Questions: Empirical Methods in Labor Economics

Question 1

Card and Krueger's (1994) study of New Jersey's minimum wage increase used which identification strategy?

• A. Instrumental variables, using political ideology as an instrument for the minimum wage
• B. Regression discontinuity, using the minimum wage threshold as a cutoff
• C. Difference-in-differences, comparing employment changes in New Jersey fast-food restaurants (treatment) to Pennsylvania fast-food restaurants (control) before and after the increase
• D. Randomized controlled trial, with randomly selected restaurants receiving higher minimum wages

Answer: (C)

Card and Krueger used a difference-in-differences (DiD) design: they surveyed fast-food restaurants in New Jersey and eastern Pennsylvania before and after New Jersey raised its minimum wage from $4.25 to $5.05 in 1992. Pennsylvania, which did not raise its minimum wage, served as the control group. The key assumption (parallel trends) is that employment in New Jersey and Pennsylvania restaurants would have changed similarly absent the minimum wage increase. They found no evidence that the increase reduced employment — a finding that challenged the standard competitive model and was consistent with monopsony.

Question 2

The 'credibility revolution' in labor economics refers to the development of new economic theories about how labor markets function.

• T. True
• F. False

Answer: False

The credibility revolution is about empirical methodology, not theory. It refers to the shift — beginning in the late 1980s and accelerating in the 1990s — from relying on structural models and cross-sectional regressions (which require strong, often untestable assumptions) toward quasi-experimental methods (DiD, RD, IV, natural experiments) that provide more credible identification of causal effects. The term was coined by Angrist and Pischke. The revolution changed what counts as convincing empirical evidence in labor economics: instead of a well-specified structural model, researchers now need a credible source of exogenous variation — either experimental or quasi-experimental — to support causal claims.

Question 3

Why did Angrist and Krueger (1991) use quarter of birth as an instrument for years of schooling in their study of returns to education, and what makes this a valid instrument?

Model answer: Compulsory schooling laws require students to stay in school until a certain age (typically 16). Because school entry is determined by calendar-year cutoffs, students born earlier in the year reach the compulsory schooling age earlier in their academic career and can legally drop out with less total schooling than those born later. Quarter of birth thus affects years of schooling (relevance) through the interaction of compulsory schooling laws and school entry dates. For validity (the exclusion restriction), quarter of birth must affect earnings only through its effect on schooling — not through any direct channel. Angrist and Krueger argued that birth timing is essentially random with respect to ability and other determinants of earnings, making it a plausible instrument.

This study is a canonical example of the instrumental variables approach in labor economics. OLS estimates of returns to education are biased because ability affects both schooling and earnings (omitted variable bias). The IV strategy uses quarter of birth to isolate variation in schooling that is plausibly exogenous — driven by the accident of birth timing interacting with institutional rules, not by individual choices. The estimated return to education was roughly 7-8% per year, similar to OLS, suggesting that ability bias in the returns-to-education literature may be smaller than feared. However, the instrument has been criticized for being weak (quarter of birth explains very little variation in schooling) and for potential violations of the exclusion restriction (birth season may correlate with family background).