Azimi S, Kianpoor S, Bakhshesh M. (2026). Income, Inequality, and Happiness in Iran (2011-2024): Testing Hypotheses of Materialism and Income Distribution with Bayesian and Machine Learning Methods. Social Welfare Quarterly. 26(2), 107-139. doi:10.32598/refahj.26.2.2474.2 URL: http://refahj.uswr.ac.ir/article-1-4555-en.html
Extended Abstract Introduction Happiness economics emphasizes subjective well-being as the ultimate indicator of societal progress, surpassing GDP growth. The income-happiness relationship is complex and non-linear (Easterlin Paradox; Ma et al., 2025): beyond a threshold, higher income does not consistently increase happiness due to hedonic adaptation and social comparisons. Oishi et al. (2022) proposed three hypotheses: (1) Persistent Materialism (income strongly predicts happiness regardless of context); (2) Inequality Hypothesis (rising inequality strengthens income-happiness correlation via social comparisons); and (3) End of Materialism (development diminishes income’s role). Iran’s economy (2011–2024) offers a unique shock-prone context: severe sanctions, chronic inflation, currency devaluation, and volatile real incomes. This raises a key question: does erosion of real purchasing power—not inequality per se—drive the income-happiness link? This study applies Oishi et al.’s (2022) framework to Iran for the first time, employing a Python-based analytical system combining Bayesian hierarchical modeling, machine learning (Ridge Regression, SHAP, PCA), and non-parametric methods. The research provides a localized framework for monitoring how macroeconomic shocks affect subjective well-being, informing evidence-based policy in developing economies under stress. Method This study used a hybrid Python-based framework (version 3.8), analyzing time-series and panel data from 2011–2024 (provincial panel: 2011–2021). Key variables: Subjective well-being: Three-year moving average of life satisfaction scores (Cantril ladder) from the World Happiness Report. Income: GDP per capita (PPP, constant 2017 international dollars) from Maddison Project (2023) and World Bank; real household income per capita (PPP-adjusted) from Statistical Center of Iran and World Bank. Inequality: Gini coefficient, bottom 50% and top 10% income shares from World Inequality Database, Iranian Central Bank, and Ministry of Cooperatives. The analysis integrated: (1) Dynamic cumulative correlation with Fisher transformation; (2) Bayesian hierarchical MCMC (10,000 iterations, 2,000 burn-in); (3) Machine learning: Ridge Regression (multicollinearity), SHAP (feature importance), PCA (latent structures). Additional techniques: interaction terms, k-fold CV (k=5), VIF, Breusch-Pagan tests, 95% confidence/credible intervals. Models evaluated with adjusted R² and BIC. This rigorous approach ensured robustness despite limited observations (effective n=12 after moving averages). Findings Analyses confirmed the Persistent Materialism Hypothesis: Real household purchasing power (PPP-adjusted income) and GDP per capita were the strongest positive predictors of happiness (Ridge coefficients: 0.15–0.25; SHAP values up to 0.12). The income-happiness correlation strengthened over time (from r≈0.25 to r≈0.64), driven by absolute material security. In contrast, the Inequality Hypothesis was rejected: After controlling for real income and GDP, inequality measures (Gini, top/bottom 10% shares) showed insignificant effects (coefficients near zero, credible intervals including zero; SHAP ≈0). PCA revealed two components: PC1 (43.5% variance) loaded on economic welfare (GDP, income, happiness >0.75), while PC2 captured inequality contrasts with minimal link to happiness. Descriptive trends showed modest Gini increase (+1.4 points), rising top 10% share, and 8.7% happiness rise amid inflation shocks. The End of Materialism Hypothesis was also rejected. These consistent results across Ridge, SHAP, PCA, and Bayesian methods highlight that in Iran’s sanction-hit, high-inflation economy, declining real purchasing power—not inequality—primarily drives happiness. Discussion This study reveals that in Iran’s turbulent economy (2011–2024), preserving real purchasing power is the core driver of the income-happiness relationship, overshadowing income inequality. Although the strengthening income-happiness correlation mirrors patterns in stable economies like the US (Oishi et al., 2022), Iran’s underlying mechanism differs. Acute instability from sanctions, chronic inflation, and currency devaluations makes absolute financial resources critical for daily survival. Consequently, relative social comparisons—central to the Inequality Hypothesis—become secondary as households prioritize basic material security. This variation underscores the need to localize happiness theories rather than applying them universally (Stevenson & Wolfers, 2008; Ma, 2025). Imported models often overlook how macroeconomic volatility reshapes responses to income. Policy implications prioritize macroeconomic stabilization: inflation control, currency support, and protecting real household incomes over solely reducing Gini coefficients. Targeted support for the bottom 50% during crises prevents sharp welfare losses. Policymakers should develop an integrated well-being dashboard tracking GDP per capita, PPP-adjusted income, and happiness using the replicable Python-based framework proposed here. Key limitations include short time-series (reduced statistical power), reliance on aggregate rather than individual-level data, and incomplete incorporation of non-economic factors (culture, family, social resilience). Future research should use nationally representative household surveys, explore non-linear threshold effects (e.g., GAM or regime-switching), and conduct comparative analyses with other shock-prone economies (e.g., Venezuela, Turkey, Russia). Ultimately, this work delivers a novel, robust, replicable analytical tool for monitoring subjective well-being in volatile developing contexts, providing clear evidence: under prolonged economic turbulence, material security remains paramount for sustaining happiness. Ethical Considerations Compliance with Ethical Guidelines This study was conducted using secondary and publicly available data and did not require approval from an ethics committee. Authors’ contributions All authors contributed to all stages of the research, including study design, data analysis, interpretation of results, and final writing. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Conflicts of interest The authors declare no financial or non-financial conflicts of interest related to this research.