This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Department of Economics University of California, Berkeley Spring 2006 Professor Woroch Problem Set 3 Suggested Answers The Excel spreadsheet wages.xls contains data from a survey of 526 U.S. workers in 1976. The data include information on each workers age, hourly wage, years working for their current employer (tenure), years of education, gender, and marital status. 1 In parts a through e , use regression analysis to investigate the relationship between wages and worker age. Construct an Excel table similar to Table 5.2 in Stock &amp; Watson (page 182). Re- port the regression coefficients, robust standard errors, and R 2 for each regression in a separate column. a. Estimate the following model using linear regression: Y i = + 1 X 1 i + i (1) where Y is the wage, X 1 is the age, and i indexes the i = 1 ... 526 workers. Do older workers earn more? Test the null hypothesis that 1 = 0. How much does the average thirty-year old make? The average sixty-year old? See Table I. The age coefficient is . 058 , which suggests that older workers may earn more than younger workers. The coefficient is statistically different than zero at the five percent level because . 058- . 119 = 4 . 85 &gt; 1 . 96 . The average wage of a thirty-year old equals 4 . 182 + 0 . 058 * 30 = 5 . 92 . The average wage of a sixty-year old equals 4 . 182 + 0 . 058 * 60 = 7 . 66 . b. It is possible that workers with more experience at their current job are more productive and earn higher wages. What is the correlation coefficient between job tenure and age? Based on this correlation coefficient, do you think the estimator for 1 from part a is biased up or down? Why? The correlation coefficient between tenure and age positive (0.0508). If tenure affects wages (as we suspect it does) then the estimate of 1 in part a is too large (biased upwards)....
View Full Document
- Spring '08