Course Overview: Applications

Overview

We have to review some core tools/ideas in statistics:

Basics of distributions, populations vs sample. Mean, variance, etc
Sample mean \(\rightarrow\) large sample theory \(\rightarrow\) population mean (estimation)
Hypothesis testing and confidence intervals
Linear models and linear regression

First, some research to motivate us.

“The Nature of Credit Constraints and Human Capital”, Lochner and Monge-Naranjo, American Economic Review (2011)
Theory:
- Incomplete markets can lead to underinvestment in human capital (agents cannot borrow against returns to investment).
- Controlling for returns, income/wealth should not predict investment unless borrowing constraints bind.
Question: Evidence of credit constraints in college enrollment decisions?
Data NLSY79 and NLSY 97 (two cohorts of young individuals)
Key idea: How do skills (measured by AFQT) and parental income predict college enrollment?

Paper: “Incentives Work: Getting Teachers to Come to School”, Duflo, Hanna, and Ryan, American Economic Review (2012)
Background: expanding access to primary education crucial in developing countries. Teacher absenteeism in India is a big problem.
Theory: providing incentives should increase attendance.
The Paper: evaluates a teacher incentive program by RCT.
Data: Attendance and student outcomes for 57 treatment schools, 56 control schools run by Seva Mandir (NGO)

Details on “treatment”:

Monitoring via tamper-proof cameras.
Financial incentives for teachers: \[ \text{Rupees per month} = 500 + 50\times\max\{0,D - 10\} \] where \(D\) is days attended per month.
Financial incentives or fear of punishment?
Notice: no incentive if: \[ \text{Days attended so far} + \text{Days left in month} \leq 10 \]