Sustainable Development Research

Measuring what matters for human progress and planetary boundaries

Developing composite indicators and analytical frameworks that bridge environmental science with policy—quantifying how societies can advance well-being within ecological limits.

PhD Environmental & Development Studies Social Progress Imperative Prague, Czechia

Research Themes

03 PILLARS
01

Composite Well-being Measurement

Advancing beyond-GDP metrics through the Social Progress Index framework. Developing time-varying weighting schemes and non-compensatory aggregation techniques to capture multidimensional societal development across 170+ countries over three decades.

Social Progress Index Entropy Weighting Panel Data
02

Environmental Efficiency of Well-being

Investigating how nations achieve social progress while minimizing environmental impact. Creating metrics like the Just Transition Score and Carbon Efficiency of Social Progress to assess sustainability performance and convergence dynamics.

Just Transition Carbon Efficiency Convergence Analysis
03

Climate Change Perception

Analyzing public attitudes toward climate risks across 100+ countries using large-scale survey data. Constructing latent variable models to understand awareness, worry, and action readiness—linking perception to policy effectiveness.

Climate Perceptions Index Latent Variables Random Forest

Publications

07 PAPERS
2025

The Evolving "Importance" of Social and Governance Measures Over Time

Htitich, M., Harmáček, J., & Krylova, P. Global Policy, 16, 1021-1038

Published
2025

Social Progress Index 1990–2020: Measuring Societal Wellbeing Over 31 Years

Krylova, P., Harmáček, J., & Htitich, M. International Journal of Social Research Methodology

Published
2025

Multidimensional Cross-Country Measurement of Climate Change Perceptions

Htitich, M., Harmáček, J., Lisney, J., & Krylova, P. Measurement: Interdisciplinary Research and Perspectives, 23(4)

Published
2025

SDG10: Inequalities Beyond Income—A Longitudinal Perspective Through the Social Progress Index

Krylova, P., Peiró-Palomino, J., Htitich, M., & Harmáček, J. The Elgar Companion to SDG Indicators

Accepted
2024

Just Transition Score: Measuring the Relative Sustainability of Social Progress

Htitich, M., Krylová, P., & Harmáček, J. Environmental and Sustainability Indicators, 23, 100440

Published
2025

Achieving Just Transition Through Convergence in the Carbon Efficiency of Social Progress

Htitich, M., Harmáček, J., & Krylova, P. Geography and Sustainability

Under Revision
2025

Dissimilarities in the Socio-economic Predictors of Climate Change Worry and Action

Htitich, M., et al. Computational Social Science

Submitted

Composite Indices

03 METRICS
JTS
Just Transition Score
Material & Carbon Efficiency
→ View methodology & policy relevance
CESP
Carbon Efficiency of Social Progress
Convergence Analysis
→ View methodology & policy relevance
CPI
Climate Perceptions Index
100k+ Survey Responses
→ View methodology & policy relevance
JTS
Just Transition Score
📊
170+ Countries
📅
1990–2020
📄
Environmental and Sustainability Indicators (2024)
Scope & Aim

The Just Transition Score measures how efficiently countries achieve social progress relative to their environmental footprint. It integrates two efficiency dimensions: material footprint efficiency (SPI per ton of material consumption) and carbon footprint efficiency (SPI per ton of CO₂ emissions).

This dual-efficiency framework captures the "safe and just space" concept from Raworth's Doughnut Economics—identifying countries that deliver high well-being outcomes while staying within planetary boundaries.

Efficiency Framework
45 Low JTS
65 Mid JTS
85 High JTS
92 Leaders
Policy Relevance

The JTS informs climate policy by identifying which development pathways achieve human flourishing without exceeding environmental limits. It helps policymakers benchmark national performance against global leaders, prioritize investments in decoupling strategies, and track progress toward the Paris Agreement's equity principles—ensuring that climate action does not compromise development goals.

CESP
Carbon Efficiency of Social Progress
📊
170+ Countries
📅
1990–2020
📄
Geography and Sustainability (Under Revision)
Scope & Aim

CESP measures how efficiently countries convert carbon emissions into social progress, calculated as the ratio of Social Progress Index to per capita CO₂ emissions. Unlike absolute metrics, CESP captures the efficiency of emissions in delivering well-being outcomes.

The study applies Phillips-Sul club convergence analysis to examine whether countries are converging toward sustainable efficiency levels or diverging into distinct performance clusters—revealing the structural barriers and enabling conditions for just transitions.

Convergence Trajectories
Policy Relevance

CESP provides critical insights for climate negotiations and technology transfer policies. By identifying convergence clubs, it reveals which country groups share common challenges and could benefit from targeted cooperation. The metric supports the Paris Agreement's differentiated responsibilities framework by quantifying emission efficiency gaps between developed and developing nations—informing equitable burden-sharing mechanisms.

CPI
Climate Perceptions Index
📊
100+ Countries
👥
100k+ Survey Responses
📄
Measurement (2025)
Scope & Aim

The Climate Perceptions Index captures public attitudes toward climate change across three latent dimensions: awareness (knowledge of climate change causes), risk perception (perceived threat level), and willingness to act (support for policy interventions).

Using Bayesian confirmatory factor analysis on data from the Lloyd's Register Foundation World Risk Poll, the CPI reveals how climate attitudes cluster across populations and identifies the socio-economic predictors that explain cross-country variation.

Three Dimensions of Climate Perception
Awareness
Risk
Action
Policy Relevance

The CPI helps governments understand where public opinion enables or constrains climate policy ambition. By decomposing perception into distinct dimensions, it reveals whether low policy support stems from awareness gaps (requiring education), risk perception deficits (requiring communication strategies), or action readiness barriers (requiring economic incentives). This granular understanding enables targeted interventions that build genuine climate constituencies.

R Package

w2m

What to Measure

A comprehensive R toolkit for sustainability research
R ≥ 4.0 MIT License v1.0.0

The w2m package provides a unified framework for constructing, analyzing, and visualizing composite indicators. It implements the methodological innovations from my dissertation research—including modified entropy weighting with temporal decay, convergence testing, and ML-based variable importance ranking.

Designed for researchers working with cross-country panel data, w2m streamlines the entire workflow from data preprocessing to publication-ready outputs.

2,246+
Lines of Code
10
Modules
40+
Functions
λ
entropy_weights.R Modified entropy with decay
β
convergence.R Phillips-Sul & σ-tests
∇
importance.R RF variable ranking
◈
spatial_viz.R Choropleth maps
Σ
aggregation.R Index construction
θ
normalization.R Min-max & z-score

Technical Methods

04 TECHNIQUES
λ
Modified Entropy Weighting
Time-varying weights with temporal decay parameters to track evolving indicator importance in longitudinal data.
→ View formula & methodology
β
Convergence Analysis
Phillips-Sul club convergence and σ-convergence tests for identifying clusters and trajectories in panel data.
→ View formula & methodology
∇
Variable Importance
Random forest and permutation-based feature ranking with cross-validated importance scores.
→ View formula & methodology
ψ
Bayesian Latent Variables
Confirmatory factor analysis using MCMC estimation to extract scores for unobservable constructs.
→ View formula & methodology
λ

Modified Entropy Weighting

Time-varying weights with temporal decay for longitudinal composite indicators

Methodology

Traditional entropy weighting assigns static weights based on cross-sectional variance. The modified approach introduces a temporal decay factor that gives greater influence to recent observations while still incorporating historical patterns—essential for tracking how indicator "importance" evolves over multi-decade panels.

Core Formulas
pij(t) = Xij(t) / Σi Xij(t)
p̃ij(t) = α(T−t) · pij(t)
Hj(t) = −k Σi p̃ij(t) log(p̃ij(t))
wj(t) = (1 − Hj(t)) / Σj(1 − Hj(t))
α = decay factor (≈ e−0.0001)
T = most recent time point
k = 1/log(n) for standardization
Evolution of Entropy Weights (1990–2020)
Basic Needs
Wellbeing
State Cap.
Democracy
β

Convergence Analysis

Phillips-Sul club convergence and σ-convergence for panel trajectories

Methodology

Convergence analysis examines whether countries are becoming more similar (converging) or more different (diverging) over time. σ-convergence tracks the dispersion across the full sample, while Phillips-Sul club convergence identifies subgroups following common transition paths—revealing hidden heterogeneity in global trends.

Key Formulas
σ-convergence: σt = √[Σi(yit − ȳt)² / N]
Phillips-Sul: log(H1/Ht) − 2log(log t) = α + β log t + ut
Ht = cross-sectional variance at time t
β ≥ 0 indicates convergence
t-stat > −1.65 confirms club
Club Convergence Trajectories
∇

Variable Importance

Random forest and permutation-based feature ranking

Methodology

Random Forest importance scores quantify each predictor's contribution to model accuracy. Permutation importance measures the increase in prediction error when a variable's values are shuffled—providing a model-agnostic assessment of feature relevance that captures non-linear relationships and interactions.

Importance Measures
VIjperm = (1/B) Σb [Err(Tb, Djπ) − Err(Tb, D)]
VIjgini = Σnodes Δ Gini(j)
B = number of trees
Djπ = data with j permuted
Δ Gini = impurity reduction
Feature Importance Rankings
.42 GDP
.35 Educ.
.28 Health
.21 Gov.
.15 Urban
ψ

Bayesian Latent Variables

Confirmatory factor analysis with MCMC estimation

Methodology

Bayesian Confirmatory Factor Analysis (BCFA) extracts latent constructs from observed indicators using probabilistic inference. The approach specifies prior distributions for factor loadings and uses MCMC sampling to estimate posterior distributions—providing credible intervals rather than point estimates and naturally handling measurement uncertainty.

Model Specification
yi = Ληi + εi , εi ~ N(0, Θ)
ηi ~ N(0, Ψ) , Λ ~ N(0, 10)
y = observed indicators
Λ = factor loadings
η = latent factors
Ψ = factor covariance
Posterior Factor Score Distributions
Aware
Risk
Action

About

Mohamed Htitich is a macroeconomist and applied econometrician specializing in sustainable development measurement. His research focuses on creating robust cross-country indicators that capture multidimensional well-being while accounting for environmental constraints.


From 2021 to 2025, he served as Research Associate at the Social Progress Imperative, co-designing the conceptual framework and statistical methodology for the Social Progress Index—a comprehensive measure of societal well-being beyond GDP covering 170+ countries.


His dissertation at Palacký University Olomouc synthesizes seven research papers across three interconnected themes: composite measurement of well-being, environmental efficiency of social progress, and public climate change perception. This work advances both methodology and empirical understanding of global sustainability patterns.

Education

PhD Environmental & Development Studies
Palacký University Olomouc, 2025


MSc Applied Econometrics
Hassan II University, Casablanca

Languages

Arabic (native) • French (bilingual) • English (advanced)

Contact

[email protected]

GitHub