Are productive capacities cursed? Evidence on the income from natural resources for African economies

doi:10.21203/rs.3.rs-4712681/v1

Download PDF

Research Article

Are productive capacities cursed? Evidence on the income from natural resources for African economies

https://doi.org/10.21203/rs.3.rs-4712681/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Over the past few decades, numerous studies have emerged on questions of natural resources. Despite this abundant literature and the importance of productive capacities for sustainable development, there is almost no study on the effect of natural resources on productive capacity. In order to fill this gap, this study analyzes the effects of natural resources on the productive capacities of African countries. Using data from 42 African countries over the period 2000–2018, we show that natural resources have on average a negative effect on the productive capacities of African countries. By distinguishing between different natural resources, we see that the effect is more amplified on diffuse resources. Our results also show that there is a U-shaped relationship between natural resources and productive capacities. Nevertheless, we note that the threshold from which natural resources improve productive capacities is included in the interval [33.90–43.85]. We therefore recommend that the leaders of African countries establish institutional mechanisms to ensure effective management of revenues drawn from natural resources and allocated to productive capacities. This involves strengthening the institutional sphere that guarantees political stability.

JEL classification: O40, O55, O43.

Productive Capacities

Natural resources

Institutions

A country's capacity to produce is linked to essential elements that promote equitable and sustainable growth and development. It has attracted great political and scientific attention over the last twenty years (Silvestre and Ţîrcă, 2019). With the aim of putting them on the path to sustainable development and making their economies more resilient to shocks, the improvement in the productive capacities ¹of developing countries and in particular African countries is dramatic (Gnangnon, 2021). According to UNCTAD ²(2020), improving and utilizing productive capacities represents a holistic approach to economic development. Thus, to study the effect of production capacity on limited resources, it is essential to develop the items of the productive capacity index and put them into practice (Xin et al., 2023).

Developing countries, more specifically Africa, rely heavily on the exploitation of their natural resources to significantly increase their national income (Gao and Tian, 2016; Hassan et al., 2019; and Xue et al., 2021). Natural resources are not harmful to the environment, but their extraction process can be. In Africa, its extractive sector is estimated to represent around 30% of global mineral reserves (UNEP, 2022). Africa is home to 40% of the world's gold deposits and around 90% of chromium and platinum (Aladejare, 2020; and UNEP, 2022). Similarly, Africa's rich oil and natural gas reserves represent 12% and 8% of global reserves, respectively (UNEP, 2022). If this wealth is an argument in favor of the development of the continent, it still does not benefit the populations enough and many African economies struggle to convert it into economic development (Rahim et al., 2021).

Several reasons explain this situation, described as a raw materials curse in Africa. For this, inadequate technological know-how, post- colonialization effect, unregulated market of natural resources, political instability, terrorism and many other deep-rooted problems are responsible for low economic growth in Africa. (Adekunie, 2021). The challenge for most African countries is therefore to be able to transform these raw materials into a source of growth and to diversify economically to depend less on these same resources.

Several works have supported the resource curse hypothesis (Sachs and Warner, 1995; Auty, 1997; Asif et al., 2020; Kim et al., 2020; Hartwell et al., 2022). This work has demonstrated that countries with abundant natural resources develop more slowly than those with limited natural resources, in part because the abundance of natural resources slows their rate of economic growth (Ahmed et al., 2016). On the other hand, many works have refuted the resource curse hypothesis and argued that the use of natural resources helps to enhance economic growth (Leite and Weidmann, 1999; Auty, 2001; Arin and Braunfels, 2018). Auty (1993), Sachs and Warner (1997) and Karl (1997) found that resource-rich countries had low economic growth compared to resource-poor countries.

The literature has continued to explore the explanatory factors of this phenomenon. Thus, various economic, political, institutional and even environmental factors have been mobilized to explain this sensation (John, 2011). Thus, the impact of natural resources is conditioned by the quality of institutions (Mehlum et al., 2006, Sala-i-Martin and Subramanian, 2008). However, there is also the institutional curse of natural resources which reflects the fact that natural resources would prevent the development of good institutions or institutional reforms conducive to economic growth (Leite and Weidmann, 1999; Isham et al., 2005; Sala-i-Martin and Subramanian, 2008).

Work related to the natural resource curse in recent years has focused on aspects of development such as innovation (Kamguia et al., 2022), wealth inequality (Njangang et al., 2022), access to drinking water and sanitation (Tadadjeu et al., 2022), political empowerment of women ( Awoa et al., 2022), human development ( Nchofoung et al., 2021), agricultural productivity ( Dorinet et al. 2021) and, export diversification ( Djimeu and Omgba , 2019). Other studies have extended the natural resource curse on structural transformation (Baldursson and von der Fehr, 2018; Xin et al., 2023). Despite this abundant literature on the natural resource curse, very little work exists on how natural resources affect productive capacities.

The contribution of this article is therefore threefold. First, it directly examines the relationship between natural resources and productive capacities. More precisely, we exploit differences in the level of natural resources between countries and over time to explain differences in the perception of productive capacities. The second contribution is to examine the different types of natural resources³ (Cockx and Francken, 2016) on productive capacities. This approach will provide a better understanding of the dynamics of productive capacity driven by natural resource types by providing a generalizable conclusion on why natural resource-rich economies are prone to poor economic performance. Third, this article goes beyond the direct relationship between natural resources and productive capacities by examining in a more appropriate non-linear framework deploying a dynamic threshold technique (Sirag et al., 2018; Sini et al., 2021). Indeed, the impact of natural resources can follow a U or inverted U shape (Liu et al., 2022; Gu et al., 2023).

To evaluate the effect of the productive capacity index on the rent of natural resources, we rely on the productive capacity index (PCI) taken from UNCTADstat database. The global index summarizes the state of productive capacities of economies around the world by calculating scores between 0 and 100 (limits not included). Thus, countries with high levels of productive capacity achieve scores close to 34. The global ICP index is subdivided into eight categories. On the other hand, to measure the rent of natural resources, we use the total rent of natural resources as a percentage of GDP as a measure of the variable dependent on natural resources (TNR). Natural resource data comes from WDI.

The estimates are carried out on asample of 42 countries in Africa over the period 2000-2018 using the fixed effect OLS estimator with standard deviations corrected by the estimation of Driscoll and Kraay (1998). This estimator has the advantage over the traditional fixed effects estimator of correcting the heteroscedasticity and autocorrelation of the residuals, but above all of taking into account the interdependence of individuals when the study period is long. We also addressed the endogeneity that the model suffers from by using the system GMM technique in the base estimation. As robustness tests, we performed the analyzes with the alternative measures of natural resources. We controlled for endogeneity with the IV-2SLS technique as a secondary method. Our results show a strong, negative and significant effect of the productive capacity index on natural resources, suggesting that natural resources decrease as ICP increases. This result remains robust to a large number of robustness checks that include alternative measures of natural resources.

In addition to this introductory section, the remainder of the article is structured as follows:

The second section presents a review of the theoretical and empirical literature on the relationship between the natural resource curse and productive capacities. The third section describes the methodology adopted. Section 4 presents the data, methodology and discusses the econometric results, while section 5 provides the conclusion and policy recommendations of the study.

To our knowledge, work on the natural resource curse and its association with productive capacities does not exist. However, several studies have been conducted to examine the relationship between natural resources and economic development. Recent studies have also shown that economies rich in natural resources (NR) are often less developed than those without them. The following analyzes provide more information on the scale of the phenomenon (Sachs and Warner, 1995; Asif et al., 2020; Khan et al., 2020). The question of the influence of the ownership of natural resources on economic growth has concerned some researchers who have made it a priority area of study. However, reading theoretical and empirical work on this subject reveals certain inconsistencies. Indeed, two hypotheses are opposed: the orthodox or optimistic hypothesis and the heterodox or pessimistic hypothesis.

2.1. Theoretical proof

According to proponents of orthodox or optimistic thinking, the ownership of natural resources has a favorable impact on economic growth. To support their point of view, they put forward both microeconomic and macroeconomic considerations.

According to microeconomic reasoning, the positive impact of natural resource ownership on economic growth can be quantified in terms of costs and benefits. Indeed, the exploitation of these resources will only be profitable if their prices on world markets exceed the marginal costs of extraction and transport, allowing these countries to obtain foreign exchange through exports, which generally have a positive impact on the country's economic growth.

Regarding the macroeconomic reasoning put forward by Smith (1776), Ricardo (1817) and other economists in the theory of international specialization, the difference in growth in developing countries can be explained by the existence or absence of natural resources. According to the latter, the presence or abundance of natural resources in a country can provide it with comparative advantages and encourage it to specialize either in their production or in their manufacturing, the latter serving as intermediate consumption. As a result, specialization will increase production while allowing economies of scale and boosting economic growth. Using the general equilibrium model, neoclassical economists such as Hecksher (1919), Ohlin (1933) and Samuelson (1953) move in the same direction, but with an emphasis on capital and arriving at a new model, the HOS model. The theoretical analyzes of Posner (1961) and Vernon (1966) go in the same direction, but with much more emphasis on innovation, that is to say the discovery of a new product or a new natural resource.

Contrary to conventional wisdom, the current heterodox view is that the possession of natural resources is a curse, implying that it has a negative impact on economic growth. It is made up of contributions from several economists, including Sachs and Warner (1995, 1999), Auty (2004), Paulo and Gary (2010) and Sala and Subramanian (2013). According to the latter, the negative link is explained by several socio-economic and environmental consequences of the extraction of primary resources, in particular: the contraction of economic activity in other sectors, the appreciation of the currency, conflicts, degradation of the ecosystem, etc.

Regarding the contraction of economic activities in other sectors caused by the discovery and exploitation of a new natural resource, she explains it by a movement of the workforce towards a more productive and better sector paid. This phenomenon is known as “Dutch disease”. Indeed, an increase in the wages of specific economic actors leads to an increase in demand and an increase in the general price level. In this case, the competitiveness of domestic products would suffer, as would exports and economic growth.

Robinson et al. (2006) focus on the political economy framework, and show how political incentives are essential for understanding the effect of resources on growth. However, Bulte et al. (2005) show that the abundance of natural resources promotes economic growth which they explain by the windfall flow (this flow has a direct and indirect effect, namely direct on the economy and indirect through the improvement of institutional quality) of revenues from resource extraction. Speaking of the curse in the literature, in Angola and Liberia there was the discovery of the diamond and this discovery rather caused conflicts instead of generating a healthy environment in terms of improving growth (Olsson, 2007; Le Billion, 2008). Similarly, the discovery of oil in Norway was a major blessing, but in Nigeria and Equatorial Guinea the discovery of oil was a source of curse and it has different effects in each country (Larsen, 2004; same, 2008; Sala and Subramanian, 2013). The Democratic Republic of Congo, for its part, although it is strongly endowed with sufficient natural resources, it has experienced conflicts and a deterioration in growth and standards of living (Lalji, 2007).

2.2. Empirical Evidence

In this section, we briefly explore the extensive literature on the relationship between natural resources and economic growth.

Ekodo and Ndam (2019) assessed the effects of income from the exploitation of natural resources, particularly oil, on economic growth in the CEMAC region over the period 1996-2016 and found, based on the GMM method, that these incomes have a positive impact on economic growth in this region.

Sachs and Warner (1995) test the effect of natural resources on long-term economic growth and find that resource-rich countries grow more slowly than resource-poor ⁴countries. The literature published after Sachs and Warner (1995) mainly studies the different transmission mechanisms of how natural resources affect growth, determining whether the natural resource curse can be avoided by improving the quality of institutions, or whether the existence of the natural resource curse depends on the means of measurement and the type of natural resources.

Several other studies analyze the importance of institutional quality and find that the natural resource curse can be avoided if institutional quality is good enough ( Isham et al., 2005 ; Mehlum et al., 2006 ; Arezki and Ploeg , 2007 ; Boschini et al., 2007; Horvath and Zeynalov , 2014). Brunnschweiler and Bulte (2008) distinguish between resource dependence (the degree to which nations rely on exports of natural resources) and resource abundance (a measure of resource wealth) and, unlike Much previous research treats institutions as endogenous. Although they are unable to establish a relationship between resource dependence and growth, they demonstrate that resource wealth is associated with better institutions and higher growth. Therefore, their findings do not support the existence of the natural resource curse.

According to Sala and Subramanian (2013), new oil discoveries tend to generate an appreciation of the real exchange rate while harming other export sectors of the economy. Gylfason and Zoega (2006) study another path and find that natural resource wealth crowds out human and physical capital, leading to slower growth in the long run. Another body of research studies the influence of natural resources on factors other than economic growth. Natural resource wealth can reinforce corruption, increase political instability and the risk of war, and hinder the functioning of democratic institutions (Barro, 1999; Ades and Tella, 1999; Ross, 2001; Jensen and Wantchekon, 2004; Collier and Hoeffler, 2005;

Lederman and Maloney (2003) estimate both cross-sectional and panel regressions and find that the results differ. Panel regressions show that natural resources have a significantly favorable influence on economic growth, but cross-sectional regressions show a negative but negligible effect. Ades and Tella (1999) analyze both cross-sectional and panel data and conclude that by using panel data, the influence of natural resources on economic growth becomes minimal.

The work of Sachs and Warner (1995; 1999) uses a large panel of countries to study the effect of natural resources and it is measured by the share of exports of primary products on GDP. Negative result of the effect of natural resources on growth. Likewise, Gylfason et al., (1999) and Auty (2001), using different variables and specifications, obtained the same result of the negative effect. Thus, obtaining negative results demonstrates what is commonly called the resource curse.

For Wantchekon (2002), using 141 countries over the period 1950-1990, demonstrates that an increase of 1% in the share of primary exports in GDP increases by 8% the probability that a government exercises authoritarian power.

On the other hand, the work of Mavrotas et al. (2011) on 56 developing countries during the period 1970-2000, find that natural resources positively affect growth when the country has a good level of governance and democracy. It is therefore recognized that wealth from natural resources affects the type of regime in place.

Brunnschweiler (2008) using OLS and DMC, finds that natural resources have a positive and direct relationship with the economic growth of a country. Supporting previous studies, Herb (2005), Boyce and Emery (2005) also suggested that natural resources increase GDP per capita and other factors of economic growth.

3.1. Data

This study uses data from a balanced panel of 42 African countries over the period 2000 to 2018. The periodicity and sample size are chosen based on data availability.

3.1.1. Dependent variable

We use as the main dependent variable the productive capacity index ⁵(PCI) taken from UNCTADstat database. The global index summarizes the state of productive capacities of economies around the world by calculating scores between 0 and 100 (limits not included).

In accordance with Fig. 1, this index is relatively below the average value whatever the sample of African countries selected. More precisely, countries with high levels of productive capacity achieve scores close to 34.

The ICP also offers country scores to help understand sources of systemic vulnerability and identify drivers of economic growth, including progress towards achieving national and global development goals. For statistical and measurement purposes, the overall ICP index is subdivided into eight categories contained in the following table:

Table 1

*description of the variables of the productive capacity index*.
Variables	Definitions
Human capital	Covers the education, skills and health conditions of the population, as well as the overall integration of research and development within society through the number of researchers and expenditures on research activities. The gender dimension is reflected by the fertility rate which, with each increase, reduces the human capital score.
Natural capital	Estimates the availability of extractive and agricultural resources, including the rents generated by the extraction of the natural resource less the cost of its extraction. To capture commodity dependence, natural capital decreases as material intensity increases.
Energy	This category measures the availability, sustainability and efficiency of energy sources. Therefore, it consists of energy use and access, distribution losses and the renewable nature of energy components and sources, and includes the GDP generated by each unit of oil to further emphasize the importance of optimal energy systems.
Transportation	Measures the ability of a system to transport people or goods from one location to another. It is defined as the capillarity of the road and rail network and air connectivity.
Information and communication technologies	Allows us to estimate the accessibility and integration of communication systems within the population. It includes landline and cell phone users, internet accessibility and server security.
The institutions	Aim to measure political stability and effectiveness through the quality and effectiveness of regulation, the success of the fight against crime, corruption and terrorism, and the safeguarding of freedom of expression and association of citizens.
The private sector	Defined by ease of cross-border trade, which includes time and monetary costs to export and import, and business support in terms of domestic credit, speed of contract execution and time to start a business.
Structural change	Refers to the movement of labor and other productive resources from low- to high-productivity economic activities. This change is currently taken into account by the sophistication and variety of exports, the intensity of fixed capital and the weight of industry and services on total GDP. Structural changes can also occur in a given sector, provided that the major constraints of the sector are identified and addressed effectively.
Source: authors

3.1.2. Independent variables

In accordance with recent work by Kamguia et al, (2022) and Mignamissi and Kuete (2021), we use the total rent from natural resources as a percentage of GDP as a measure of the natural resource dependent variable (TNR). This variable is obtained from the World Bank database, (2022). Indeed, it is one of the most widely used measures in the resource curse literature, given its availability. This indicator is the sum of five rents including oil, natural gas, coal, minerals and forestry, as a proportion of GDP.

3.1.3. Control Variables

To support the relationship between the curse on productive capacities and natural resources and to avoid variable omission bias our basic analysis controls for a subset of contemporary controls that have been shown to be important for natural resources (Headey and Hodge 2009; Bloom et al., 2011; Tahir, 2014; Keho, 2017). They include: (i) trade openness or trade which is the sum of exports and imports of goods and services measured as a percentage of gross domestic product, (ii) inflation which here represents the level of inflation per relative to GDP. Inflation, which results in a general rise in prices, can have a positive effect on inclusive growth when it is controlled and regulated by policy makers, (iii) gross fixed capital formation which is the expenditure on assets such as buildings and machinery, either to replace the existing stock of fixed assets or to increase it. It also represents the value of acquisitions of new or existing fixed assets and expressed as a percentage of real GDP (Oluwatobi et al., 2021) because it shows the extent to which new value added in the economy is invested rather than consumed (Dladla and Khobai, 2018).

For Kanu and Ozurumba (2014) it is a component of expenditure linked to gross domestic product (GDP) and therefore indicates the extent to which new value added in the economy is invested rather than consumed. (iv) Population is based on the de facto definition of population, which counts all residents regardless of their legal status or citizenship. Values shown are mid-year estimates, (v) internet, which is measured by the proportion of individuals using the internet. Access to ICT encourages development due to its positive effects on economic growth, productivity and trade (Edquist and Henrekson, 2017). Indeed, there is some consensus in the existing literature that the knowledge economy (ICT) is the key to economic development (Kuada, 2015; Asongu and Le Roux, 2017; Tchamyou, 2017) due to their potential for wider and faster adoption and penetration. (vi) Foreign direct investment (FDI) is measured by FDI inflows in GDP and is used here from the work of Asongu and Odhiambo (2020), Ofori and Asongu (2021) which justify the importance of FDI as an indicator that contributes to increasing productivity growth. Thus, FDI is a wealth-creating determinant which contributes to improving the standard of living and the quality of economic growth.

3.2. Empirical strategies

This study aims to examine the effect of natural resources on productive capacities. Based on the methodological framework of previous work by Zallé (2019) and Naila (2019), we estimate the dynamic panel model specified below:

$$\:{ICP}_{it}=\alpha\:+{\beta\:}_{1}{ICP}_{it-1}+{\beta\:}_{2}{TNR}_{it}+{\beta\:}_{3}{X}_{it}+{\mu\:}_{i}+{\gamma\:}_{t}+{\epsilon\:}_{it}$$

Where $\:{ICP}_{it}\:$represents the productive capacity index and $\:{ICP}_{it-1}$its lagged value.$\:{TNR}_{it}$ represents the total natural resources. $\:{X}_{it\:}$is the vector of control variables. $\:{\mu\:}_{i}\:$is an unobserved country-specific effect and $\:{\gamma\:}_{t}$is a time-specific effect and $\:{\epsilon\:}_{it}\:$is the error term.

We adopt a multi-step econometric approach for this purpose and use several estimation techniques from the literature, namely OLS with fixed effects, Driscoll-Kraay and generalized least squares (GLS) for the basic estimator. The use of these methods makes it possible to resolve a multitude of econometric problems, namely:

Firstly, we apply least squares dummy variables (LSDV) to control the potential problem of heterogeneity linked to the individual specificities present in the panel. Secondly, note that if the fixed effects take into account the heterogeneity specific to each country and make it possible to correct this problem, then given that we are in a panel, there may be a problem of cross-sectional dependencies (See appendices Table A2). However, to correct for this, we need to use the estimation technique of Driscoll and Kraay (1998). Third, the fixed effects estimator and Driscoll-Kraay, however, assume a static relationship between the different variables in the model, whereas this is not always the case. The application of these two estimation techniques does not always take into account a certain unobserved heterogeneity of certain variables and also allows the possibility of endogeneity and autocorrelation linked to the absence of independence between the observations of the variables of our model. To resolve the autocorrelation problem in our model, we use the GLS estimation initially developed by Aitken (1936).

Indeed, the GLS estimation predicts that there is a certain degree of correlation between the residuals and thus allows us to take this into account when estimating the regression model as well as the unknown parameters. This method is more efficient than OLS and weighted least squares because taking into account correlations of residuals in their estimation methods can be statistically inefficient. Fourth, the fixed effects, Driscoll-Kraay and GLS, although they allow us to improve our model, do not take into account the unobserved heterogeneity of certain variables, thus suggesting a presumption of heteroskedasticity and possible endogeneity in our model ( Baum et al, 2003).

Furthermore, due to the specification of Eq. (1), our modeling can thus suffer from Nickel's bias (1981), the use of the GMM estimator therefore becomes crucial for our analysis to control the presence of the effect memory on the delayed dependent variable. Common to the existing literature, we distinguish two GMM approaches, in difference and in system (Roodman, 2009a; Arellano and Bond, 1991; Arellano and Bover, 1995; Blundell and Bond, 1998). Initially developed by the work of Arellano and Bond (1991), the GMM estimator for linear dynamic panel data resolves endogeneity problems. On the one hand, differentiation is accompanied by the use of the level of lag values as instruments, which gives rise to the well-known difference GMM estimator (Arellano and Bond, 1991). Despite its superiority over the OLS estimator, a problem with this difference GMM estimator is that lagged levels have proven to be weak instruments for first differences (Bound et al., 1995). The efficiency can be increased by adding the original equation in levels to the system, which is known as the two-step system-GMM (Arellano and Bover, 1995). We use the two-step GMM system proposed by Arellano and Bover (1995) and Blundell and Bond (1998). Two-step GMM estimators generally have lower bias and standard errors than one-step estimates, which generally leads to higher significance levels. In this regard, some studies preferred to use one-stage estimation due to the downward bias of two-stage standard errors. However, after using the standard error correction proposed by Windmeijer (2005), the two-step estimators account for the downward bias. Therefore, the two-step estimator becomes superior to the one-step estimator in terms of efficiency. We therefore use the two-step GMM estimator after applying the standard error correction proposed by Windmeijer (2005)⁶. Although the GMM system is robust in addressing the endogeneity issues mentioned above, it has a weakness in that it can generate too many instruments leading to over-adjustment of endogenous variables, weakening the Hansen test of joint validity of the instruments and producing biased estimates (Roodman, 2009a). To solve this problem, we include the Collapse option in our GMM matrix in order to limit the proliferation of instruments.

The GMM system estimator is associated with two main tests: the model over-identification test (Hansen test), by which the validity of the instruments used is verified, in the sense that they must be correlated with the variables and not with error terms; and the error autocorrelation test of Arellano and Bond (1991), which tests the first order serial correlation of residuals in level (AR1) and also tests the second order serial correlation of errors in difference (AR2), given that the error terms expressed as first difference are correlated to the first order by the construction of the GMM system estimator. According to this, the validity of the GMM system estimator is based on the quality of the choice of instruments (Hansen test) and on the way in which the second order errors in the difference equation do not autocorrelate (AR2).

4.1. Basic results

Table 2 presents the results of estimating the relationship between productive capacity and natural resources using the LSDV method. We can see that our variable of interest (TNR) and a set of our control variables are overall significant. The coefficient associated with the total natural resource variable in column (1) in relation to the productive capacity reduces the productive capacity by 0.117 and the magnitude indicates that a 1% increase in the natural resource rate results in a decrease of approximately 0.117% of productive capacity. This result allows us to confirm the intuition of Fig. 1 thanks to the elements of column (1) of Table 1. As we add the control variables, the coefficient of the natural resource rate variable decreases from column (2) to column (6) while keeping the expected sign and the significance at the 1% level. Subsequently, the coefficient will increase slightly in column (7) while keeping the expected sign and the significance at the 1% threshold. So, this result reflects the fact that an increase in natural resources leads to a decrease in productive capacities while the productive capacity index reflects the overall transformation of the economy.

This result can be justified by the fact that resource-rich countries develop more slowly than resource-poor countries (Sachs and Warner, 1995). However, the literature identifies economic and political-institutional factors as the two main factors that explain this curse of natural resources. In relation to economic factors, the work of Sachs and Warner (2001), Auty (2001), Ploeg and Poelhekke (2009) identifies the volatility of income from primary factors or the misallocation of production factors as an explanatory factor of Dutch disease.. Low human capital development is a valid reason for the resource curse in countries rich in natural resources (Butkiewicz and Yanikkaya, 2010).

The control variables in our study reveal a significant correlation between trade openness and productive capacity. In column (2), the commercial openness variable has a coefficient of 0.539. It has a positive and significant impact at the 1% threshold on productive capacity. When trade openness increases by 1%, productive capacity increases by 0.539%. As the control variables increase in the model, the coefficient of trade openness keeps the same positive sign from column (1) to (7). The positive sign and significance is confirmed with the work of Keho (2017) for whom trade openness has positive effects on economic growth in the short and long term.

Regarding the Internet, our results show that the Internet contributes positively and significantly to productive capacity. This suggests that a 1% increase in Internet contributes to an increase in productive capacity of 0.176 to 0.172%, respectively. This result can be justified by the fact that ICT enters as an input into the economic supply in the form of capital and leads to improvements in the production process by deepening capital and boosting technological progress in production mechanisms.

Regarding foreign direct investments, it has a negative effect and is significant at the 1% level with a coefficient of 0.608. These results are contradictory with the works of Siddique et al. (2012) who suggest that direct investments have a positive influence on economic growth.

Table 2

Basic estimation by Least Squares Dummy Variables
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	OLS
VARIABLES	PCI index
TNR	-0.117***	-0.113***	-0.110***	-0.106***	-0.0511***	-0.0329***	-0.0371***
	(0.0125)	(0.0120)	(0.0122)	(0.0123)	(0.0101)	(0.00864)	(0.00870)
Trade		0.539***	0.584***	0.566***	0.340***	0.189***	0.220***
		(0.0372)	(0.0382)	(0.0401)	(0.0336)	(0.0293)	(0.0309)
Inflation			-0.102	-0.115	-0.0487	-0.0300	-0.0184
			(0.0705)	(0.0705)	(0.0557)	(0.0461)	(0.0459)
GFCF				0.0256	0.0642***	0.0439***	0.0586***
				(0.0170)	(0.0136)	(0.0113)	(0.0123)
Pop					-0.0245***	-0.0172***	-0.0169***
					(0.00124)	(0.00113)	(0.00112)
Internet						0.176***	0.172***
						(0.0100)	(0.0101)
FDI							-0.608***
							(0.204)
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constant	0.227***	0.193***	0.192***	0.188***	0.243***	0.236***	0.233***
	(0.00675)	(0.00697)	(0.00701)	(0.00750)	(0.00652)	(0.00543)	(0.00554)
Observations	768	679	677	660	660	649	649
R- squared	0.180	0.364	0.381	0.385	0.619	0.743	0.747

Standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1

Source: authors

However, even if the results of the LSDV mentioned above show a negative result between natural resources and productive capacity linked to the fact that it takes into account the heterogeneous character of each country and allows to correct this problem, given that the if we work in a panel, there may also remain a problem of cross-sectional dependence and possibilities of autocorrelation linked to the absence of independence between the observations of the variables in our model. Thus, to correct these different problems, we will respectively use the technique of Driscoll and Kraay (1998) and the GLS developed by Aiken (1936) after having carried out the various appropriate tests (see Appendix Table A2).

Table 3

Controls for the cross-sectional dependence and autocorrelation problem with Driscoll and Kraay and GLS.
	(1)	(2)	(3)	(4)	(5)	(6)
	LSDV		Driscoll Kraay		GLS
VARIABLES	PCI index
TNR	-0.117***	-0.0371***	-0.117***	-0.0371*	-0.0894***	-0.0550***
	(0.0125)	(0.00870)	(0.0199)	(0.0194)	(0.00815)	(0.00676)
Trade		0.220***		0.220***		0.322***
		(0.0309)		(0.0756)		(0.0271)
Inflation		-0.0184		-0.0184		-0.0479**
		(0.0459)		(0.0452)		(0.0233)
GFCF		0.0586***		0.0586***		0.0445***
		(0.0123)		(0.0107)		(0.00854)
Pop		-0.0169***		-0.0169***		-0.0156***
		(0.00112)		(0.00355)		(0.000960)
Internet		0.172***		0.172***		0.153***
		(0.0101)		(0.0162)		(0.00802)
FDI		-0.608***		-0.608		-0.670***
		(0.204)		(0.381)		(0.147)
Constant	0.227***	0.233***	0.227***	0.233***	0.237***	0.224***
	(0.00675)	(0.00554)	(0.00205)	(0.00802)	(0.00137)	(0.00397)
Observations	768	649	768	649	768	649
R- squared	0.180	0.747	0.180	0.747
Number of groups			42	38

Standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1

Source: authors.

Driscoll – Kraay and GLS method estimation presented in Table 3 above shows that natural resources continue to have a negative effect on productive capacity. We note that the coefficients associated with LSDV seem similar to those associated with the technique of Driscoll-Kraay in column (3) and (4) but the standard deviations are more robust with the technique Driscoll-Kraay compared to LSDV and therefore the presence of the strong cross-sectional dependence has been corrected. Thus, an increase of 1% in the natural resource rate leads to a decrease of 0.117% in column (3) and 0.0371% in column (4), respectively, for the method of Driscoll and Kraay, (1998) and 0.0894% in column (5) and 0.0550% in column (6), respectively, for the second method, in particular that of GLS. Our results confirm that natural resources are at the origin of “a curse on productive capacities”. Our results remain consistent with those found by Sachs and Warner (1995); Sachs and Warner (2001); Auty (2001); Easterly and Levine (2003); Bulte et al, (2005); Olsson (2007); The Billion (2008); Van der Ploeg and Poelhekke (2009), Butkiewicz and Yanikkaya (2010), Tsani (2013); Sala-i-Martin and Subramanian (2013), who find that African countries with many natural resources have low economic growth. Furthermore, we find that all our control variables continue to have a significant influence on productive capacity while keeping the same sign.

Furthermore, by virtue of the model specification of Eq. (1), our model may suffer from Nickel's (1981) bias due to the correlation between the dependent lagged variable and the error term, as well as that robust results corrected for the bias of cross-sectional dependence and autocorrelation may be biased due to the fact that they do not take into account the unobserved heterogeneity of certain variables thus suggesting a presumption of heteroskedasticity and possible endogeneity in our model (Baum et al., 2003). So to overcome this problem we apply system GMMs which seem more robust than difference ones (Arellano and Bover, 1995; Blundell and Bond, 1998). The results of the estimation are contained in Table 4 below, it illustrates just as in the previous analyzes a curse of total natural resources on productive capacities, this results in the negative and statistically significant effect of the latter on productive capacities as we add control variables.

Table 4

Control of endogeneity with Sys-GMM
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	SYS-GMM
VARIABLES	PCI index
L.PCI index	0.941***	0.917***	0.913***	0.933***	0.991***	0.934***	0.936***
	(0.00242)	(0.00658)	(0.00722)	(0.00693)	(0.0120)	(0.0340)	(0.0320)
TNR	-0.00787***	-0.0106***	-0.00995***	-0.00506***	-0.0151***	-0.0158**	-0.0166***
	(0.00101)	(0.00177)	(0.00161)	(0.00177)	(0.00546)	(0.00612)	(0.00565)
Trade		0.0825***	0.0909***	0.110***	0.165***	0.176***	0.177***
		(0.00966)	(0.0106)	(0.0102)	(0.0272)	(0.0263)	(0.0256)
Inflation			-0.0457***	-0.0529***	-0.145***	-0.161***	-0.146***
			(0.00955)	(0.0165)	(0.0199)	(0.0246)	(0.0234)
GFCF				-0.0292***	-0.0790***	-0.0846***	-0.0804***
				(0.00264)	(0.00986)	(0.00902)	(0.00974)
Pop					0.0102***	0.00884***	0.00861***
					(0.00124)	(0.00103)	(0.00108)
Internet						0.0116***	0.0114***
						(0.00350)	(0.00321)
FDI							-0.0837
							(0.114)
Constant	0.0169***	0.0167***	0.0173***	0.0174***	-0.0114***	0.00471	0.00421
	(0.000628)	(0.00150)	(0.00166)	(0.00142)	(0.00378)	(0.00756)	(0.00736)
Observations	728	644	642	626	626	615	615
number of groups	42	39	39	38	38	38	38
Instruments	35	35	35	35	35	35	35
AR( 1)	0.000	0.000	0.000	0.000	0.007	0.000	0.000
AR (2)	0.604	0.986	0.944	0.909	0.699	0.370	0.318
Hansen (OIR)	0.296	0.242	0.211	0.199	0.240	0.209	0.200

Standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1

Source: authors

4.2. Robustness test

The first test uses natural resource rent typologies. In the second, we apply an alternative control test for the potential endogeneity problem based on the application of IV-2SLS. Finally, we conclude this section by controlling for the limited nature of our dependent variable.

4.2.1. Curse of productive capacities: evidence on the different types of natural resource rents

The following Table 5 analyzes the results of five typologies of natural resource rents on the productive capacity index. Columns (1) to (4) summarize the results for point resources, while the last column summarizes the results for diffuse resources (column 5). We note that whatever the nature of the natural resources, its sign on the productive capacity index does not change; but however the effect is more amplified for diffuse resources, i.e. all things being equal, an increase of 1% in forestry rent is associated with a decrease of 0.0829% in the level of productive capacity.

Table 5

Curse of productive capacities: evidence on the different types of natural resource rents
	(1)	(2)	(3)	(4)	(5)
	SYS-GMM
VARIABLES	PCI index
L.PCI index	0.924***	0.996***	0.984***	0.843***	0.869***
	(0.0333)	(0.0139)	(0.0115)	(0.0311)	(0.0317)
Oils rents	-0.0127***
	(0.00289)
Mineral rents		-0.0180***
		(0.00235)
Coal rents			-0.00752***
			(0.00119)
Natural gas				-0.0189***
				(0.00266)
Forest rents					-0.0829***
					(0.0179)
Variable control	Yes	Yes	Yes	Yes	Yes
Constant	0.00354	0.000137	0.00156	0.0183**	0.0142*
	(0.00701)	(0.00439)	(0.00367)	(0.00755)	(0.00740)
Observations	615	615	615	615	615
Group number	38	38	38	38	38
Instruments	37	33	32	34	35
AR( 1)	0.000	0.000	0.000	0.000	0.000
AR( 2)	0.232	0.592	0.653	0.298	0.717
Hansen (OIR)	0.247	0.296	0.336	0.202	0.319

Standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1

Source: authors

4.2.2. Alternative control of endogeneity: IV-2SLS

The question of endogeneity is treated differently in the econometric literature.

When the data suffer from heteroskedasticity, the GMM estimator is recommended,

except that the latter induces additional constraints on the proportionality between the number

of temporal observations (J) and the number of individuals or countries (N). Although this condition of low temporal observation in relation to the number of individuals or countries (J less than N) is respected, a methodological dilemma arises in instrumental variable techniques on the choice of instruments. There are two competing approaches to instrumentation, namely internal and external instrumentation. Due to the theoretical difficulties encountered in the choice of a purely external instrument, we adopt the internal instrumentation approach (Baum et al., 2006; Lewbel, 2012). The internal instrumentation of this relationship is carried out through the delay of the suspected endogenous variable in the case we use the delay of natural resource rents. The results obtained remain stable and robust regardless of the natural resource rent used (Table 6). They once again confirm the strength of the relationship between the different natural resource rents and the productive capacity index. The diagnostic tests show that the selected instruments are valid, more specifically the statistics developed by Stock and Yogo (2005) and Cragg and Donald (1993) illustrate that our instruments although they are internal are relevant. This is also confirmed through the test of Kleibergen and Paap (2006) and the statistic of Hansen, (1982) and Sargan, (1983) concludes their validity on the basis of the p-value all above the critical threshold of 10%.

Table 6

Alternative control of endogeneity with IV-2SLS
	(1)	(2)	(3)	(4)	(5)	(6)
	IV-2SLS
VARIABLES	PCI index
TNR	-0.0350*
	(0.0185)
Oils rents		-0.0514***
		(0.0174)
Mineral rents			-0.0856***
			(0.0284)
Forest rents				-0.0764***
				(0.0246)
Coal rents					0.0221
					(0.0208)
Natural gas						-0.0611***
						(0.0139)
Trade	0.227***	0.235***	0.210***	0.184***	0.208***	0.359***
	(0.0560)	(0.0626)	(0.0517)	(0.0515)	(0.0523)	(0.0513)
Inflation	0.0453	-0.00954	0.00847	0.000716	-0.00388	-0.0899*
	(0.0858)	(0.0668)	(0.0791)	(0.0802)	(0.0764)	(0.0497)
GFCF	0.0633***	0.0702***	0.0608***	0.0457***	0.0551***	0.0670***
	(0.0143)	(0.0140)	(0.0142)	(0.0136)	(0.0141)	(0.0159)
Pop	-0.0185***	-0.0174***	-0.0194***	-0.0185***	-0.0191***	-0.0137***
	(0.00267)	(0.00286)	(0.00276)	(0.00272)	(0.00280)	(0.00305)
Internet	0.139***	0.140***	0.143***	0.140***	0.146***	0.129***
	(0.00798)	(0.00889)	(0.00964)	(0.00946)	(0.00947)	(0.00846)
FDI	-0.576**	-0.646**	-0.429*	-0.296	-0.484**	-0.611***
	(0.246)	(0.272)	(0.242)	(0.240)	(0.244)	(0.157)
Country fixed effects	No	Yes	No	No	No	Yes
Constant	0.238***	0.233***	0.238***	0.244***	0.237***	0.225***
	(0.00809)	(0.00956)	(0.00887)	(0.00865)	(0.00906)	(0.00878)
Observations	581	581	583	583	581	581
R- squared	0.734	0.759	0.729	0.732	0.728	0.837
10% maximum IV size	19.93	19.93	19.93	19.93	19.93	19.93
15% maximum IV size	11.59	11.59	11.59	11.59	11.59	11.59
20% maximum IV size	8.75	8.75	8.75	8.75	8.75	8.75
Hansen J p-value test	0.707	0.830	0.981	0.742	0.782	0.691
Kleibergen -Paap rk p-value	0.000	0.000	0.000	0.000	0.000	0.000
Kleibergen -Paap rk Wald	51.35	105.5	30.86	66.52	33.39	32.49

Standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1

Source: authors

4.2.3. Control over the bias of the limited dependent variable

As the dependent variable is bounded in the interval [0-100], our results may be biased by OLS or other related techniques. Indeed, OLS are also inappropriate with limited dependent variables for which there is a large number of varieties: this is the case for our productive capacity index variable, which varies and is defined in the interval 0 to 100. Sometimes, a variable dependent can be continuous over one or more intervals, but take one or more values with finite probability. Limited dependent variable models are designed to deal with samples that are truncated or censored. To remedy this bias, certain estimators are appropriate. In this article, we use respectively the Tobit estimator and the IV-Tobit of Newey (1987) for censoring and truncation control while correcting the potential endogeneity problem that may exist in the relationship. To do this we apply the two-step IV-Tobit approach using the delay of the endogenous variable as an instrument. These models are referred to as counting models because they count occurrences of an event. More specifically, they take into account censorship and truncation issues. On the one hand, a sample is truncated if some of its observations which should have been there have been systematically excluded. On the other hand, a sample is said to be censored if no observation has been systematically excluded, but if certain information contained by these observations has been removed. These two explanations could be the case for the extreme values (0 and 100) of the productive capacity index. The results obtained in Table 7 remain consistent with previous results: the rent from natural resources reduces the productive capacity index. However, the effect seems more important when we control for endogeneity in the counting models (see column 3 of Table 7 where the associated coefficient is 0.0831% at the 1% threshold).

Table 7

Control over the bias of the limited dependent variable
	(1)	(2)	(3)
	OLS	TOBIT	IV-TOBIT two step
VARIABLES	PCI index
TNR	-0.0371**	-0.0372***	-0.0831***
	(0.0157)	(0.00854)	(0.0134)
Trade	0.220***	0.219***	0.151***
	(0.0516)	(0.0303)	(0.0333)
Inflation	-0.0184	-0.0182	0.0755
	(0.0448)	(0.0451)	(0.112)
GFCF	0.0586***	0.0586***	0.0621***
	(0.0133)	(0.0120)	(0.0142)
Pop	-0.0169***	-0.0170***	-0.0167***
	(0.00242)	(0.00110)	(0.00132)
Internet	0.172***	0.172***	0.132***
	(0.0115)	(0.00987)	(0.00865)
FDI	-0.608***	-0.605***	-0.351
	(0.230)	(0.200)	(0.226)
Constant	0.233***	0.233***	0.244***
	(0.00799)	(0.00543)	(0.00569)
Observations	649	649	348
R- squared	0.747
Log likelihood		-1513
P-value		0.000	0.000
Exogeneity statistic			3,800
Exogeneity test p-value			0.0512

Standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1

Source: authors

4.3. Sensitivity test

Based on the great debate in the existing literature on the scope of natural resources in economic growth, this section aims to test the sensitivity of our results. More specifically, it consists firstly of examining whether the resource curse is a reality for all the sub-regions of our sample, and secondly it allows us to revisit the divisions in the theoretical literature on the persistence of a relationship non-linear.

4.3.1. Regional heterogeneity

Although the relationship between natural resources and the productive capacity index is negative in the overall sample of African countries, there may be asymmetric differences in sign or magnitude between the main sub-regions of the continent. For the sake of freedom and sufficient parsimony of the different specifications, we retain five large regions, namely East Africa, Central Africa, North Africa, Southern Africa and South Africa. 'West. The results obtained in Table 8 reveal a type of asymmetries tested, namely an amplitude asymmetry. More specifically, the curse on productive capacities is more visible in the Southern sub-region Africa. In fact, in this sub-region, countries like Angola experienced high levels of natural resources almost above the average over the period of our sample but unfortunately, their level of productive capacity was well below the average over our period of study (See Fig. 1). In addition to this sub-region, the curse is also evident in dominant economies such as: Western Africa, Eastern Africa and Northern Africa.

Table 8

Regional heterogeneity
	(1)	(2)	(3)	(4)	(5)
	Eastern Africa	Middle Africa	Northern Africa	Southern Africa	Western Africa
VARIABLES	PCI index
TNR	-0.100***	-0.0198	-0.0425***	-0.311**	-0.109***
	(0.0298)	(0.0371)	(0.00903)	(0.154)	(0.0229)
Trade	0.0675	-0.0678	0.799***	0.370***	0.447***
	(0.0715)	(0.191)	(0.150)	(0.0847)	(0.0603)
Inflation	-0.0891	-0.0826***	0.101	-0.277	0.349**
	(0.139)	(0.0166)	(0.179)	(0.272)	(0.138)
GFCF	0.116***	0.0172	-0.0293	0.0357	0.0262
	(0.0237)	(0.0269)	(0.0228)	(0.0363)	(0.0219)
Pop	-0.0134***	0.00778	-0.000888	-0.00285	-0.00720**
	(0.00333)	(0.00483)	(0.00138)	(0.00454)	(0.00330)
Internet	0.181***	0.0718***	0.0737***	0.0616***	0.137***
	(0.0187)	(0.00885)	(0.00836)	(0.00925)	(0.0117)
FDI	-0.624**	-0.227	0.902	-0.583	-0.203
	(0.313)	(0.234)	(0.825)	(0.647)	(0.267)
Constant	0.233***	0.182***	0.214***	0.300***	0.199***
	(0.00918)	(0.0163)	(0.0152)	(0.00986)	(0.00941)
Observations	177	79	64	63	198
R- squared	0.830	0.957	0.872	0.883	0.589
10% maximum IV size	19.93	19.93	19.93	19.93	19.93
15% maximum IV size	11.59	11.59	11.59	11.59	11.59
20% maximum IV size	8.75	8.75	8.75	8.75	8.75
Hansen J p-value test	0.289	0.946	0.326	0.203	0.478
Kleibergen -Paap rk LM p-value	0.000	0.002	0.000	0.006	0.000
Kleibergen -Paap rk Wald	42.79	11.87	19.94	10.11	51.23

Standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1

Source: authors

4.3.2. Checking the quadratic effect

This section consists of verifying the possible existence of a quadratic effect declining the non-linear relationship that may exist between the total rent of natural resources and the productive capacity index. This verification involves two tests, firstly we carry out an empirical verification of the behavior of the relationship as illustrated in Fig. 2 below; secondly we test the intuition present in this figure. To do this in Table 9 below we estimate in accordance with the work of (Nchofoung et al., 2022; Ngounou et al, 2023) the specification of the following model:

$$\:{ICP}_{it}=\alpha\:+{\beta\:}_{1}{ICP}_{it-1}+{\beta\:}_{2}{TNR}_{it}+{\beta\:}_{3}{{TNR}^{2}}_{it}+{\beta\:}_{3}{X}_{it}+{\mu\:}_{i}+{\gamma\:}_{t}+{\epsilon\:}_{it}$$

Where $\:{TNR}^{2}$is the total natural resources squared, thus making it possible to highlight the quadratic form, the shape of this specification is based respectively on the sign and the significance of the coefficients ( ) $\:{\beta\:}_{2}et\:{\beta\:}_{3}\:$. The highlighting of this specification is recorded in Table 9, the coefficients associated with these specifications are respectively negative and positive, significant in most cases when we control the addition of the control variables commonly to the result recorded in Table 3. Thus, our results show the existence of a U-shaped curve because $\:{\beta\:}_{2}<0$ and $\:{\beta\:}_{3}>0$. So, without wishing to be exhaustive, there is a threshold beyond which the curse of natural resources becomes a myth and not a reality. In accordance with the work of ( Abdulahi et al., 2019; Kengdo et al., 2020; Dramani et al., 2022, Nchofoung et al., 2022) on the basis of Eq. 2 the determination of this threshold is done in fact as follows :

$$\:\:Seuil\left(\%\right)=-\frac{{\beta\:}_{2}}{2{\beta\:}_{3}}*100\:\:\:\:\:\:$$

The results recorded in Table 9 illustrate a quadratic relationship between the rent of natural resources and the productive capacity index. More specifically, these results demonstrate the existence of a U-shaped effect between natural resources and productive capacities for the sample of African countries selected (see Fig. 3 above), this is reflected in Table 9 by the signs respectively negative and positive of the total rent of natural resources and its high squared value. In accordance with this result, we can deduce that there is a threshold effect from which the resource curse becomes a myth. Thus, the analysis of the basic relationship in column (1) of Table 9 shows us a significant presumption of the threshold at 34.14%, as we add the control variables the threshold tends to vary up to 43.85% with significance at the 1% level. Ultimately, in accordance with Fig. 2 we can see that the threshold from which natural resources improve productive capacities belongs to the interval [33.90–43.85] with significance at the 1% threshold. Our results thus reject the existence of the inverted U hypothesis of Kuznets (1967) and are in line with those of Abdulahi et al. (2019); Kengdo et al. (2020) and Dramani et al. (2022).

Table 9

Analysis of the threshold effect
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	SYS-GMM
VARIABLES	PCI index
L.PCI index	0.940***	0.918***	0.912***	0.932***	0.992***	0.920***	0.923***
	(0.00375)	(0.00670)	(0.00719)	(0.00780)	(0.0126)	(0.0348)	(0.0333)
TNR ($\:{\beta\:}_{2})$	-0.0226***	-0.0204***	-0.0301***	-0.0109	-0.0200**	-0.0360***	-0.0371***
	(0.00433)	(0.00397)	(0.00452)	(0.00829)	(0.00766)	(0.00839)	(0.00969)
TNR ² ($\:{\beta\:}_{3})$	0.0331***	0.0211**	0.0444***	0.0120	0.0154	0.0424***	0.0423**
	(0.00808)	(0.00808)	(0.00956)	(0.0175)	(0.0208)	(0.0153)	(0.0198)
Trade		0.0784***	0.0814***	0.105***	0.144***	0.182***	0.180***
		(0.0112)	(0.0126)	(0.0102)	(0.0460)	(0.0300)	(0.0314)
Inflation			-0.0746***	-0.0560**	-0.157***	-0.179***	-0.161***
			(0.0228)	(0.0223)	(0.0266)	(0.0263)	(0.0282)
GFCF				-0.0270***	-0.0752***	-0.0800***	-0.0745***
				(0.00308)	(0.0116)	(0.00867)	(0.00955)
Population					0.00953***	0.00893***	0.00859***
					(0.00180)	(0.00100)	(0.00106)
Internet						0.0124***	0.0121***
						(0.00334)	(0.00310)
FDI							-0.105
							(0.129)
Thresholds (%)	34.14***	48.34**	33.90***	45.42	64.93	42.45***	43.85***
Constant	0.0178***	0.0174***	0.0197***	0.0179***	-0.00939*	0.00755	0.00696
	(0.00111)	(0.00171)	(0.00215)	(0.00236)	(0.00509)	(0.00774)	(0.00780)
Observations	728	644	642	626	626	615	615
Number of groups	42	39	39	38	38	38	38
Instruments	35	35	35	35	35	35	35
AR (1)	0.000	0.000	0.000	0.000	0.007	0.001	0.001
AR (2)	0.617	0.987	0.994	0.928	0.714	0.375	0.308
Hansen (OIR)	0.246	0.206	0.188	0.175	0.211	0.186	0.165

Standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1

Source: authors

Regardless of the immense wealth that natural resources can bring to a country, it is regularly observed that resource-rich countries have worse social, political and economic outcomes than their resource-poor counterparts. This trend has been observed so often that it has been called the “resource curse” or “paradox of abundance.” Using data from 42 African countries during the period 2000-2018 and the Generalized Moments method in a two-step system, this article examines the evidence of a curse on an original indicator of perception of the level of development, notably productive capacities. The results obtained show that the total rent of natural resources has a negative effect on the index of productive capacities. However, a sensitivity analysis distinguishing point resources from diffuse resources demonstrates that the magnitude of the effect is very pronounced on diffuse resources. Furthermore on the basis of controlling for regional heterogeneity; we arrive at empirical evidence of the paradox of abundance in the different regions selected but more specifically in the Southern Africa. Furthermore, our results remain robust when we control the limited nature of our dependent variable, in particular by applying the Tobit estimator and the Tobit estimator of the two-stage instrument variables (IV-Tobit). The results of the two-stage instrumental variables (IV-2SLS) are not left out because, just like those of the system GMMs, the latter are consistent with the basic results. Without wishing to claim to be exhaustive, our results also highlight evidence of the existence of a U-shaped relationship between the total rent of natural resources and the index of productive capacities.

We draw several conclusions from this study. Given the importance of improving the level of productive capacity and accelerating economic growth, we encourage resource-rich governments to allocate a relatively large share of resource revenues to finance the tools that constitute the ' exponentialization of productive capacities. Upstream, particular emphasis must be placed on human capital. These additional revenues would make it possible to strengthen the private sector, promote the strengthening of energy policies, boost the attractiveness of multinationals, which will facilitate the creation of digital infrastructures (acceleration of ICT access kits), while strengthening the economic and social fabric either through travel modes or the increase in educational infrastructures. Downstream, governments should put in place institutional mechanisms to ensure effective management of revenues derived from natural resources and allocated to productive capacities. This involves strengthening the institutional sphere that guarantees political stability.

Abdulahi , M.E., Shu, Y., & Khan, M.A. (2019). Resource rents , economics growth , and the role of institutional quality : A panel threshold analysis . Resources Policy , 61 , 293-303.
Adekunle, I. A. (2021). On the search for environmental sustainability in Africa: the role of governance. Environmental Science and Pollution Research , 28 (12), 14607-14620.
Adekunle, IA, Maku , O., Williams, T., Gbagidi , J., & Ajike , EO (2023). Natural resource endowments and growth dynamics in Africa: evidence from panel cointegrating regression. Management of Environmental Quality : An International Journal .
Ades , A., & Di Tella , R. (1999). Rents , competition , and corruption. American economics review , 89 (4), 982-993.
Ahmad, N., & Yadav , D.K. (2013). Inclusive Growth and Food Inflation: An Analysis of their Nullification Effect . Available at SSRN 2739349 .
Ahmed, K., Mahalik , M.K., & Shahbaz, M. (2016). Dynamics between economic growth, labor, capital and natural resource abundance in Iran: An application of the combined cointegration approach. Resources Policy , 49 , 213-221.
Aitken, A. C. (1936). IV. On least squares and linear combination of observations. Proceedings of the Royal Society of Edinburgh, 55, 42–48.
Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The review of economic studies , 58 (2), 277-297.
Arellano, M., & Bover , O. (1995). Another look at the instrumental variable estimation of error-components models. Journal of econometrics , 68 (1), 29-51.
Arezki, R., & van der Ploeg , R. (2007). Can the natural resource curse be turned into a blessing? The role of trade policies and institutions.
Arin, K.P., & Braunfels, E. (2018). The resource curse revisited: A Bayesian model averaging approach. Energy Economics , 70 , 170-178.
Asif , M., Khan, KB, Anser, MK, Nassani , AA, Abro , MMQ, & Zaman, K. (2020). Dynamic interaction between finance development and natural resources : Evaluating the 'Resource curse' hypothesis . Resources Policy , 65 , 101566.
Asongu , SA, & Le Roux, S. (2017). Enhancing ICT for inclusive human development in sub-Saharan Africa. Technological Forecasting and Social Change , 118, 44–54.
Asongu , S.A., & Odhiambo , N.M. (2020). Foreign direct investment , information technology and economics growth dynamics in sub-Saharan Africa . Telecommunications Policy , 44(1), 101838.
Auty, R. M. (1997). Natural resource endowment, the state and development strategy. Journal of International Development: The Journal of the Development Studies Association , 9 (4), 651-663.
Auty , R. M. (2000). How natural resources affect economic development. Development Policy Review , 18 (4), 347-364.
Auty, R.M. (Ed.). (2001). Resource abundance and economics development . Oxford university press .
Auty , RM (Ed.). (2001). Resource abundance and economic development . Oxford University Press.
Awoa , PA, Ondoa , HA, & Tabi, HN (2022). Women's political empowerment and natural resource curse in developing countries. Resources Policy , 75 , 102442.
Baldursson , FM, & von der Fehr, NHM (2018). Natural resources and sovereign expropriation. Journal of Environmental Economics and Management , 92 , 580-607.
Barro, R. J. (1999). Determinants of democracy . Journal of Politics economics , 107 (S6), S158-S183.
Baum, CF 2006. An Introduction to Modern Econometrics Using Stata. College Station, TX: Stata Press .
Baum, C.F., Schaffer, M.E., & Stillman, S. (2003). Instrumental variables and GMM: Estimation and testing. The Stata Journal , 3 (1), 1-31.
Belarbi , Y., Hamdi, F., Khalfi , A., & Souam , S. (2021). Growth, institutions and oil dependence: a buffered threshold panel approach. Economic Modeling , 99 , 105477.
Bloom, DE, Canning, D., & Rosenberg, L. (2011, February). Demographic change and economic growth in South Asia. St. Gallen, Switzerland: WDA-Forum, University of St. Gallen.
Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models . Journal of Econometrics , 87(1), 115–143.
Boschini , A.D., Pettersson , J., & Roine , J. (2007). Resource curse or not: A question of appropriability . Scandinavian Journal of Economics , 109 (3), 593-617.
Bound, J., Jaeger, DA, Baker, RM, 1995. Problems with instrumental variables estimation when the correlation between the instruments and the endogenous explanatory variable
Boyce, J.R., & Emery, J.H. (2005). A hotel explanation of “the curse of nature resources ”. University of Calgary, Department of Economics Discussion Paper , 6 .
Brunnschweiler , C.N. (2008). Cursing the blessings? Natural resource abundance , institutions, and economics growth . World development , 36 (3), 399-419.
Brunnschweiler , C. N., & Bulte , E. H. (2008). Linking natural resources to slow growth and more conflict . Science , 320 (5876), 616-617.
Brunnschweiler , C. N., & Bulte , E. H. (2008). The resource curse revisited and revised: A tale of paradoxes and red herrings. Journal of environmental economics and management , 55 (3), 248-264.
Bulte , E.H., Damania , R., & Deacon, R.T. (2005). Resource intensity , institutions, and development . World development , 33 (7), 1029-1044.
Butkiewicz , J.L., & Yanikkaya , H. (2010). Minerals , institutions, openness , and growth : an empirical analysis . Land Economics , 86 (2), 313-328.
Chirwa , T.G., & Odhiambo , N.M. (2016). Macroeconomics determinants of economics growth : A review of international literature . The South East European Journal of Economics and Business , 11 (2).
Cockx , L., & Francken , N. (2016). Natural resources: a curse on education spending?. Energy Policy , 92 , 394-408.
Collier, P., & Hoeffler , A. (2005). Resource rents , governance , and conflict . Journal of conflict resolution , 49 (4), 625-633.
Djimeu , EW, & Omgba , LD (2019). Oil windfalls and export diversification in oil-producing countries: Evidence from oil booms. Energy Economics , 78 , 494-507.
Dladla , K., & Khobai , H. (2018). The impact of taxation on economic growth in South Africa.
Dorinet , E., Jouvet , PA, & Wolfersberger, J. (2021). Is the agricultural sector cursed too? Evidence from Sub-Saharan Africa. World Development , 140 , 105250.
Dramani , J.B., Abdul Rahman, Y., Sulemana , M., & Owusu Takyi , P. (2022). Natural resource dependence and economics growth in SSA: are there threshold effects ?. Development Studies Research , 9 (1), 230-245.
Driscoll , J. C., & Kraay , A. C. (1998). Consistent covariance matrix estimation with spatially dependent panel data. Review of economics and statistics , 80 (4), 549-560.
Easterly , W., & Levine, R. (2003). Tropics , germs , and crops : how endowments influence economics development . Journal of monetary economics , 50 (1), 3-39.
Edquist , H., & Henrekson , M. (2017). Swedish lessons : How important are ICT and R&D to economic growth ? Structural Change and Economic Dynamics , 42 , 1-12.
Ekodo , R., & Ndam , M. (2019). Possession of natural resources and economic growth in the CEMAC zone: case of oil resources. Economic Benchmarks and Perspectives , 3 (1).
Erum , N., & Hussain, S. (2019). Corruption, natural resources and economics growth : Evidence from OIC countries. Resources Policy , 63 , 101429.
Frankel , J.A., & Romer, D. (2017). Does trade causes growth ? In Global Trade (pp. 255-276). Routledge.
Freund, C., & Bolaky , B. (2008). Trade, regulations , and income . Journal of Development Economics , 87, 309–321.
Gnangnon , S.K. (2021). Effect of productive capacities on economic complexity. Journal of Economic Integration , 36 (4), 626-688.
Gu, X., Badeeb , RA, Ali, S., Khan, Z., Zhang, C., & Uktamov , KF (2023). Nonlinear impact of natural resources and risk factors on the US economic growth. Resources Policy , 82 , 103570.
Gylfason , T., & Zoega , G. (2006). Natural resources and economics growth : The role of investment . World Economy , 29 (8), 1091-1115.
Gylfason , T., Herbertsson , T.T., & Zoega , G. (1999). A mixed blessing: natural resources and economics growth . Macroeconomics dynamics , 3 (2), 204-225.
Hansen, L.P. (1982). Large sample properties of generalized method of moments estimators. Econometrica : Journal of the econometric society , 1029-1054.
Hartwell, C., Horvath, R., Horvathova , E., & Popova, O. (2022). Natural resources and income inequality in developed countries: synthetic control method evidence. Empirical Economics , 1-42.
Headey , D.D., & Hodge , A. (2009). The effect of population growth on economics growth : A meta-regression analysis of the macroeconomic literature . Population and development review , 35 (2), 221-248.
Heckscher, E.F. (1919). The effect of foreign trade on the distribution of income . Readings in the theory of international trade . Blakiston Saries of Republished Artides on Economics , (4), 386.
Henri, PAO (2019). Natural resources curse : A reality in Africa . Resources policy , 63 , 101406.
Herb , M. (2005). No representation without taxation? Rents , development , and democracy . Comparative Politics , 297-316.
Horváth, R., & Zeynalov , A. (2014). The natural resource curse in post-soviet countries: The role of institutions and trade policies (No. 341). IOS Working Papers.
Hye , Qazi Muhammad Adnan. 2012. Long-term effect of trade openness on economic growth in the case of Pakistan. Quality and Quantity 46:1137–49.
Isham , J., Woolcock, M., Pritchett , L., & Busby , G. (2005). The varieties of resources experience : natural resource export structures and the politics economics of economics growth . The World Bank Economic Review , 19 (2), 141-174.
Isham , J., Woolcock, M., Pritchett, L., & Busby, G. (2005). The varieties of resource experience: natural resource export structures and the political economy of economic growth. The World Bank Economic Review , 19 (2), 141-174.
Jawaid , Syed Tehseen . 2014. Trade openness and economic growth: a lesson from Pakistan. Foreign Trade Review 49: 193–212.
Jensen, N., & Wantchekon , L. (2004). Resource wealth and politics regimes in Africa . Comparative politics studies , 37 (7), 816-841.
John, J. (2011). Is there really a resource curse? A critical survey of theory and evidence. Global governance: A review of multilateralism and international organizations , 17 (2), 167-184.
Kamguia , B., Keneck-Massil , J., Nvuh-Njoya , Y., & Tadadjeu , S. (2022). Natural resources and innovation: Is the R&D sector cursed too ?. Resources Policy , 77 , 102725.
Kanu, SI, & Ozurumba , BA (2014). Capital formation and economic growth in Nigeria. Global Journal of human -social science: Economics , 14 (4), 43-58.
Karl, T. L. (1997). The paradox of plenty: Oil booms and petro -states (Vol. 26). University of California Press.
Keho , Y. (2017). The impact of trade openness on economic growth: The case of Cote d’Ivoire. Cogent Economics & Finance , 5 (1), 1332820.
Kengdo , AAN, Nchofoung , T., & Ntang , P.B. (2020). Effect of external debt on the level of infrastructure in Africa . Economics Bulletin , 40 (4), 3349-3366.
Khan, Z., Hussain, M., Shahbaz , M., Yang, S., & Jiao , Z. (2020). Natural resource abundance , technological innovation, and human capital nexus with finance development : a case study of China. Resources Policy , 65 , 101585.
Kim, D.H., Chen, T.C., & Lin, S.C. (2020). Does oil drive income inequality? New panel evidence. Structural Change and Economic Dynamics , 55 , 137-152.
Kleibergen , F., & Paap , R. (2006). Generalized reduced rank tests using the singular value decomposition. Journal of econometrics , 133 (1), 97-126.
Kuada , J. (2015). Private enterprise-led economic development in Sub-Saharan Africa : the human side of growth . Springer.
Kuznets, S. (1967). Population and economics growth . Proceedings of the American Philosophical Society , 111 (3), 170-193.
Lalji , N. (2007). The resource curse revised : Conflict and coltan in the Congo. Harvard International Review , 29 (3), 34-38.
Lashitew , A.A., & Werker , E. (2020). Do natural resources help or hinder development? Resource abundance, dependence, and the role of institutions. Resource and Energy Economics , 61 , 101183.
Le Billon, P. (2008). Diamond wars ? Conflict diamonds and geographies of resource wars . Annals of the Association of American Geographers , 98 (2), 345-372.
Lederman , D., & Maloney , W. F. (2003). R&D and development . Available at SSRN 402480 .
Leite , M. C., & Weidmann , J. (1999). Does mother nature corrupt? Natural resources, corruption, and economic growth . International Monetary Fund.
Lewbel, A. (2012). Using heteroscedasticity to identify and estimate mismeasured and endogenous regressor models . Journal of Business & Economics Statistics , 30(1), 67–80.
Liu, H., Alharthi , M., Atil , A., Zafar, MW, & Khan, I. (2022). A non-linear analysis of the impacts of natural resources and education on environmental quality: Green energy and its role in the future. Resources Policy , 79 , 102940.
Marelli , E., & Signorelli, M. (2011). China and India : Openness , trade and effects on economics growth . The European Journal of comparative economics , 8 (1), 129.
Mavrotas , G., Murshed , S.M., & Torres, S. (2011). Natural resource dependence and economic performance in the 1970–2000 period . Review of Development Economics , 15 (1), 124-138.
Mehlum , H., Moene , K., & Torvik , R. (2006). Cursed by resources or institutions?. World Economy , 29 (8), 1117-1131.
Mehlum , H., Moene , K., & Torvik , R. (2006). Institutions and the resource curse . The economic journal , 116 (508), 1-20.
Meyer, D.F., & Sanusi, K.A. (2019). A causality analysis of the relationships between big fixed capital formation, economic growth and employment in South Africa . Studia Universitatis Babes -Bolyai Oeconomica , 64 (1), 33-44.
Ncanywa , T., & Makhenyane , L. (2016). Can investment activities in the form of Capital Formation influence economic growth in South Africa ?
Nchofoung , TN, Achuo , ED, & Asongu , SA (2021). Resource rents and inclusive human development in developing countries. Resources Policy , 74 , 102382.
Nchofoung , TN, Asongu , SA, Njamen Kengdo , A.A., & Achuo , E.D. (2022). Linear and non‐ linear effects of infrastructures on inclusive human development in Africa . African Development Review , 34 (1), 81-96.
Newey , W. K. (1987). Specification tests for distributional assumptions in the Tobit model. Journal of Econometrics , 34 (1-2), 125-145
Ngounou , BA, Tekam Oumbe , H., Ongo Nkoa , BE, & Noubissi Domguia , E (2023). Inclusive growth in the face of increasing urbanization : What experience for African countries?. Review of Development Economics .
Njangang , H., Asongu , SA, Tadadjeu , S., Nounamo , Y., & Kamguia , B. (2022). Governance in mitigating the effect of oil wealth on wealth inequality: a cross-country analysis of policy thresholds. Resources Policy , 76 , 102561.
Ofori , I.K., & Asongu , S.A. (2021). ICT diffusion, foreign direct investment and inclusive growth in sub-Saharan Africa . Telematics and Informatics , 65, 101718.
Ohlin, B. (1933). Interregional and international trade Harvard University Press Cambridge Massachusetts USA.
Olsson, O. (2007). Conflict diamonds . Journal of Development Economics , 82 (2), 267-286.
Oluwatobi , AA, Festus, FA, & Grace, OO (2021). Tax revenue, capital formation, and economic growth in Nigeria. Research in World Economy , 1 (12), 101-112.
Ongo , EN, & Vukekeng , AW (2014). Does gross capital formation matter for economic growth in the CEMAC sub- region ?. EuroEconomica , 33 (2).
Peach , J., & Starbuck , C. M. (2011). Oil and gas production and economics growth in New Mexico. Journal of Economic Issues , 45 (2), 511-526.
Posner, M. V. (1961). International trade and technical change. Oxford Economics papers , 13 (3), 323-341.
Rahim, S., Murshed , M., Umarbeyli , S., Kirikkaleli , D., Ahmad, M., Tufail, M., & Wahab, S. (2021). Do natural resources abundance and human capital development promote economic growth? A study on the resource curse hypothesis in Next Eleven countries. Resources, Environment and Sustainability , 4 , 100018.
Ricardo, D. 1817. Principles of Political Economy and Taxation, London.
Rigobon , Roberto, and Dani Rodrik . 2005. Rule of law , democracy , openness , and income : Estimating the interrelationships . Economics of Transition 13: 533–64.
Robinson, JA, Torvik , R., & Verdier, T. (2006). Politics foundations of the resource curse . Journal of development Economics , 79 (2), 447-468.
Roodman , D. (2009a). How to do xtabond2: An introduction to difference and system GMM in Stata. The stata journal , 9 (1), 86-136.
Ross, M. L. (2001). Does oil hinder democracy ? World politics , 53 (3), 325-361.
Sachs, J.D., & Warner, A. (1995). Natural resource abundance and economic growth .
Sachs, J.D., & Warner, A. (1995). Natural resource abundance and economic growth.
Sachs, J.D., & Warner, A.M. (1999). The big push, natural resource booms and growth . Journal of development economics , 59 (1), 43-76.
Sachs, J.D., & Warner, A.M. (2001). The curse of nature resources . European economic review , 45 (4-6), 827-838.
Sala- i -Martin, X., & Subramanian, A. (2013). Addressing the natural resource curse: An illustration from Nigeria. Journal of African Economies , 22 (4), 570-615.
Samuelson, P. A. (1953). Prices of factors and goods in general equilibrium . The Review of Economics Studies , 21 (1), 1-20.
Sargan , J.D. (1983). Identification and lack of identification. Econometrica : Journal of the Econometric Society , 1605-1633.
Shahbaz , M., & Rahman, M. M. (2012). The dynamic of finance development , imports, foreign direct investment and economic growth : cointegration and causality analysis in Pakistan. Global Business Review , 13 (2), 201-219.
Shahbaz, M., Destek , M.A., Okumus , I., & Sinha, A. (2019). An empirical note on comparison between resource abundance and resource dependence in resource abundant countries. Resources Policy , 60 , 47-55.
Siddique , A., Selvanathan , E.A., & Selvanathan , S. (2012). Remittances and economics growth : empirical evidence from Bangladesh, India and Sri Lanka. Journal of development studies , 48 (8), 1045-1062.
Silvestre, BS, & Ţîrcă , DM (2019). Innovations for sustainable development: Moving toward a sustainable future. Journal of cleaner production , 208 , 325-332.
Singh, B.P., & Pradhan , K.C. (2022). Institutional quality and economic performance in South Asia. Journal of Public Affairs , 22 (1), e2401.
Sini, S., Abdul-Rahim, A.S., & Sulaiman , C. (2021). Does natural resource influence conflict in Africa? Evidence from panel nonlinear relationship. Resources Policy , 74 , 102268.
Siraj, M., Zhang, K., Xiao, W., Bilal, A., Gemechu , S., Geda , K., & Xiaodan , L. (2018). Does participatory forest management save the remnant forest in Ethiopia?. Proceedings of the National Academy of Sciences, India Section B: Biological Sciences , 88 , 1-14.
Smith, A. 1776. Research on the nature and causes of the wealth of nations, Paris,
Guillaumin, 3, ed fr1859 p.
Tadadjeu , S., Njangang , H., Ningaye , P., & Nourou , M. (2022). Oil dependence and access to water and sanitation in African countries: Does the Extractive Industries Transparency Initiative matter?. African Development Review , 34 (1), 54-67.
Tahir, M., Haji, DHNBP, & Ali, O. (2014). Trade openness and economic growth: A review of the literature. Asian Social Science , 10 (9), 137.
Tchamyou , VS (2017). The role of knowledge economy in African business . Journal of the Knowledge Economy , 8(4), 1189–1228.
Tsani , S. (2013). Natural resources , governance and institutions quality : The role of resource funds . Resources Policy , 38 (2), 181-195.
Ulasan , B. (2015). Trade openness and economics growth : panel evidence . Applied Economics Letters , 22 (2), 163-167.
UNCTAD (2020). The Least Developed Countries Report 2020: Productive Capacities for the New Decade. United Nations publication. Sales No. E.21.II.D.2 New York and Geneva.
Vamvakidis , A. (2002). How robust is the growth-openness connection ? Historical evidence . Journal of Economics Growth , 7 , 57-80.
van der Ploeg , R., & Poelhekke , S. (2009). The volatility curse : revisiting the paradox of plenty .
Wantchekon , L. (2002). Why do resources abundant countries have authoritarian governments ? Journal of Africa Development , 5 (2), 145-176.
Windmeijer , F., 2005. A finish sample correction for the variance of linear efficient twostep GMM estimators . J. Econom . 126 (1), 25–51.
Xin, Y., Ajaz , T., Shahzad, M., & Luo, J. (2023). How productive capacities influence trade-adjusted resources consumption in China: Testing resource-based EKC. Resources Policy , 81 , 103329.
Xue , J., Rasool, Z., Nazar , R., Khan, AI, Bhatti, SH, & Ali, S. (2021). Revisiting natural resources—Globalization-environmental quality nexus: Fresh insights from South Asian countries. Sustainability , 13 (8), 4224.
Zallé , O. (2019). Natural resources and economics growth in Africa : The role of institutional quality and human capital. Resources Policy , 62 , 616-624.

UNCTAD defines productive capacities as “the productive resources, entrepreneurial capabilities and production linkages that together determine the capacity to produce goods and services enabling the growth and development of a country” (UNCTAD, 2006, 2020a, 2020b).
The United Nations Conference on Trade and Development (UNCTAD) released a new Productive Capacity Index in February 2021 to help diagnose the performance of productive capacities and assess appropriate policies needed to strengthen them.
These are divided into “point” resources (such as fuels, ores and metals) and “non-point” resources (such as agricultural and forestry products).
This is what we call the natural resource curse.
The UNCTAD Productive Capacity Index is a dynamic and practical tool to help developing countries understand the state of their productive capacities and how they can be improved. It builds on UNCTAD's long-standing work on productive capacities, which are essential for generating inclusive and sustainable economic growth and achieving sustainable development.
We carry out the estimations using the Stata xtabond2 module according to Roodman (2009b).

Download PDF

Reviewers agreed at journal
26 Jul, 2024
Reviewers invited by journal
19 Jul, 2024
Editor assigned by journal
12 Jul, 2024
First submitted to journal
10 Jul, 2024

You are reading this latest preprint version

Are productive capacities cursed? Evidence on the income from natural resources for African economies

Status:

Version 1

Abstract

Figures

1. Introduction

2. Literature Review

3. Empirical data and strategies

3.1. Data

3.1.1. Dependent variable

3.1.2. Independent variables

3.1.3. Control Variables

3.2. Empirical strategies

4. Results and discussions

4.1. Basic results

4.2. Robustness test

4.2.1. Curse of productive capacities: evidence on the different types of natural resource rents

4.2.2. Alternative control of endogeneity: IV-2SLS

4.2.3. Control over the bias of the limited dependent variable

4.3. Sensitivity test

4.3.1. Regional heterogeneity

4.3.2. Checking the quadratic effect

5. Conclusion

References

Footnotes

Status:

Version 1