Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

ECO 395K

Oettinger

Option III M.A. Labor Economics

Spring 2023

Empirical Exercise #1 due at 2:00 PM on Wednesday, March 8

Go to IPUMS web page (www.ipums.org) and extract data from the 2021 American Community Survey (ACS) Public Use Microdata Sample. When creating your extract,     limit your sample to observations on individuals aged 25-54 who live in the state that I     have assigned for your analysis (shown on the last page below).  Your data extract should include the following IPUMS variables: STATEFIP, NCHILD, NCHLT5, AGE, SEX,     MARST, RACE, HISPAN, EDUC, OCC, CLASSWKR, WKSWORK1, UHRSWORK,   INCWAGE, and FTOTINC.

Read the IPUMS documentation to see what each of these variables measures and how each is coded. Next, create the following variables in Stata (or your preferred statistical software package, if different):

•   Age and age squared;

•   A dummy variable for Hispanic”;

•   Dummy variables for Black” and Asian”;

•   A dummy variable for currently married”;

•   Dummy variables for each of the following educational attainment categories: (i) Completed less than 12 years of schooling; (ii) Completed exactly 12 years of schooling; (iii) Completed 1-2 years of college; (iv) Completed exactly 4 years of college; and (v) Completed 5+ years of college;

•   Dummy variables for each of the 12 major occupation categories (i.e., “Management, Business, Financial Occupations”; “Computer, Engineering, and Science Occupations”; “Education, Legal, Community Service, Arts, and Media Occupations”; etc.);

•   A dummy variable for public sector worker”;

•   A dummy variable for wage worker last year” that equals 1 if the individual worked 1 or more weeks last year and was not self-employed and was not an unpaid worker, and equals 0 otherwise;

•   An annual hours worked” variable, computed as usual hours worked per week multiplied by the individual’s weeks worked last year;

•   A nonlabor income” variable, calculated as total family income minus own wage income;

•   An hourly wage” variable, calculated as wage income last year divided by annual hours worked;

•   The natural logs of annual hours worked, nonlabor income, and the hourly wage.

Finally, perform the following analyses using Stata:

a.   Produce detailed summary statistics that describe the sample distributions of weekly hours of work, annual hours of work, and hourly wages — separately for males and  females.  Examine your data to make sure that the variables in your data set take on  meaningful values for all of the observations in your analysis samples. (You may drop observations with missing data on any of the key variables, but you should note what fraction of observations are dropped for this reason.)

b.   Estimate labor supply models — one for males and one for females — using only

data on individuals who worked positive hours in the previous year in paid, wage employment (i.e., not self-employed).  Your measure of labor supply (the dependent variable) should be the natural log of annual hours worked.  Your explanatory variables should be the natural log of the hourly wage, the natural log of nonlabor income, and a vector of any additional explanatory variables that you believe are useful controls for variation across individuals in tastes for market work” at a given wage and nonlabor income.

c.   Interpret the estimates obtained in part (b), addressing the following questions:

•   Which estimated coefficient gives an estimate of the uncompensated (i.e., with nonlabor income held constant) wage elasticity of labor supply, denoted as 7w(u)? Do the signs of your estimated coefficients for 7w(u)  agree with the predictions of  static labor supply theory?

•   Which estimated coefficient gives an estimate of the (nonlabor) income elasticity of labor supply, denoted as 7V ?  Do the signs of your estimated coefficients for 7V agree with the predictions of static labor supply theory?

•   The Slutsky equation implies that the compensated (i.e., with utility held constant) wage elasticity of labor supply, denoted as 7w(c) , is related to 7w(u)  and 7V by the equation:

7w(c)  = 7w(u)  − () ∙ 7V  = 7w(u)  − () ∙ 7V ,

Use your coefficient estimates and the mean sample values of wage income and “full income” to obtain estimates of 7w(c) .  (Note: In calculating “full income” ,    assume that all individuals in your samples have an annual time endowment of T = 4000 hours.)  Do the signs of your estimates for 7w(c)  agree with the predictions of static labor supply theory?

•   What are the major differences in your estimates for men versus women?  Discuss some possible explanations for these differences.

•   Discuss why your estimated elasticities might be biased/inconsistent. In particular, discuss how measurement error, omitted variables, and/or sample selection may influence your estimates.

When you have completed these tasks, please turn in: (i) your Stata (or other software) code, (ii) the Stata (or other software) output for parts (a) and (b) of the problem, and (iii) your answers to part (c).

You should freely share information with your classmates any expertise that you acquire regarding topics such as: how to download the raw data, how the raw data are coded, appropriate code to create variables and estimate models in Stata, etc.

Assignment of Students to States for the Empirical Assignment

Student

State

PHILLIP AN

California

ARINDAM CHATTERJEE

Florida

SOO JEE CHOI

Georgia

ASHA CHRISTENSEN

Illinois

ADRIAN DURAN

Michigan

KE JIN

New Jersey

SHREYA KAMBLE

New York

ZHENGYI LIN

North Carolina

CHEN YEN LIU

Ohio

HAYATO MIYAZAKI

Pennsylvania

DANIEL MOONEY

Texas

JAE YOU

Virginia