Abstract
This study explores the macroeconomic and social determinants of poverty in the United States and Canada from 1980 to 2023, using the poverty headcount ratio at $4.20/day. It employs the Autoregressive Distributed Lag (ARDL) model and Error Correction Model (ECM) to examine both short-run and long-run dynamics between poverty and six key variables: GDP per capita growth, income share of the bottom 20%, school enrollment, inflation, labor force participation, and government consumption expenditure. The results for the United States indicate strong long-run relationships, with income distribution, education, inflation, and labor force participation showing significant impacts on poverty. The ARDL model explains 95% of the variation in poverty, and the ECM confirms a stable adjustment toward long-run equilibrium. In contrast, the Canadian model explains 58% of the variation, with inflation, income share, and labor market variables showing notable effects, while education and government spending play more modest roles. These differences reflect how national welfare systems and institutional responses to macroeconomic pressures shape poverty outcomes. The comparative analysis highlights how differing institutional settings, Canada’s universal welfare state versus the United States' liberal model, mediate macroeconomic impacts on poverty. The study provides actionable insights for regional policy design, suggesting that enhancing income redistribution, improving educational access, and stabilizing inflation can significantly reduce poverty in liberal welfare regimes such as the U.S., while reaffirming the effectiveness of universalist policies in the Canadian context. These findings underscore the importance of redistributive mechanisms and investment in human capital in mitigating poverty.