This statistical assignment gets your hands “dirty”with R.1  This statistical assignment is due in-class on March 22. You are also welcome to hand into my mailbox (6th floor, 19 West 4th Street) beforehand or up to noon on the 22nd . You may work in teams of 2 or 3 (or by yourself...) and hand in one assignment for the team. Make sure that all names of the team are on the assignment. 20% will be docked for each day late the assignment is handed in. Please staple your assignment together.

1    Dataset Details

The data comes from the Current Population Survery (CPS) conducted by the Bureau of Census.  The sample in your dataset has been restricted to age 25-59 non-hispanic men that worked at least one hour in the year. The dataset consists of 80,456 observations from 1976 to 1984. There are 5 variables in the dataset which are defined as:

❼ year: calendar year

❼ metarea:  CPS metropolitan area code.  There are 44 of these.  Please go to https:

//cps.ipums.org/cps-action/variables/METAREA#codes_section to get the code to metropolitan area link. The met code for Miami is 5000.

❼ age: age of individual

❼ wage: individual’s hourly wage

❼ edcode:  Individual’s education.  This variable is coded as:  1 =high school dropout,

2 =high school graduate, 3 = some college, 4 = college graduate.

General Instructions

Please hand in assignment as one (stapled) document. Any graphs should be legible and should be properly labelled.  For instance, if there are two lines please label both lines ap- propriately in the legend (e.g., as “Miami”and “Rest of USA”). Then answer the questions. Besides the graphs (which can take up a lot of room), the writing for this assignment should not be too onerous and should represent well under one page of writing.  This assignment follows our classroom discussion about the Mariel Boatlift. In this assignment, we will recre- ate Card (1990) but assign different control cities. You will see that the results can be quite sensitive to this choice.

Question 1

Graph the average wages for individuals in Miami vs. the rest of the USA by year. Briefly (2-3 sentences) describe the pattern that you see.

Question 2

Graph the average wages for individuals in Miami vs. the rest of the USA by year for high school dropout and non-high school dropouts (e.g., you should create two graphs).  Briefly (2-3 sentences) describe the pattern that you see.

Question 3

Define what you believe to be the most reasonable control group for your difference-  in-differences to answer the question “Does immigration affect the wages of locals?” Your control group must consist of at  least two cities.  State why you picked these cities as ‘control’ in 3-5 sentences.  [note:  please pick a different control group from that of Card  (1990). Your treatment group should obviously be Miami due to the Mariel Boatlift.]

Question 4

Graph the average wages for your treatment and control group by year.  Then, run an explicit difference-in-differences regression using city and year fixed effects. Please copy and paste the (one line) R code you used to do this in the document.2  Report the point estimate from the regression and interpret it in one sentence.

Question 5

Based on your results discuss whether you believe immigration affects the wages of locals. To do so, please discuss the internal validity of your design by evaluating the likelihood that the parallel trends assumption will hold given your results.  3-5 sentences should suffice. [note:  No right or wrong answer here and answer very well might depend on your control group definition.  To be clear, though, your evaluation of the parallel trends assumption should be specifically evaluated using your data and not just a general discussion of it].