Homework 4 Fixed Effect and First Difference Model Estimation

1. One of my master’s students worked on a paper examining the effect of tribal colleges on
various outcomes for American Indians. Tribal colleges were formed in different states over
the past 30 years, beginning with Dine College in 1968. The last tribal college formed was in
1997. Suppose I want to estimate the effects of tribal colleges on American Indian heart
disease.

i. What would be the “ideal” dataset for examining this question? Describe (1) who would
be included in the dataset, (2) who the treatment and control groups would be, and (3)
how the treatment would be assigned.

ii. What would your estimating equation be using this data? Interpret your coefficients of
interest.

iii. Of course, I do not have ideal data. Instead, I have a large nationally, representative
survey of adults in the United States that has been conducted over many years. The data
are repeated cross sections—individuals are not followed over time. I know the county in
which an adult lives, the colleges in that county, the individual’s race and ethnicity, the
individual’s education level, and information about the individual’s health. Give at least
two examples of potential treatment and control groups.

iv. Give the estimating equations that correspond to the treatment and control groups you
described. The notation in your equations (subscripts, etc) needs to be precise. Interpret
your coefficients of interest.

2. David Neumark and William Wascher published a study in 1992 of the effect of minimum
wages on teenage employment using a U.S. state panel. The paper used annual observations
for the years 1977-1989 and included all 50 states plus the District of Columbia. The
estimated equation is of the following type:

Eit = β0 + β1 (Mit /Wit ) + γ2D2i + • • • + γnD51i + δ2B2t + • • • + δTB13t + uit

where E is the employment to population ratio of teenagers, M is the nominal minimum

wage, and W is average wage in the state. In addition, other explanatory variables, such as the
prime-age male unemployment rate, and the teenage population share were included. D2-
D51 are the state dummies, B2-B13 are the time dummies.

(a) Briefly discuss the advantage of using panel data in this situation rather than pure cross-
sections or time series.

(b) Estimating the model by OLS but including only time dummy variables results in the
following output

2
Eit = β0 – 0.33 × (Mit /Wit ) + 0.35 × (SHY it ) – 1.53 × uram it ; R̄ =.20
(.08) (.28) (.13)
 where SHY is the proportion of teenagers in the population, and uram is the prime-age male
unemployment rate. Coefficients for the time fixed effects are not reported.
Numbers in parentheses are homoskedasticity-only standard errors.

Comment on the above results. Are the coefficients statistically significant? Since these are
level regressions, how would you calculate elasticities?

(c) Adding state fixed effects changes the above equation as follows:

2
Eit = β0 + 0.07 × (Mit /Wit ) – 0.19 × (SHY it ) – 0.54 × uram it ; R̄ = 0.69
(0.10) (0.22) (0.11)
Compare the two results. Why would the inclusion of state dummy variables change the
coefficients in this way?

2
(d) The significance of each coefficient decreased, yet R̄ increased. How is that possible?
What does this result tell you about testing the hypothesis that all of the state fixed effects can be
restricted to have the same coefficient? How would you test for such a hypothesis?

3. Do Computer exercises in Wooldridge 13.6 and 13.12

One thing to note: the Wooldridge data sets for these problems have the “change” variables
already constructed, usually labeled dXXX. You may find it useful to replicate the
construction of these variables so that you understand how the models work.