The synergetic effect of economic complexity and governance on quality of life: policy thresholds

The study addresses whether and how a country's productive structure—captured by the economic complexity index (ECI)—improves quality of life (QoL), and the extent to which governance (GOV) modulates this relationship. Traditional development metrics such as GDP fail to capture qualitative aspects of development and productive capabilities. ECI better reflects embedded knowledge and diversification of production and exports and has been linked to multiple human development dimensions. However, prior findings on ECI’s effects on well-being are mixed, partly due to narrow well-being measures and overlooking institutional contexts. This study posits that governance can shape how economic complexity translates into social progress by influencing innovation, human capital, property rights, and efficient resource allocation. Focusing on developing countries—where economic structures are less diversified and institutions are often weaker—the paper investigates the synergistic and potentially nonlinear interactions among ECI, governance, and QoL. The hypotheses are: H1: ECI has a positive unconditional impact on QoL in developing countries. H2: The impact of ECI on QoL is contingent on governance settings (moderation). H3: There exist governance thresholds that condition the sign and magnitude of ECI’s impact on QoL.

The literature links structural change and productive sophistication to human development beyond income growth. Numerous studies report positive associations of ECI with health outcomes, human capital, employment opportunities, and more equitable income distribution, suggesting that complex economies enhance capabilities and social welfare. Conversely, some research warns of adverse or ambiguous effects: potential increases in inequality, offshoring of undesirable activities, polarization, environmental costs, and context-dependent outcomes where institutional frameworks and the type of diversification matter. A second strand explores governance as both a driver and outcome of economic complexity. Higher-quality institutions foster innovation, human capital, and R&D conducive to complexity; yet bureaucratic frictions and certain institutional constraints may hinder it. Governance also directly promotes social progress by expanding access to services and opportunities, with consistent evidence that better institutions improve QoL measures, including SPI and HDI, and reduce poverty and inequality. Few studies examine the tripartite dynamics among ECI, governance, and QoL or account for nonlinearity and moderation. Emerging work indicates that the relationship between complexity and well-being can depend on income levels, globalization, or governance quality. This paper contributes by jointly modeling ECI, governance, and QoL within a large developing-country panel, using SPI as a comprehensive QoL metric, testing moderation, and identifying policy-relevant governance thresholds.

Data: Unbalanced panel of 75 developing countries, 2011–2021 (825 observations). Outcome: Quality of life measured by the Social Progress Index (SPI), which aggregates three non-economic pillars (Basic Human Needs, Foundations of Wellbeing, Opportunity) into 60 indicators. Main regressor: Economic Complexity Index (ECI). Moderators: Six World Governance Indicators (WGI: voice and accountability, political stability, government effectiveness, regulatory quality, rule of law, control of corruption, scaled −2.5 to 2.5). To reduce collinearity and capture distinct institutional domains, principal component analysis (PCA) aggregates the six WGI into four governance dimensions: general (all six), political (voice and accountability, political stability), economic (regulatory quality, government effectiveness), and institutional (rule of law, control of corruption). Controls: Foreign direct investment inflows (% GDP), general government final consumption expenditure (% GDP), and ICT penetration (mobile + telephone). Empirical strategy: Two-step System GMM (SYS-GMM) is used to estimate dynamic models with lagged dependent variable, addressing endogeneity, persistence, omitted-variable bias, and measurement error. The model includes unconditional specifications (direct effects of ECI and governance) and conditional specifications with interaction terms (ECI × governance dimension or indicator) to estimate moderation. Following guidance on interaction models, marginal/net effects are computed, and policy thresholds are derived where the marginal effect of ECI on SPI becomes zero. Diagnostics: Arellano–Bond AR(1)/AR(2) tests confirm appropriate serial correlation structure; Hansen/Sargan tests assess instrument validity; instrument count kept below number of groups. Multicollinearity checks (VIF < 5) and correlation matrices corroborate model reliability. Robustness checks: (i) Alternative estimator (LIML) with Anderson–Rubin and Basmann tests of instruments; (ii) Alternative QoL measure (HDI) in SYS-GMM; (iii) Sensitivity analysis excluding top/bottom 2% outliers by ECI/QoL, re-estimating net effects and thresholds.

- Unconditional effects: ECI and governance are each positively and significantly associated with higher SPI across GMM specifications. Among individual governance indicators, government effectiveness and voice and accountability exhibit the strongest direct positive associations with SPI; corruption control is positive but relatively smaller. ICT is consistently positive for SPI, while FDI and government consumption often show negative associations. - Moderation: Governance significantly modulates the ECI–QoL relationship. The interaction terms for general, institutional, and economic governance are negative and significant, indicating that as governance increases, the marginal effect of ECI on SPI diminishes; political governance’s interaction is statistically insignificant. Despite negative interactions, the overall net effects remain positive at observed governance levels. - Policy thresholds: Thresholds at which the marginal effect of ECI on SPI turns zero are identified. Reported thresholds are: government effectiveness 0.8435; rule of law 1.846; control of corruption 1.717; general governance 5.59; institutional governance 3.14; economic governance 3.32. Below these thresholds, governance is necessary and sufficient for ECI to raise QoL; beyond them, complementary policies are needed to sustain positive impacts. - Robustness: Results hold across alternative estimators (LIML), using HDI as an alternative QoL measure, and after outlier removal. - Magnitudes (illustrative from SYS-GMM): SPI displays strong persistence (lag coefficient ~0.89). ECI’s direct coefficient is positive across models (e.g., 0.33–0.55). Positive coefficients for governance dimensions (e.g., general governance ~0.16) accompany negative ECI×governance interaction terms (e.g., −0.065 for general; −0.105 for institutional and economic). Control variables align with expectations (ICT positive; FDI/EXP often negative or insignificant).

The findings validate that more complex economies tend to achieve higher social progress, and that institutional quality is a key channel shaping how productive sophistication translates into QoL gains. Governance enhances the capacity to convert complex productive structures into social outcomes by supporting innovation, human capital, property rights, and efficient service delivery. However, the negative interaction terms indicate diminishing marginal returns of ECI on QoL as governance becomes tighter or more advanced, reflecting potential regulatory burdens or reduced incremental conversion efficiency at higher governance levels. This reconciles mixed prior evidence by showing that both complexity and institutions are beneficial on average, yet their synergy is subject to thresholds beyond which added governance, without complementary measures, can neutralize the incremental QoL payoff from complexity. The policy relevance is twofold: (i) strengthening governance—especially government effectiveness, rule of law, and control of corruption—is vital at earlier stages to unlock ECI’s social dividends; (ii) maintaining positive marginal returns beyond identified thresholds requires complementary policies in skills, innovation, competition, and market functioning to avoid overregulation or capability bottlenecks that blunt ECI’s QoL impacts. Consistency of results across specifications, indicators, and robustness checks underscores their generalizability within developing countries over the study period.

This paper contributes the first large-sample empirical assessment of governance’s moderating role in the ECI–QoL nexus for developing countries using SPI as a comprehensive QoL measure and SYS-GMM for causal inference. It documents that ECI and governance each unconditionally raise QoL, that governance moderates the ECI–QoL link with an overall positive net effect, and that policy-relevant thresholds exist for general, institutional, economic governance and key indicators (government effectiveness, rule of law, control of corruption). Policy implications are to: prioritize institutional strengthening to catalyze ECI’s social returns at early stages; focus on government effectiveness, rule of law, and anti-corruption; and deploy complementary policies—human capital development, innovation support, business environment reforms, trade facilitation, and targeted industrial upgrading—to sustain positive marginal effects beyond thresholds. Future research should undertake country-specific analyses, explore additional moderators (finance, ICT, FDI), assess nonlinearities across income groups, and extend time coverage to capture longer-run dynamics.

The analysis is limited by the relatively short observation window (2011–2021) and panel data constraints that necessitated interactive regressions for threshold estimation. Cross-country panel results may mask country-specific heterogeneity in institutions and structural transformation pathways. Data availability and reliance on export-based ECI may omit aspects of domestic capability formation not captured in trade patterns.