Growth for Whom? State Capacity, Consumption Exposure, and the Limits of Instrumental Identification in Global Social Outcomes

Economic growth, administrative capability, and the distribution of consumption are potentially complementary mechanisms of human development, but their causal effects cannot be inferred from the significance of country-panel coefficients alone. This study links a 1996–2024 harmonized macro-social panel to the Quality of Government Standard Time Series dataset and to the World Bank Global Consumption Database 2010, including the latter’s original household survey provenance. We assess two explicit instrumental-variable designs: historical settler mortality for perceived government effectiveness and an interactionbetween net oil exports per capita measured in 1995 and subsequent global oil-price changes for growth. The latter is evaluated over matched prospective outcomes using controls dated before the shock, country and year effects, country-clustered and two-way score variance, weak-identification-robust Anderson–Rubin regions, exposure-year substitutions, crisis exclusions, and cross-fitted partially linear IV scores. The oil interaction is weak in each main sample: for future extreme poverty, poverty at the international $4.20 line, child mortality, and secondary enrollment, the country-clustered squared first-stage t statistics are respectively 0.49, 0.49, 0.24, and 0.08. Two-way-clustered Anderson–Rubin tests do not yield bounded effects over the prespecified exploratory grid. Settler mortality has measurable cross-sectional relevance for effectiveness but faces historical and ecological channels that prevent credible exclusion for present-day welfare. The consumption data show that the lowest spending group devotes approximately 58.1 percent of expenditures to food across 92 observed countries, providing an economically meaningful direct price-to-welfare pathway. Neither instrumental candidate supports an independently identified causal ranking of growth versus state capacity. The results clarify the empirical requirements for studying the social conversion of economic expansion, rather than equating growth with development.