This was developed using STATA as part of Econometrics final project. See PDF for full report.
A few months ago, I faced the difficult decision of enrolling into graduate school at USC or starting my full-time job at Ernst & Young (E&Y). I was very grateful to have these choices but felt that either decision would determine my career for at least the next five years. This prompted me to question, which option would result to having the higher yearly compensation and to what extent would other factors affect my salary? Hence, I developed an econometric model that focuses on compensation salary in the United States versus age and other influential variables such as race, gender, education, and region.
I used 2019 public U.S. Census Data on Current Population Survey (CPS) Annual Social and Economic (ASEC) Supplement. The data focuses on the person record type and I renamed many variables from the provided data dictionary to help make the model more understandable (See STATA code in Appendix for variables renamed). Because the data was from the U.S. Census Survey, many important explanatory variables such as tenure-level, years in the work force, or cost of living,
Age is positively correlated with salary; however, age starts to have a diminishing effect on salary after the age of ~53 years old. I have determined that gender, race, educational degree, and the region of work are all significantly correlated with salary. Workers with a graduate degree are also expected to make more than those who hold only a bachelor’s degree. These results largely appear to be consistent with findings from other research studies. Since this is also an observational study, I can only conclude correlation rather than causation. Moving forward, an inclusion of additional explanatory variables (i.e. years in the work force, tenure level, etc.) would further strengthen the model.