TRAINING SCHOOL

The Training School on Econometrics for the Environment, organized in conjunction with the CE² 2026 International Conference, is designed to provide PhD students, early-career researchers, and practitioners with advanced theoretical and practical skills for analyzing environmental, climate, energy, and natural resource issues.

The programme will focus on three major methodological areas: Panel Data Econometrics and Structural Equation Modeling. Through a combination of lectures and hands-on applications using relevant statistical software and real-world data, participants will gain practical experience in applying these modern econometric approaches to environmental and sustainability challenges.

The Training School will provide an excellent opportunity to strengthen research capacities, exchange ideas with leading international scholars, and develop innovative empirical approaches for addressing complex environmental and climate-related issues.

Lecture #1: Panel Data Econometrics: Theory and Applications Using Stata

December 18th, 2026

By Yassine Sbai Sassi, New York University, USA, 

Yassine Sbai Sassi is an Assistant Professor of Economics at New York University. He received his Ph.D. in Economics from the University of California, Berkeley. His research focuses on econometric theory, specializing in the statistical analysis of network models, dyadic interactions, and panel data.

 Ahmed El Ghini, Mohammed V University in Rabat, Morocco,

Ahmed El Ghini is Professor of Econometrics at Mohammed V University in Rabat, Morocco, and a former postdoctoral researcher at CNRS and the University of Lille, France.
His research focuses on econometrics, time series analysis, and statistical methods applied to economics, the environment, energy, and finance.
He is the Founding Chair of the international symposia Days of Econometrics for Finance and Econometrics for the Environment.

                                 and Slimane Regui, Mohammed V University in Rabat, Morocco.

Slimane Regui is an Assistant Professor of Statistics at Mohammed V University in Rabat, Morocco. He received his Ph.D. in Applied Mathematics, specializing in Statistics, from Abdelmalek Essaâdi University. His research focuses on statistical inference and econometric theory, with particular interests in periodic coefficient regression models, asymptotic theory, and hypothesis testing. 

Course Description

This short course provides a comprehensive introduction to the theoretical foundations and empirical applications of panel data econometrics. It combines methodological lectures with hands-on computer sessions using Stata, enabling participants to develop both a solid understanding of panel data methods and the practical skills needed to implement them in applied economic research.

The course is organized into two complementary parts.

Part I: Theoretical Foundations of Panel Data Econometrics (Yassine Sbai Sassi, 3 hours. Syllabus)

The first part focuses on the fundamental concepts, identification assumptions, and estimation techniques underlying panel data models. It begins by examining the advantages of panel data, with particular attention to unobserved heterogeneity, within-unit versus between-unit variation, and the distinction between persistence due to individual heterogeneity and genuine dynamic effects.

The course then covers the identification and estimation of linear panel data models, including fixed effects, first differences, random effects, and correlated random effects. It also addresses statistical inference, serial dependence, and cluster-robust standard errors. Particular attention is devoted to the limitations of conventional fixed-effects estimation when strict exogeneity fails, as well as to dynamic panel data models, Nickell bias, and the Arellano–Bond difference GMM estimator using internal instruments.

Throughout this part, the emphasis is on understanding identification assumptions, interpreting estimators, and determining the appropriate econometric approach for different empirical settings.

Part II: Applications in Economics Using Stata (Ahmed El Ghini and Slimane Regui, 2 hours)

The second part is devoted to the practical implementation of panel data econometric methods using Stata. Through guided computer exercises and applied economic examples, participants will learn how to prepare and structure panel datasets, estimate alternative panel data models, conduct specification tests, and interpret empirical results.

The practical sessions will cover:

• Data management and exploration: Structuring panel datasets, declaring panel identifiers and time variables, and exploring within-unit and between-unit variation.

• Estimation of static panel models: Implementing pooled OLS, fixed-effects, first-difference, and random-effects estimators, and comparing their results.

• Model specification and inference: Examining the relevance of individual effects, testing model assumptions, and implementing cluster-robust standard errors.

• Endogeneity and feedback effects: Exploring the implications of alternative exogeneity assumptions and their consequences for empirical estimation.

• Dynamic panel data models: Estimating dynamic specifications and introducing the Arellano–Bond difference GMM estimator, including the use of internal instruments.

• Economic interpretation and empirical reporting: Interpreting estimated coefficients, assessing the robustness of results, and presenting findings in a form suitable for academic research and policy analysis.

Applications will draw on standard economic examples, such as production functions, wage equations, and other relationships involving repeated observations over time. The methods can also be applied to empirical questions in environmental and energy economics, climate change, agricultural economics, and related fields.

Learning Objectives

By the end of the course, participants will be able to:

• Understand the main identification assumptions underlying panel data estimators.

• Select appropriate econometric methods for static and dynamic panel data models.

• Implement panel data estimation and inference procedures in Stata.

• Recognize potential sources of endogeneity and understand the role of internal instruments.

• Interpret and critically assess empirical results in applied economic research.

Prerequisites

Participants should have some familiarity with ordinary least squares (OLS), instrumental variables, and the generalized method of moments (GMM). Basic familiarity with Stata is helpful but not essential.

Target Audience

The course is intended for Master's and PhD students, early-career researchers, and applied economists seeking to strengthen their theoretical and empirical skills in panel data econometrics. It will be particularly relevant to researchers working in environmental economics, energy economics, climate change, and other fields involving longitudinal or country-level economic data.

Main References

The theoretical part draws primarily on lecture notes by Manuel Arellano, Badi Baltagi, Victor Chernozhukov and Iván Fernández-Val, and Guido W. Imbens and Jeffrey M. Wooldridge, as well as the foundational contribution by Arellano and Bond (1991) on specification testing and dynamic panel data estimation.

Lecture #2:  Structural Equation Modeling

December 18th, 2026

Title: Structural Equation Modeling for Environmental and Sustainability Research 

By Prof. Zouhair El Hadri, Mohammed V University in Rabat

Short Bio: Zouhair El Hadri is a Full Professor of Mathematics at the Faculty of Sciences of Mohammed V University in Rabat, Morocco. He is a member of the Laboratory of Mathematics, Statistics and Applications.

His research focuses on multivariate statistics and statistical modeling, with particular emphasis on Structural Equation Modeling (SEM), including both Partial Least Squares SEM (PLS-SEM) and Covariance-Based SEM (CB-SEM). His research interests also include Path Analysis, Factor Analysis, and Multiblock Data Analysis.

1. Course Description and Learning Objectives

This advanced training course provides a methodological and applied introduction to Structural Equation Modeling (SEM) for empirical research in environmental economics, sustainability science, climate change, and related fields. It is designed for Master’s and Ph.D. students, researchers, and academics who wish to develop the theoretical understanding and practical skills required to formulate, estimate, assess, and interpret SEM models.

The course presents SEM as an integrated framework combining regression analysis, path analysis, factor analysis, and latent-variable modeling. Particular attention is given to the distinction between the measurement model and the structural model, model specification and identification, estimation and model assessment, and the conceptual differences between Covariance-Based SEM (CB-SEM) and Partial Least Squares SEM (PLS-SEM).

Throughout the course, environmental and sustainability-related examples will be used to illustrate how SEM can be employed to analyze complex relationships involving environmental awareness, sustainable behavior, climate change perceptions, green finance, corporate social responsibility, environmental policy, renewable energy, water management, and other sustainability outcomes.

Learning Objectives

By the end of the training, participants should be able to:

1. Understand the transition from conventional regression models to Structural Equation Modeling (SEM).

2. Distinguish between observed/manifest variables and latent/unobserved constructs, and understand their roles in empirical research.

3. Understand and interpret PCA, EFA, and CFA as complementary tools for dimensionality reduction, construct identification, and measurement validation.

4. Specify, estimate, and interpret Path Analysis models for analyzing direct and indirect relationships among observed variables.

5. Distinguish clearly between the measurement and structural components of an SEM and understand how they interact.

6. Master the fundamental terminology, notation, assumptions, and graphical representation of SEM.

7. Understand the main stages of SEM analysis, including model specification, identification, estimation, evaluation, and interpretation.

8. Understand the conceptual and methodological differences between CB-SEM and PLS-SEM, including their respective objectives, assumptions, and fields of application.

9. Apply SEM techniques to environmental and sustainability research questions, using real or realistic empirical examples.

10. Use SEM software to implement empirical models, with practical demonstrations using lavaan and plspm R packages. 

11. Critically evaluate and communicate SEM results in the context of academic research and applied environmental policy analysis.

2. Pedagogical Structure and Program

The training combines methodological lectures, graphical illustrations, empirical examples, and software demonstrations. The program is organized progressively, moving from conventional multivariate methods to advanced SEM applications.

Part I - From Regression to Structural Equation Modeling

1. Multiple Regression Model - 15 min

• From regression analysis to multivariate modeling

• Direct relationships among observed variables

• Limitations of conventional regression for complex causal structures

• Motivation for Structural Equation Modeling

2. Path Analysis - 20 min

• Concept and graphical representation

• Direct and indirect effects

• Recursive path models

• Path coefficients and interpretation

• Environmental applications: determinants of sustainable behavior and environmental outcomes

3. Principal Component Analysis (PCA) - 15 min

• Motivation and objectives

• Dimensionality reduction

• Principal components and interpretation

• Applications to environmental and sustainability indicators

4. Exploratory Factor Analysis (EFA) - 15 min

• Identifying underlying dimensions and latent constructs

• Factor extraction and rotation

• Factor loadings and interpretation

• Applications to environmental attitudes, perceptions, and sustainability-related indicators

5. Confirmatory Factor Analysis (CFA) - 15 min

• From exploratory to confirmatory analysis

• Specification of measurement models

• Factor loadings and measurement errors

• Environmental and sustainability measurement examples

6. SEM: Vocabulary, Notation, and Graphical Representation - 10 min

• Latent and observed variables

• Exogenous and endogenous variables

• Indicators and measurement errors

• Structural paths and covariances

• Model diagrams and SEM notation

Coffee & Networking Break - 30 min

Part II - Covariance-Based Structural Equation Modeling

7. CB-SEM: From Model Specification to Empirical Estimation - 35 min

• Model specification

• Identification conditions

• Estimation methods

• Model fit and goodness-of-fit measures

• Assessment of measurement models

• Structural model evaluation

• Direct, indirect, and total effects

• Interpretation and reporting of SEM results

8. Applied Example 1: Sustainable Water Management - 10 min

Understanding Users’ Behavioral Intention to Adopt a Search Engine that Promotes Sustainable Water Management

• Conceptual model and hypotheses

• Latent constructs and observed indicators

• Measurement and structural models

• Estimation and model assessment

• Interpretation of direct and indirect effects

• Software demonstration: lavaan R package

Part III - Partial Least Squares Structural Equation Modeling

9. PLS-SEM: Concepts, Procedures, and Estimation - 35 min

• Objectives and characteristics of PLS-SEM

• Reflective and formative measurement models

• PLS-SEM modes

• Weighting schemes

• PLS algorithm

• Parameter estimation

• Assessment of measurement models

• Structural model assessment

10. Applied Example 2: Green Finance, Green Taxation, and Corporate Social Responsibility - 10 min

Unveiling the Impact of Green Taxation and Green Financing on Corporate Social Responsibility

• Conceptual framework

• Latent constructs and indicators

• PLS-SEM estimation

• Interpretation of structural relationships

• Assessment of mediation and/or indirect effects

• Software demonstration: plspm R package

3. Expected Outcomes

At the end of the training, participants will have developed a practical and conceptual understanding of SEM, enabling them to:

• formulate SEM-based research questions and conceptual frameworks;

• translate theoretical concepts into measurable latent constructs;

• design and assess measurement instruments;

• specify and estimate CB-SEM and PLS-SEM models;

• evaluate measurement and structural models;

• interpret direct, indirect, and total effects;

• use lavaan and plspm for empirical applications; and

• apply SEM methodology to environmental, climate, sustainability, and green-economy research.

The training is particularly relevant for empirical studies dealing with environmental behavior, climate-change perceptions, sustainable consumption, water and natural-resource management, renewable energy adoption, green finance, environmental policy, corporate sustainability, and ESG-related research.