Analysis of CO2 Emissions, FDI, and Environmental Policies in Sub-Saharan Africa

Introduction

The econometric methods employed in the research paper “CO2 Emissions, Legal Strength & Environmental Sustainability Policies; and FDI Nexus in Sub-Saharan Africa,” written by Michael Asiedu, Emmanuel Mensah Aboagye, Benedict Arthur, and Gabriel Kyeremeh, are critically examined in this assessment. The study aims to look at the link between CO2 emissions and foreign direct investment (FDI) inflows in 54 African nations between 2004 and 2020 (Asiedu et al., 2022). The authors use the two-step Generalized Method of Moments (GMM) approach, which is especially well-suited for analyzing dynamic panel data. More accurate and dependable estimations are possible with this technique, which also helps to address any endogeneity problems. The suitability of the employed econometric techniques and the reliability of the analysis’s results will be evaluated.

Research Question

The main research question is: How do environmental sustainability regulations and legal strength affect CO2 emissions, and foreign direct investment flows in Sub-Saharan Africa?

Variables

Dependent Variables

1. CO2 emissions
2. FDI inflows

Independent Variables

1. Legal Strength
2. Environmental sustainability policies
3. GDP growth
4. Government expenditure on education (GovntExpEdu)
5. Internet usage
6. Inflation
7. Logistics performance
8. Exchange rate
9. Trade of GDP

Endogenous Variables

1. CO2 emissions
2. FDI inflows

Exogenous Variables

1. Access to electricity
2. GDP growth
3. Government expenditure on education
4. Internet usage
5. Inflation
6. Trade of GDP

Instrumental Variables

1. Property rights
2. Exchange rate
3. Logistics performance

Sampling Technique and Sample Size

Sampling Technique

This study’s sampling strategy uses panel data collected from 54 African nations between 2004 and 2020, a period of 17 years. Panel data, or longitudinal or cross-sectional time-series data, is beneficial when analyzing changes over time within particular entities, like nations. This method captures differences throughout time and across different nations, facilitating a more detailed understanding of economic trends and dynamics. Panel data, as opposed to conventional cross-sectional or time-series data, allow researchers to examine each nation’s general trend and unique trajectories, offering a more comprehensive understanding of economic events. As a result, panel data are an effective means of revealing intricate connections and trends that might not be apparent from other kinds of data.

Sample Size

The study used an extensive sample size of 918 observations (54 nations * 17 years) to examine the influence of Foreign Direct Investment (FDI) on CO2 emissions across 54 African countries from 2004 to 2020. This provides a solid foundation for analysis. Every observation captures a unique combination of nation-specific traits and time variations, providing a thorough overview of the dynamics between FDI inflows and environmental results during the specified time frame. This large dataset makes exploring slight differences, emerging heterogeneities, and nuanced patterns in various historical and geographical situations easier. The study’s wide range of country-year observations allows it to identify trends, clarify connections, and provide a valuable understanding of the complex interactions among foreign investment, environmental sustainability, and economic development in Sub-Saharan Africa.

Type of Data

This research uses panel data, which offers a solid basis for in-depth analysis. The unique quality of panel data is that it incorporates several observations across time for each unit, in this example, the nations of Africa. By simultaneously examining cross-sectional and time-series dimensions made possible by this design, changing patterns and dynamics may be better understood. Panel data facilitates the discovery of subtle associations that could go unnoticed and allows for more accurate calculations by accounting for unobserved heterogeneity at the country level. Furthermore, its capacity to record dynamic impacts across time improves the study’s capacity for prediction and provides an understanding of the persistence of occurrences. Panel data significantly deepens analysis, which makes it a priceless tool for researching intricate socioeconomic issues.

Analysis Techniques

Generalized Method of Moments (GMM), a two-step approach well-suited for dynamic panel data analysis, is the analytic technique used in this work. Endogeneity and simultaneity are two problems frequently occurring in dynamic panel setups and pose difficulties for conventional estimating techniques. Effectively addressing these issues, GMM reduces biases resulting from missing variables or measurement mistakes and provides more reliable parameter estimations. GMM is a two-step process that uses moment circumstances to estimate model parameters and then uses a system of equations to refine those estimates. This iterative technique improves speed and accuracy to investigate complex economic interactions in panel data sets. Therefore, implementing GMM highlights the research’s dedication to thoroughly examining the intricate relationship among foreign direct investment, greenhouse gas emissions, and environmental regulations throughout Sub-Saharan Africa.

Estimation Technique

Panel data regression models are often used in econometric analysis and are most likely the primary estimation method used here. Models with fixed effects (FE) or random effects (RE) may be utilized to consider unobserved heterogeneity between nations. Furthermore, sophisticated techniques such as the Generalized Method of Moments (GMM) or Two-stage least squares (2SLS) are probably used to handle endogeneity problems. These methods ensure the trustworthiness of the estimation findings by allowing the management of simultaneity using instrumental variables and demonstrating robustness in managing endogeneity problems. The assumptions about the type of unobserved heterogeneity determine which model to use: FE or RE. In panel data contexts, 2SLS and GMM provide versatile ways to deal with endogeneity issues.

Econometric Specification

This study’s econometric model aims to examine the factors that influence CO2 emissions in Sub-Saharan Africa. Here are the model’s specifications:

CO2 emissions 𝑖 𝑡 = 𝛽 0 + 𝛽 1 Legal strength 𝑖 𝑡 + 𝛽 2 Env. sustainability policies 𝑖 𝑡 + 𝛽 3 FDI 𝑖 𝑡 + 𝛽 4 𝑋 𝑖 𝑡 + ϵ 𝑖 𝑡

The link between CO2 emissions (the dependent variable) and environmental sustainability policies, legal strength, foreign direct investment (FDI), and control variables (represented by 𝑋 𝑖 𝑡) is shown in this equation. In this case, 𝑖 stands for the nation and 𝑡 for the era.

FDI 𝑖 𝑡 = 𝛾 0 + 𝛾 1 Legal strength 𝑖 𝑡 + 𝛾 2 Env. sustainability policies 𝑖 𝑡 + 𝛾 3 CO2 emissions 𝑖 𝑡 + 𝛾 4 𝑍 𝑖 𝑡 + 𝜈 𝑖 𝑡

The correlation between FDI (the dependent variable) and environmental sustainability policies, CO2 emissions, legal strength, and control factors denoted by 𝑍 𝑖 𝑡 is modeled by this equation.

Where:

• Legal strength 𝑖 𝑡: A measure of the strength of the country’s legal system at time 𝑏.
• Env. sustainability policies 𝑖 𝑡: A measure of the environmental sustainability policies that were put into place at time 𝑡 in nation 𝑖.
• FDI 𝑖 𝑡: Inflows of foreign direct investment into the nation 𝑖 during time 𝑡.
• CO2 emissions 𝑖 𝑡: Emissions of carbon dioxide in nation 𝑖 at time 𝑡.
• 𝑋 𝑖 𝑡: Control variable vector that affects CO2 emissions.
• 𝑍 𝑖 𝑡: Control variable vector affecting foreign direct investment.
• ϵ 𝑖 𝑡 and v 𝑖 𝑡: Error words that indicate deviations in the model that cannot be explained.

Assumptions of the Estimation Technique

This study uses the two-step system Generalized Method of Moments (GMM) estimate approach. The validity of this approach is predicated on many assumptions:

1. Absence of serial correlation in the error terms: This premise necessitates the absence of cross-temporal correlation among the model’s mistakes.
2. Instrument homogeneity: The instruments utilized in the GMM estimate must be exogenous so they cannot be associated with the regression equation’s error term.
3. Homoscedasticity: The error term’s variance should be constant throughout all observations.
4. Validity of instruments: For the estimation to be valid, the instruments must be uncorrelated with the error term and associated with the endogenous regressors, FDI, and CO2 emissions.

These presumptions provide substantial insights into the link between CO2 emissions, FDI, legal strength, and environmental sustainability policies in Sub-Saharan Africa. They also assure the validity and reliability of the estimation findings derived using the two-step system GMM approach.

Critical Evaluation

Assumptions Plausibility

A key component of guaranteeing the validity of the statistical analysis is the assumption of homoscedasticity and the absence of serial correlation. It is crucial to recognize that these presumptions might be broken in real life, which could provide biased outcomes. For example, missing variables or incorrect model specifications may result in serial correlation, the correlation between error terms over time. Similarly, if the variability of residuals varies over time or between groups, homoscedasticity—which presupposes constant variance of error terms across observations—may be broken. To evaluate how resilient the results are to deviations from these presumptions, robustness tests and sensitivity analysis should be conducted.

Endogeneity

Considering the endogeneity between Foreign Direct Investment (FDI) and CO2 emissions and the dynamic character of the panel data, the two-step system Generalized Method of Moments (GMM) approach was used. When an explanatory variable and the error term are associated, endogeneity occurs, which results in skewed and inconsistent parameter estimations. According to this study’s background, FDI and CO2 emissions are probably influenced by several variables simultaneously, including environmental regulations and economic growth. Through using instrumental variables and unobserved heterogeneity control (GMM), the study attempts to reduce the possibility of biased estimations and enhance the validity of the findings.

Instrumental Variables

In panel data analysis, the validity of instrumental factors is crucial for resolving endogeneity and generating objective estimates. One uses tools like property rights and currency rates to instrument for the possibly endogenous explanatory factors, like FDI. To guarantee that the instrumental variables satisfy the relevance and exogeneity requirements, it is crucial to choose and justify them carefully. Relevance is the capacity of the instruments to forecast the endogenous variables correctly; on the other hand, exogeneity indicates that there is no correlation between the instruments and the error term. To guarantee that instrumental variables are appropriate for the study, thorough testing and validation must be done. Sensitivity analysis, such as the Sargan test for overidentifying constraints, can be used to evaluate how resistant the results are to instrument choice and how genuine the instruments are.

Methodological Rationale

The choice of research technique is an important consideration that dramatically impacts the reliability and validity of the study’s conclusions. The use of the two-step system Generalized Method of Moments (GMM) technique in the research article “CO2 Emissions, Legal Strength & Environmental Sustainability Policies; and FDI Nexus in Sub-Saharan Africa,” written by Michael Asiedu et al., shows that the research question, the characteristics of the data, and the necessity of addressing potential methodological challenges were all carefully considered. This section examines different estimating strategies that may be compatible with the data type and research issue and the reasoning for the methodology’s selection.

The Rationale for Choosing Two-Step System GMM

Dynamic Nature of Panel Data: Over 16 years (2004-2020), the study looks at how Foreign Direct Investment (FDI) affects CO2 emissions in 54 African nations. Panel data, observations made over time on several entities, are dynamic since variables will likely be associated with time. Since the two-step system, GMM, permits parameter estimation while accounting for endogeneity and unobserved heterogeneity, it is ideally suited for modeling dynamic panel data.

Endogeneity Concerns: Biased parameter estimates result from endogeneity, the correlation between explanatory factors, and the error term. According to the study, FDI and CO2 emissions may be influenced by several variables simultaneously, including environmental regulations and economic growth. Using instrumental variables and accounting for unobserved heterogeneity, the two-step GMM reduces the possibility of biased estimates and handles endogeneity.

Strategy for Instrumental Variables: Instrumental variables can be used with the GMM framework to handle endogeneity. Property rights and currency rates are two examples of instrumental variables included in the study that are exogenous to the model and theoretically relevant. By capturing FDI variance that is not directly impacted by CO2 emissions, these instruments should improve the identification technique and increase the validity of the findings.

Efficiency and Consistency: The GMM estimator is known for being efficient and consistent, especially when endogeneity and unobserved heterogeneity are present. The study attempts to achieve impartial and consistent parameter values using the two-step estimation approach, offering more trustworthy insights into the link between FDI and CO2 emissions in Sub-Saharan Africa.

Alternative Estimation Techniques

Other estimating techniques offer Different data analysis methods, which can also give complimentary insights into the connection between variables. The Instrumental Variable (IV) method uses exogenous instruments to address endogeneity, whereas Fixed Effects (FE) and Random Effects (RE) models help capture unobserved variability. A valuable tool for assessing policy initiatives is Difference-in-Differences (DID) analysis, whereas Structural Equation Modeling (SEM) facilitates exploring intricate interactions between several variables. Every approach has advantages and disadvantages; therefore, the selection should be based on the study’s goals and the nature of the data. Investigating other approaches can strengthen the validity of results and provide a more thorough comprehension of the phenomenon being studied.

Fixed Effects (FE) and Random Effects (RE) Models: In panel data analysis, the Fixed Effects (FE) and Random Effects (RE) models provide two different ways to deal with unobserved heterogeneity. FE models can effectively account for time-invariant unobserved influences by including dummy variables for every entity while capturing individual-specific effects. By treating individual impacts as random and uncorrelated with the regressors, RE models, on the other hand, offer effective estimates of the average impact of FDI on CO2 emissions. While RE models help estimate population-average impacts across entities, FE models are effective when the focus is on estimating individual-specific effects, such as country-specific policies or characteristics. The link between FDI and CO2 emissions may be better understood using the FE and RE models, which give researchers flexibility in modeling unobserved variability while accounting for time-invariant variables.

Instrumental Variable (IV) Approach: The Instrumental Variable (IV) method provides a useful tactic for addressing endogeneity problems by using instrumental variables to instrument for possibly endogenous regressors. IV estimate is based on distinct identification procedures and assumptions about the instrument’s validity, in contrast to GMM. IV approaches are especially appropriate when there is uncertainty regarding the exogeneity of instrumental factors or when researchers want to evaluate outcomes across various identification processes. Using theoretically relevant and exogenous instrumental variables, IV estimation improves the validity of empirical findings by providing researchers with objective and consistent parameter estimations. Furthermore, IV techniques can offer essential insights into the underlying mechanisms influencing the association between variables of interest and enable the investigation of alternate causal paths.

Difference-in-Differences (DID) Analysis: A popular technique for assessing the causal influence of interventions or policy changes is called Difference-in-Differences (DID) Analysis, which compares the variations in outcomes over time between a treatment group and a control group. This method is quite helpful when examining the effects of specific policy interventions or regulatory changes on Foreign Direct Investment (FDI) and CO2 emissions. A greater understanding of the cause-and-effect link between policy changes and economic and environmental consequences is made possible by DID analysis, which isolates the impacts of these interventions from other confounding factors. In the context of Sub-Saharan Africa or any other pertinent region, researchers can determine the causal impact of interventions by comparing treated and untreated entities in an organized manner over time. This helps clarify policy measures’ efficacy in influencing FDI inflows and mitigating CO2 emissions.

Structural Equation Modeling (SEM): A robust statistical technique that makes it possible to investigate complex interactions between many variables is structural equation modeling or SEM. SEM might provide insights into the relationship between Foreign Direct Investment (FDI) and CO2 emissions, considering intermediary factors, including institutional quality, economic development, and environmental legislation. It is advantageous when examining complex routes of effect. SEM may provide a detailed knowledge of the underlying dynamics by clarifying the intricate pathways through which FDI affects CO2 emissions by simulating both direct and indirect effects. This methodological approach makes it easier to explore complex causal pathways and illuminates the complex relationship between foreign direct investment (FDI) and environmental outcomes in Sub-Saharan Africa. It may also reveal subtle interactions that are not always visible using more conventional analytical techniques.

Conclusion

In summary, Asiedu, Aboagye, Arthur, and Kyeremeh’s study “CO2 Emissions, Legal Strength & Environmental Sustainability Policies; and FDI Nexus in Sub-Saharan Africa” provides insightful information about the complex interplay between FDI and CO2 emissions in the context of Sub-Saharan Africa. The study provides a detailed picture of the relationship between FDI inflows and CO2 emissions through rigorous econometric analysis using the two-step system Generalized Method of Moments (GMM) technique. It also emphasizes the importance of environmental policies and legal strength in preventing potential environmental degradation. Through a thorough examination of an extensive dataset covering 54 African nations over an extended period, the study offers a solid basis for comprehending the intricacies of foreign direct investment’s ecological ramifications in the area. Though the selected methodology provides insightful information, more investigation should look at different methods, such as Structural Equation Modeling (SEM), to better understand the complex causal processes at play. This study contributes substantially to the conversation about environmental policies and sustainable development in Sub-Saharan Africa by highlighting the significance of solid environmental governance frameworks and balanced economic growth.

Reference

Asiedu, M., Aboagye, E. M., Arthur, B., & Kyeremeh, G. (2022). CO2 emissions, legal strength & environmental sustainability policies; And FDI nexus in sub-Saharan Africa.