Investing in Human Capital: The Neoclassical Barro-Becker Model and its Implications

This lecture delves into the neoclassical Barro-Becker model of human capital investment, a foundational framework in development economics. It explores how parents make decisions regarding their children's education, considering these investments as purely economic choices with associated costs and returns. The model is used to analyze the potential contribution of family-level decisions and credit constraints to the persistence of poverty and inequality, and to understand the role of public policy in this context. The discussion begins with a general utility function allowing for material and symbolic consumption, and complex intergenerational linkages, before simplifying to the more constrained Barro-Becker preferences which assume only material consumption and a single discount rate applied across generations.

A key aspect of the model is its production technology for human capital, which posits that parental education can reduce the cost of educating their children, though at a declining rate. Crucially, the model explicitly rules out non-convexities in skill production and strong complementarities between parental and child human capital, which are features of other models (like Galor-Zeira or Dasgupta-Ray) that generate persistent inequality or multiple steady states. The assumption of perfect credit markets, where individuals can borrow or lend at a gross interest rate, is central to the model's predictions, eliminating commitment and contracting problems between generations.

Under these neoclassical assumptions, the model yields a striking result: the optimal level of human capital investment is determined solely by the marginal return to education (higher wages for the child, lower future education costs for grandchildren) relative to its marginal cost, irrespective of parental preferences, patience, or income. This is because any individual parent can adjust their savings or borrowing to achieve their desired consumption path, leaving the education investment decision to be purely efficiency-driven. Consequently, in this simplified framework, everyone converges to the same human capital level in one generation, and large shocks to human capital are quickly sorted out.

The lecture also examines policy instruments within this framework, such as education subsidies and progressive income taxation. It highlights that lump-sum subsidies (e.g., free primary education) have no effect on marginal investment decisions, while marginal subsidies or taxes do. The model suggests that the most effective education policy is one that enhances the value of education in the labor market, rather than direct educational interventions. This framework, while powerful for analytical tractability, is acknowledged to be highly simplified, abstracting away from behavioral complexities like hyperbolic discounting or real-world credit market imperfections and intergenerational contracting issues.

"today is going to be a bit of a fall back to a much more neoclassical territory because I do want some time talking about the model that in a sense was the Workhorse model still is the Workhorse model of how to think about human capital investment decision"

"parents treat investment in human capital of themselves and their children as economic decisions exclusively which have costs and returns"

"if your parents are themselves educated and it becomes cheaper to educate the the child"

"the cost of education is convex so it becomes more and more expensive to acquire subsequent skill"

"the fundamental assumption that makes everything tick is that there are non-convexity so for example the return to primary education is zero and then there is a fixed cost and we you start getting return once you get at least to middle school or at least to college"

"you care only one generation below so you care about the utility not the consumption the utility of you of the future generation in turn they care about the utility of the future generation discounted Delta and in term the future generation care about the utility of the future generation and so on so far"

"the only thing that matters is the returns to education and how that compares to the cost and nothing else because a more patient parent will will invest less but that will all go through the how much they put in assets there is no reason to not do the efficient thing in in respect of with respect to education"

"in this model subsidizing primary of education is development this is a fixed this is lump sum doesn't matter to the choice"

"the best education policy is no education policy the best education policy is one where which which is in the labor market where it becomes valuable more valuable to get an education"