Question

In 1985, neither Florida nor Georgia had laws banning open alcohol containers in vehicle

passenger compartments. By 1990, Florida had passed such a law, but Georgia had not.

1. [19 Points] Suppose you can collect random samples of the driving-age population in both states, for 1985 and 1990. Let arrest be a binary variable equal to unity if a person was arrested for drunk driving during the year. Without controlling for any other factors, write down a linear probability model that allows you to test whether the open container law reduced the probability of being arrested for drunk driving. Which coefficient in your model measures the effect of the law?

2. [11 Points] Why might you want to control for other factors in the model? What might some of these factors be?

Hint:

You can apply the difference-in-difference approach here. Write down the linear probability model (linear regression model) with arrest as a dependent variable and explanatory variables: dummy variable for 1990 (2nd period dummy), dummy for Florida (treatment dummy), and interaction between the two dummy variables. The treatment effect will be measured by the coefficient on the interaction term.

Hints:

For part (B), explain your reasoning and include some of the factors. =====

Remarks: The above problems are largely based on panel data and pooled cross section data models. In addition to problems in Exercise 1 and Illustration 2, the following key terms might help in preparing for the material on Instrumental Variables Estimation.

Answer 1

Part A

In this case we need to construct a linear regression model. That is,

in this equation,

yr90 is the dummy variable representing the year 1990.

FL is the dummy variable representing Florida

So the variable should given the interest it describes the given law. If we restrict the given equation with 0 and 1 variable we can use the logit form. That is,

Part B

The dependent variable Arrest may be depends upon some other exogenous variables. Some of them are:

1. Gender

2. Education

3. Income

4. Police patrol

5. Fine imposed

These are factors affecting the arrest which could be reason for drunk driving. So it is noted that may be get some good coefficient to estimate the value.