Write the name of all group members, and your responses on a single document to submit.
Estimating the Returns to Schooling
For this section we will analyze data describing individual earnings, their education and experience profile, and other characteristics. We will dive into the economic relationship between education and wages in this example.
Q1a. Load wagedata.csv in your workstation and name the data frame wages main.
The variable wage represents monthly wages. We want to convert this value into 2020 dollar values. First create a variable CPI and set it equal to 0.4. Then, create a new variable wages main$wage2020 by dividing the wage by the CPI. Use the describe function and report the summary statistics for monthly wages in 2020 dollars. Include the mean, standard deviation, median, max, min, Q1 and Q3.
Q1b. Let’s create an hourly wage variable. The current variable hours is average hours worked per week. Create a variable annualhours equal to weekly hours multiplied by 52. Also, create a varaiable annualwages, noting that the current 2020 wage measure is in months. Next, create a variable hourlywages. Use the describe function and report the summary statistics for hourly wages. Note any cause for concern you may have with the data.
Q1c. The education variable educ reports the highest level of education attainment for each person in the sample. Count the number of people in the sample with each education level. Describe what you see in 2-3 sentences. Overall, does the data align with what you would expect?
Q1d. Produce five scatter plots and paste them here. Note that tenure is defined as years in the current job, and experience is defined as overall years of work experience.
• Education vs. IQ (IQ on x-axis)
• Hourly Wages vs. IQ (IQ on x-axis)
• Hourly wages vs education (education on x-axis)
• Hourly wages vs tenure (tenure on x-axis)
• Hourly wages vs experience (experience on x-axis)
Q1e. Describe the findings of your scatterplots in 2-3 paragraphs. Define discrete and continuous variables, then categorize each variable in this assignment as one or the other. Explain the relationship between education, IQ and wages. Then describe the what you find in terms of tenure, experience and wages. Any surprises on this visual inspection of the data?
Describe the estimate you get for bβ1. Based on your results, how much does the model predict your hourly wage to rise for one additional year of schooling completed? What makes this estimate difficult to interpret?
Q1g. Next estimate
Log(HourlyW agei) = β0 + β1Educi + βexperiencei + β3experience2 i + εi. Report the coefficient for bβ1. How did it change and why?
Q1h. Substitute the tenure variables for the experience variables in 2f and run the regression again. Do your results for bβ1 change much?
