Book Name: Stock and Watson's Introduction to Econometrics, 3^rd Updated Edition.

http://wps.aw.com/aw_stock_ie_3/178/45691/11696965.cw/index.html

Chapter 10 Empirical Exercise: Use the Seatbelts data set to complete the exercises below.

1. Run three regressions

a. Run a regression of the traffic fatality rate on seat belt usage, speed limits, the blood alcohol limit, the drinking age, log of per capita income and mean age using ordinary least squares estimation.

b. Fit the same model controlling for state fixed effects.

c. Fit the same model controlling for both state and year fixed effects.

2. Create a table with results for each regression.

3. Answer the following questions:

a. How do you interpret the coefficient on the seat belt usage variable in each of the three regressions?

b. Does controlling for state fixed effects affect your results? If so, what does the change in the coefficient on the seat belt usage variable tell you about the relationship between the unobserved state fixed effects, seat belt usage and traffic fatalities?

4. Estimate the model including a time trend instead of the year fixed effects. Which model is more appropriate? How do you interpret the time trend?

5. Run a regression of seat belt usage on speed limits, the blood alcohol limit, the drinking age, log of per capita income and mean age and the seat belt enforcement variables. Interpret the coefficients on the seat belt enforcement variables.

a. How does changing from secondary to primary enforcement affect traffic fatalities?

Chapter 11 Empirical Exercise:

1. Are the self-employed less likely to have health insurance than wage earners? Is this difference economically significant? Is it statistically significant?

2. In what other ways do the self-employed differ from wage earners? Discuss

3. Estimate a probit model examining the determinants of whether an individual has health insurance.

a. Estimate a basic model including only the self-employment variable and benchmark model.

b. Calculate and interpret the marginal effect of self-employment on the probability of having health insurance.

4. Estimate a probit model for health status, including marginal effects. What do your results indicate about the effect of self-employment on health? Estimate three models:

i. a simple model controlling only for self-employment,

ii. a second model adding the health insurance variable,

iii. a third model adding in other covariates which you believe belong in the model.

5. Repeat part (4) using a logit model in place of the probit model. Are there any significant differences?