How popularising higher education affects economic growth and poverty alleviation: empirical evidence from 38 countries

The study situates the popularisation of higher education within UNESCO's Sustainable Development Goal 4, noting rapid expansion of higher education and its complex implications for management, purpose, and societal outcomes. Prior work underscores education's role in economic development and poverty reduction through returns to education, yet also raises concerns about diminishing marginal economic effects as higher education scales. Gaps identified include limited simultaneous empirical analysis of how popularising higher education affects both economic growth and poverty alleviation across countries. The research aims to examine how population segments with higher education and adult education relate to GDP growth and poverty reduction in OECD countries from 1995–2021. The study articulates four research questions: (Q1) What is the effect of a population segment with higher education on GDP growth? (Q2) What is the effect of adult education on GDP growth? (Q3) What impact does a population segment with higher education have on reducing the proportion of those experiencing poverty? (Q4) What is the relation between an increased level of adult education and the proportion of the population experiencing poverty?

Popularising higher education and economic growth: The literature portrays higher education as a driver of human capital, productivity, and technological advancement, with multiple studies documenting positive associations between higher education expansion and economic development in OECD and other contexts. Evidence includes significant correlations in EU countries and findings that higher education has the strongest impact on labor productivity among education levels. However, some studies suggest marginal or trade-off effects between quantity expansion and quality upgrading, particularly in graduate education, with effects concentrated in high-tech fields. International cooperation and spillovers enhance growth impacts. Popularising higher education and poverty alleviation: Higher education expansion can reduce poverty and inequality by improving skills, fostering intergenerational mobility, and enhancing total factor productivity, though education alone is not a panacea. Educational poverty is influenced by student and school factors. Sectoral alignment (e.g., agricultural human capital in Africa) and quality improvements are crucial. Theoretical underpinnings: Grounded in Solow (technological progress), Schultz (human capital and returns to education), Lucas (internal/external effects of human capital with increasing returns), and Romer (endogenous technological change), the literature supports pathways through which higher education affects growth and poverty via productivity, innovation, and human capital accumulation.

Design: Empirical panel analysis of 38 OECD countries (1995–2021) using Ordinary Least Squares (OLS) and panel quantile regression. Data sources: OECD database and World Bank indicators. Variables: Independent variables capture higher education levels and popularisation. A = proportion of population aged 25–64 that completed higher education (overall level of higher education among working-age population). B = proportion of population aged 25–34 that completed higher education relative to peers (degree of popularisation). Dependent variables: C = GDP (total annual domestic economic production, USD), D = poverty line (share living on less than $2.15/day, 2017 PPP). Analytical models: Based on the Mankiw–Romer–Weil extension of the Solow model with higher education measures as core growth inputs. Log-linear forms specified: ln(Yit) = θ0 + α ln(Qit) + β ln(Pit) + μit; with specific implementations ln(Cit) = α ln(Ait) + β ln(Bit) and ln(Dit) = α ln(Ait) + β ln(Bit). Hypotheses: H1: Higher-educated population share positively impacts GDP growth. H2: Adult education/popularisation positively impacts GDP growth. H3: Higher-educated population share reduces poverty. H4: Adult education/popularisation reduces poverty. Estimation: OLS for GDP and poverty equations; model diagnostics included White and Breusch–Pagan tests indicating heteroscedasticity; robust standard errors used. Quantile regression estimated at τ = 0.10 to 0.90 (increments of 0.10) to capture heterogeneous effects across the distributions of GDP and poverty. Sample sizes: OLS GDP model n = 859; OLS poverty model n = 635.

OLS results for GDP (C) on A and B: Model R^2 = 0.097; F(2,856) = 46.137, p < 0.001. Estimated equation: C = 4.763 − 0.813·B + 1.553·A. A (25–64 higher education share) positive and significant (t = 7.141, p < 0.01), indicating higher working-age tertiary attainment is associated with higher GDP. B (25–34 higher education share) negative and significant (t = −4.300, p < 0.01), suggesting at the mean that increased popularisation among young adults is associated with lower GDP, potentially reflecting overcapacity or lagged returns during expansion phases. OLS results for poverty line (D): R^2 = 0.134; F(2,632) = 48.771, p < 0.001. Estimated equation: D = 2.284 − 0.048·B + 0.005·A. B negative and significant (t = −4.305, p < 0.01), implying higher popularisation among 25–34 associates with lower poverty. A not significant (t = 0.428, p = 0.669). Variance decomposition (covariance-based): For GDP, B explains 50.74% and A 49.26% of change; for poverty, B explains 54.12% and A 45.88%. Pearson correlations (p < 0.01 unless noted): 25–34 with 25–64: r = 0.931; 25–34 with GDP: r = 0.209; 25–64 with GDP: r = 0.279; 25–34 with poverty: r = −0.365; 25–64 with poverty: r = −0.329; GDP with poverty: r = −0.048 (ns). Quantile regression (GDP): Effect of A: not significant at τ = 0.1–0.3; significantly negative at τ = 0.4–0.6; significantly positive at τ ≥ 0.7, indicating that higher education level contributes positively at higher GDP quantiles but may impose short-run costs at middle quantiles. Effect of B: no effect at τ = 0.1; significantly positive at τ = 0.2–0.6; not significant at τ = 0.7; significantly negative at τ > 0.7, consistent with benefits in early to mid development and potential overexpansion at higher development levels. Quantile regression (poverty): A generally not significant across quantiles. B: not significant at τ = 0.1; significantly positive at τ = 0.2 (suggesting a short-run increase in poverty with expansion under certain conditions); not significant at τ = 0.3–0.4; significantly negative at τ ≥ 0.5, implying that popularisation reduces poverty at median and higher poverty quantiles.

Findings indicate that expanding the share of working-age adults with higher education is associated with higher GDP, addressing Q1 affirmatively. Popularisation among younger cohorts shows heterogeneous effects: it supports GDP growth in less- and mid-developed contexts but can be neutral or negative at higher development stages due to overcapacity and lagged returns, addressing Q2 with nuance. Regarding poverty, increasing popularisation among 25–34-year-olds is associated with lower poverty overall and especially at higher poverty quantiles, while the overall level of higher education among 25–64-year-olds does not significantly reduce poverty in the OECD context, addressing Q3 and Q4. These results align with human capital and endogenous growth theories, suggesting that quality, timing, and alignment of higher education expansion with labor market absorption and innovation capacity are critical. Overexpansion risks include credential inflation, mismatches, and resource waste, which can dampen economic returns. Policy emphasis should thus balance quantity with quality, equitable access, and innovation ecosystems to translate higher education expansion into sustainable growth and poverty reduction.

Using panel data from 38 countries (1995–2021), the study shows that higher working-age tertiary attainment is positively associated with GDP growth, and the popularisation of higher education contributes positively at early stages of development. However, the poverty-reducing effect is chiefly linked to the popularisation among young adults rather than the overall higher-educated share. The results underscore that simply expanding higher education is insufficient; targeted, equitable, and quality-focused expansion is needed to support both economic growth and poverty alleviation.

The analysis is limited to OECD member countries and available indicators for 1995–2021, which may constrain generalisability. The study relies on observational panel data and primarily correlational models (OLS and quantile regression); causal inference is limited. Inclusion of qualitative components could illuminate contextual cultural and historical mechanisms. Future research should employ methods suited to causal identification and broaden samples and variables to move from correlation to causation.