Diffusing knowledge: Reoptimizing redistribution for growth and inequality

Abstract

We examine the effects of income redistribution on output and welfare by generalizing Bénabou’s (2002) economy to incorporate two new elements: physical capital and social knowledge externalities. Income inequality, sustained by unequal privileges to private education and parental networking in the absence of a credit market, interacts with these added features. These interactions theoretically link four seemingly unrelated global trends: increased capital share in production due to automation, rising income inequality, slower knowledge diffusion, and declining productivity growth, offering new insights into growth- and welfare-enhancing redistributive policies. Automation, reflected in the growing importance of physical capital in production, and capital-ownership concentration worsen unequal educational opportunities and, in turn, income inequality, which slows productivity growth through two underexplored channels. First, we provide empirical support for the idea that higher inequality hampers knowledge diffusion to lower the economy’s social knowledge stock, thereby hindering children’s learning from the existing know-how through the knowledge externality channel. Second, greater heterogeneity in knowledge absorption due to more unequal access to education reduces average human capital accumulation because of diminishing returns to investment. Progressive redistribution helps counteract these adverse effects, pushing the economy’s productivity frontier outward, especially for countries with lower social cohesion, going beyond Bénabou’s (2002) finding of reduced resource misallocation. Optimal redistribution balances these benefits against potential distortions to labor supply and savings. Simulations using OECD data show sizable gains in output, aggregate efficiency (as defined in Bénabou, 2002), and welfare from moving towards optimal redistributive rates, though with varying effects across countries.