Econ 309:  Econometrics

Link to slides

Lecture Notes on Econometrics

For interest:  

• • Causality and Milton Friedman's thermostat

  • Doing treatment and control groups is very hard, even in medicine.  (NPR, March 8, 2010)

• Stronger placebos?  (Wired, Aug. 28, 2009)

  • Identifying event:  Young people reaching the age limit on their parents' health insurance policies.  "The Effect of Health Insurance on the Use of Medical Services", by Anderson, Dobkin, and Gross (NBER 15823, March 2010)

  • Does Drinking Impair College Performance?  Evidence from a Regression Discontinuity Approach", by Carrell, Hoekstra, and West (NBER 16330, September 2010)

  • Does immigration lower wages?  Card versus Borjas.  (New Republic, August 6, 2010)

  • Identifying event:  An exogenous change in the rules.  "Regulating Abortion:  Impact on Patients and Providers in Texas", by Coleman & Joyce (NBER 15825, March 2010)  

  • What effect does Teach for America have on its members?  (New York Times, Jan. 3, 2010)

  • How good are Charter Schools?  (The New Republic, October 5th, 2009)

  • How much is it worth to get a college degree?  $800,000?  $300,000?  Much less since unconditional estimates fail to control for ability, which correlated with the probability of getting a college degree?  (Wall Street Journal, Feb. 2, 2010)

  • Does vigorous exercise bring better grades?  (New York Times, June 3, 2010)

  • Did the stimulus bill have any effect in the places where it was spent?  (Eddie Glaeser, New York Times, June 1, 2010)

  • "The Credibility Revolution in Empirical Economics:  How Better Research Design is Taking the Con out of Economics", Joshua Angrist and Jorn-Steffen Pischke, NBER 15794

  • Awesome identifying event for a Kellogg Marketing professor (Slate, May 20, 2011)

For amusement:

• • Here is a method that was not tested via careful comparison of treatment and control groups

  • The eternal statistics question

• The Null Hypothesis

• Jelly beans

• Time series causality

• Sperm count and cola consumption

• Freakonomist research

  • The Journal of Spurious Correlations

  • Post hoc, ergo propter hoc (Huh?  Dilbert)