Abstract
<title>Abstract</title> <p>Infant mortality remains unequally distributed across Latin America and the Caribbean. We examined whether within-country changes in poverty, income inequality, education, government health expenditure, and universal health coverage (UHC) were associated with infant mortality. We created an auditable panel from a frozen World Bank API snapshot for 2000–2023. From 7,056 indicator-country-year records, we constructed 1,008 country-years; the primary complete-case sample included 290 country-years from 17 countries with at least two informative observations. We estimated log-linear country and year fixed-effects models using country-clustered CR1 intervals, a t(16) reference distribution, restricted Webb-weight wild cluster bootstrap-t inference, and cluster-jackknife checks. No predictor showed robust evidence of association in the primary model. A one-point UHC increase corresponded to 1.51% lower infant mortality (95% CI, -3.51% to 0.54%; bootstrap p = 0.316), and a one-percentage-point increase in government health expenditure corresponded to 2.80% lower infant mortality (95% CI, -7.18% to 1.79%; bootstrap p = 0.219). First-difference models produced stronger inverse estimates, but these post hoc results were unstable across specifications. Ecological measurement, missingness, unbalanced coverage, and few countries constrain inference. The findings do not establish causal or policy effects.</p>