Problem 1) Come up with instrumental variables by yourself- First, briefly ecplain the causal question of interest, and then explain why the suggested IV(s) is valid, (You can forget about the data availability to answer this question, The purpose of this question is to check your understanding about the concept

Problem 1) Come up with instrumental variables by yourself- First, briefly ecplain the causal question of interest, and then explain why the suggested IV(s) is valid, (You can forget about the data availability to answer this question, The purpose of this question is to check your understanding about the concept.)

Problem 2.

Take any data set you introduced in Problem Set I (or you can use the wagepan.dta data from http: www.stata.com texts/eacsap ). Pick any fixed effect regression. In the following you are asked to run it in three different ways.

i) Use xtreg, fe command in stata (using straight standard errors, and clustering by person) ii) Do the fixed effects regression, i.e- regress(Yit — % ) on (Xlt — Si), You can construct the variable by using the egen command with by. Get the standard errors two ways-the standard way and clustering by person.

iii) Also first difference the data and get standard errors with and without the cluster command How do all of these results compare? What happends if you only use two periods?

Problem 3.

Now take the data set jtrainl (from http: www.stata.com texts eacsap/) This data set has data on firms and the amount of job training they get.

A) Only use the data from 1987 and 1988. Construct the difference in differences estimator in two different ways:

i) Construct the 4 means (control,treatmentx before,after) ii) Run the regression hrsenup;t + + — 1988) + ßgEi + where Ei is a dummy variable for being a treatment (i.e. someone who would receive the grant in 1988).

iii) Run the fixed effect regression: hrsentpit + grantit + (year 1988) + Do you get exactly the same answer? Why or why not?

B) Now include a firm specific time trend in the model in two different ways:

i) Use the xi command (something like xi: reg y x i.fcode*year) ii) For each firm, run a regression of x and y on an intercept and a time trend, take the residuals and run them on eaéh other (you could again use egen with by)

Problem 4.

Now use the data set

 that you can get from

You can read it into stata using the comand:

infile coll merit male black &ian ,ymr state chst using regm.raw.txt Now the difference in difference model 4 different ways:

1. Standard regression using all data (construct standard errors 3 ways, robust, cluster by state year, cluster by state)
2. Standard regression using all data but weighted so that all states get the same weight
3. Now take the mean of all variable by state x year and run the differences-in-differences regression

(robust se, and clustering by states)

4. Do the same as in c), but weight by state so it looks like the population, How does this all compare?