. It is also noted that the unemployment situation in African cities favors the establishment of small informal industries that cause a nuisance to the immediate environment due to the lack of restrictions on zoning and density, the immediate environment due to pollution, noise and odors. However, accelerated industrialization, GDP/capita, and domestic investments contribute to significantly reduce environmental inequalities. Indeed, the alignment of industrial policy forces firms to exploit the productive forces of cities, and to respond to urban demand without policy support, in the form of a favorable regulatory framework, training and skills-building opportunities, and prioritization of infrastructure that supports value chains. In this way, it enables resilient, green, cross-sectoral, and multilevel governance, better intrinsic democracy, and environmentally friendly urban mobility through multimodal options. The findings corroborate those of Wei et al. (2017).

Robustness
To guarantee the robustness of our results and to highlight the non-linear effect of the model represented here by the square of industrialization, we estimate two panel, non-linearity models. The Developed by Hansen (1999), PTR models imply that individual observations can be divided into homogeneous classes based on the value of an observed variable. Specifically, these models assume a transition from one regime to another based on the value of a threshold variable. The change is important because it accurately specifies the inflection point between the evolution of the independent variable (industrialization) and that of the dependent variable (inequality). PTR models are criticized Empirical evidence using the PTR and PSTR models is more widely observed in finance. Gonzà les et al. (2005) illustrate the imperfections of capital markets on investments. Eggoh and Villieu (2013) re-examine the non-linearity between financial development and economic growth. A few studies in environmental economics have used these models to appreciate the nonlinearity between endogenous and exogenous variables (Heidari et al., 2015). In addition to adding to the empirical literature on nonlinearity between macroeconomic and environmental variables, we test for model linearity before estimating the PTR and PSTR models. This exercise, which has been swept up in several works, allows us to determine the existence of a non-linearity, and to determine the variable causing the non-linearity and on which the study focuses. Several tests of linearity are proposed in the literature (Ramsey, 1969;Tsay, 1989;Hansen, 1996). The test of Tsay (1989) In this form, Tsay (1989) postulates the null hypothesis of linearity: 0 : 1 = 2 ( = 1 … … ) against the alternative hypothesis: 1 : 1 ≠ 2 ( i=1......m). Subsequently, we test the nonlinearity relationship in a regime-switching model. The determination of the two regimes is identified by the threshold effects estimation method with smooth panel transition (PSTR). The PSTR technique allows an endogenous determination of the thresholds. These advantages are essential. They are suitable for explaining the gradual effect of the change in the relationship in relation to a higher and higher level.
The crowding-out effect does not appear to be radical, hence the need for a certain degree of gradualism if African states decide to accompany their urbanization policy with some reforms requiring the reduction of inequalities.
To detect the non-linearity relationship between industrialization and inequality, we use a PSTR model as developed by Gonzales et al. (2005). Let us consider y it as the explained value, x it as the explanatory variable and u it as the transition variable. We present the simple case of a PSTR with two regimes and a simple function between industrialization and inequality in Africa. The choice of Africa is justified by the fact that this region concentrates multiple forms of inequality and has undergone a considerable www.scholink.org/ojs/index.php/jepf Journal of Economics and Public Finance Vol. 7, No. 4, 2021 Published by SCHOLINK INC.
industrialization process. The model takes the following form: Where g(u it ,γ,c) is the transition function. This function is continuous and depends on the threshold c of the transition variable u it ; γ is the transition parameter. The transition function is a normalized and bounded function between 0 and 1, with extreme values associated with the coefficients β 0 and (β 0 + β 1 ). Gonzales et al. (2005) consider this function to be a logistic transition function of the form: The slope of the parameter γ determines the smoothness of the transition. For m=1, the model exhibits the two regimes separating the lower and upper values of u it with a simple monotonic transition of the coefficients of β 0 and (β 0 + β 1 ) when u it increases. As the slope of the parameter increases, the transition becomes rougher and the transition function g (u it ,γ,c) becomes a function of type g (u it ,γ,c). When the smoothing parameter tends to infinity, the transition function is equal to unity, i.e., g (u_it,c)=1 if u it >c; the transition function is zero (g (u it ,γ,c)=0) otherwise (u it <c). When γ is close to 0, the transition function is a constant. In this case, the PSTR tends to the PTR as developed by Hansen (1999). In general, for all values of m, the transition function g (u it ,γ,c) is constant when γ is close to 0.
The procedure for estimating the PSTR model requires three steps (Gonzales et al., 2005) 1 the sum of the residual squares under H 1 (the PSTR model with two regimes), the Wald test can be written as follows: The maximum likelihood test can be written as follows: The parameters ( 0 , 1 , , ) of the baseline model are estimated in two steps. In the first, we remove individual effects by subtracting the means. In the second step, we apply Nonlinear Least Squares (NLLS) on the transformed data to determine the parameter values that minimize the sums of squares of the residuals.
The number of regimes test allows determining the number of regimes of the transition function. In most cases, the existence of a decoupling phenomenon in the presence of one regime is shown. But, in www.scholink.org/ojs/index.php/jepf Journal of Economics and Public Finance Vol. 7, No. 4, 2021 Published by SCHOLINK INC.
the presence of two regimes, a more pronounced transition is observed. Gonzales et al. (2005) propose a sequential approach to test the null hypothesis of the number of non-linearities in the transition function. In the basic theory of PSTR model estimation, we assume that the linearity hypothesis is rejected. The task is then to test whether there is one transition function ( 0 : = 1) or whether there are at least two transition functions. To test for the number of regimes that link urbanization and inequality, we draw on the procedure of Gonzales et al. (2005). The authors then assume that if the function has the following form: We should test the nullity or not of the parameter β_2^'. Thus, the test for nonlinearity is defined by 0 : 2 ′ = 0. Let us note SSR 0 the sum of squares of the residuals under H 0 then denoting a PSTR with a transition function. Let us also note SSR 1 the sum of squares of the residuals under H 1 , i.e., the transformed model. Taking these assumptions into account, the regime number is given if the null hypothesis is rejected (Note 1).  Industrial transformation reduces income inequality up to the threshold of 24.55%. This is justified by the fact that when a continent industrializes, there is an income catch-up. People working in the primary sector see an increase in their income as their production is purchased locally and processed.
The strong local demand then contributes to increasing the purchase price in the field. Progressively, The structural transformation reduces environmental inequalities before the rate of 21.60%. It is as if industrial production, especially in urban centers, increases air pollution. On a small scale, this pollution is not felt, but when industrial zones are created near urban centers, environmental degradation is felt by the urban population. The people who suffer the most from this degradation are those with limited financial means. It is important to note that of the three industrialization thresholds identified, the industrial transformation will quickly impact the environment. The value of the transition parameter is γ=0.926. This result reveals the dilemma that African governments must resolve.
Sustainable consolidation taking into account the protection of the environment induces the establishment of free trade zones.
Industrial transformation reduces the inequalities of habitat before the rate of 32.75%. Indeed, the construction of new factories in peripheral and urban areas will decongest urban centers, thus reducing the rate of informal settlements and the proliferation of slums. After this threshold, the attractiveness of the industrial zone, whose construction has generated positive externalities, will contribute to the Foreign direct investment does not affect inequality. Democracy presents results that differ from one form of inequality to another.

Conclusion
This paper has assessed the effects of industrialization on disaggregated indices of inequality (income, environmental and housing). Although exacerbated, these inequalities, to the best of our knowledge, have not been the subject of simultaneous empirical investigations. We implement it econometrically on a panel of 48 African countries over the 1980-2016 time horizon using the Generalized Method of Moments in System. The results indicate that industrialization significantly reduces inequality in Africa. To achieve these results, we mobilized a theory anchored on developments related to the Kuznets curve. A dynamic panel data model was used as the econometric basis, with a quadratic (non-linear) specification.
The results of the basic model remained robust by applying the PTR and PSTR models. Their implementation allowed us to determine the critical thresholds of urbanization and industrialization that could reveal a decoupling effect on the different declines of inequalities. Specifically, below the www.scholink.org/ojs/index.php/jepf Journal of Economics and Public Finance Vol. 7, No. 4, 2021 Published by SCHOLINK INC.
Above these thresholds, industrialization would increase inequality provided that good industrial transformation policies are in place.
Four main recommendations can be made: (i) economic, social and environmental disparities must be taken into account; (ii) the implementation of sanitation or waste flow management policies must contribute to reducing inequalities between rich and poor neighborhoods; (iii) in the search for sustainable cities, the decoupling of socioeconomic, demographic and territorial growth from resource scarcity and environmental degradation requires the management of public actions; (iv) the industrial transformation of Africa must be aimed not only at increasing productivity, but also at improving the quality of life of the population.