# The Inframarginal Returns to College | Hoover Institution

Source: https://www.youtube.com/watch?v=nDeOKQmBKZU
Recap page: https://rapidrecap.app/video/nDeOKQmBKZU
Generated: 2026-02-10T08:32:34.863+00:00

---
## Quick Overview

The research models the inframarginal returns to college by developing an equilibrium model with rich occupational and individual heterogeneity, concluding that policies dramatically reducing college costs will have minimal effects on completion or individual welfare because tuition represents a small fraction of overall college costs, which are dominated by effort, time, and disutility of studying.

**Key Points:**
- Policies that significantly lower the cost of college education will yield very little effect on college completion or individual welfare because these broad policies often transfer resources to higher-income individuals who already planned to attend.
- The model replicates marginal income returns to college found in prior research, but actual utility gains are small because prior studies neglect substantial costs beyond tuition, such as effort and time investment.
- Tuition is estimated to be a small part of overall college costs; for example, the average total cost estimated might be around $300,000, while net tuition averages about $60,000 over four years.
- The model incorporates rich occupational heterogeneity, individual heterogeneity (absolute and comparative advantage), and costs associated with college, including effort (human capital investment) and time.
- The speaker notes that the share of men aged 25 to 34 with a bachelor's degree showed very little movement between 1975 and 2000 but then shot up by about 10 percentage points starting in the mid-2000s.
- The model is structured as a static, repeated cross-section model for 25-year-olds, where time variation is introduced by exogenous shifts in occupational productivities and returns to human capital.
- The utility function includes disutility from effort (S) in accumulating human capital, and this effort cost helps explain why lawyers are paid more than teachers on average.

**Context:** The presentation introduces a brand new paper analyzing the inframarginal returns to college, contrasting its approach with existing applied micro research that focuses only on individuals right on the margin of the college decision, often using RD specifications to estimate large income returns. The speaker aims to develop an equilibrium model incorporating occupational and individual heterogeneity to better understand how people far away from these margins respond to policies intended to expand college access, such as making college free.

## Detailed Analysis

The core of the presentation is the development of an equilibrium model featuring rich occupational heterogeneity (e.g., carpentry vs. lawyering) and individual heterogeneity, defined by absolute advantage (alpha) and comparative advantage (fee) in different sectors. Individuals choose one of three education levels: no college, some college (associates or dropouts), or a bachelor's degree or more. Wages are determined endogenously based on occupation productivity, comparative advantage, and accumulated human capital (which depends on effort, S). A key finding anticipated is that policies targeting tuition reduction will have limited impact because tuition is a small component of total college costs; the model suggests total costs might average around $300,000, dwarfing the $60,000 tuition estimate. Utility is reduced by the disutility of effort (S) and idiosyncratic costs (epsilon) associated with college, which load up factors like opportunity cost, psychic cost, and information cost. The model, while static in structure (ignoring cohort dynamics for now), allows exogenous shocks to occupational parameters (productivity beta, human capital returns gamma, and average costs kappa) to drive sorting, and the speaker plans to use microdata moments, including higher-order wage moments, to discipline the distributional parameters of the model.

### Research Goal

- Assessing returns to college for individuals away from the margins of the decision
- Developing an equilibrium model with rich occupational and individual heterogeneity
- Goal is to better understand potential policy proposals for expanding college access

### Model Primitives and Heterogeneity

- Individuals possess absolute advantage (alpha) and comparative advantage (fee) in occupations
- Costs are idiosyncratic (epsilon) and average (kappa, including tuition and opportunity cost)
- Occupations differ by productivity (beta) and returns to human capital (gamma)

### Utility and Cost Structure

- Utility depends on wage (endogenous) and disutility from effort (S) in accumulating human capital
- Costs (kappa) include tuition, opportunity cost, and psychic costs, which are estimated to be significantly larger than tuition
- Model suggests marginal individuals gain income but offset utility due to these costs

### Model Limitations and Future Work

- The current model is static, treating time as repeated cross-sections of 25-year-olds, lacking cohort structure and extensive labor supply margins
- The speaker intends to use the model to explain aggregate trends, such as the recent sharp increase in bachelor's attainment since the mid-2000s, and explore policies like the GI Bill retrospectively

### Wage Determination

- Wages are log-linear functions of occupation productivity (beta), comparative advantage (sorting), human capital investment (scaled by returns gamma), and an aggregate productivity shifter
- Sorting effects are included, meaning increased supply of workers in an occupation lowers the average wage in that occupation