Economics 310 Professor Wallace Fall 2013 Homework #2 Instructions: In this homework assignment you will be asked to use the CPS ORG data that we have been working with in class (and on the last homework) to verify some of the implications of the human capital model. In constructing your answers to the questions below please reference appropriately constructed figures, tables, and descriptive statistics. 1. Start a Do-file in which you load the STATA date file cpsorg.dta. In the first several command lines of the Do-file you should eliminate all observations associated with 1979. For a remainder of the analysis you will work with only data from 1987. 2. In your Do-file use generate and replace to create a new educational attainment variable called ed_level which aggregates across the categories of the variable ed so that values of ed below the value of 12 correspond to a value of ed_level equal to 1, a value of ed equal to 12 corresponds to a value of ed_level equal to 2, values of ed greater than 12 and smaller than 16 correspond to values of ed_level equal to 3, and values ed greater than or equal to 16 correspond to a value of ed_level equal to 4. (Hint: This ed_level variable is the same as the new_ed variable you were asked to create in your last homework assignment) 3. For both men and women separately use descriptive statistics and/or graphical techniques to address the extent to which wages rise with education level (ed_level). Does the relationship between education level and wage appear to stronger for men or for women? Explain. 4. The data file cpsorg.dta contains a variable referenced as ex. This variable, known as potential experience, measures the length of time that individual have been out of school and (potentially) in working in the labor force. Create a new variable called ex_level which equals 0 if potential experience less or equal to than 10, 10 if potential experience is greater than 10 and smaller than or equal to 20, 20 if potential experience is greater than 30 and smaller than or equal to 40, 40 if potential experience is greater than 40 and smaller than or equal to 50, and 50 if potential experience is greater than 50 and smaller than or equal to 60. (Hint: For a shortcut see the Do-file from September 10 and look carefully at the way the frequency distribution of wage was created. An alternative would be to use one gen and 5 replace statements to create the variable) 5. An Experience-wage profile is a plot of mean wages by experience levels (vertical axis) against experience level. Make two line graphs. The first graph should show plots of mean wages by ex_level for men in all four education levels (separately) against the ex_level on the horizontal axis (i.e., there should be 4 profiles ploted on the same graph). The line graph second should show the same information for women. 6. Common features of experience-wage profiles from most data sources are: i. A convex shape whereby wages rise rapidly at low levels of experience, level off, and then decline at high levels of experience; ii. Profiles that fan out with experience, so that education related differences in earnings are larger at higher levels of experience; iii. Profiles for men tend to be more convex and fan out more than those for women.To what extent to the wage experience profiles that you created share these features? 7. For the rest of the assignment you will not need to access data pertaining to Hispanics (race=3). In your Do-file drop all Hispanics. 8. Produce a descriptive statistics or a figure(s)-graph(s) that document the differences mean wages between black and white men and women (separately). Describe these differences. 9. For the rest of the assignment you will not need to access data pertaining to women. In your Do-file drop all women. 10. One possible explanation for racial differences in mean wages is rooted in human capital theory. To the extent that labor market experience and education are associated with higher wage rates, lower levels of experience and education among blacks may result in lower mean wage rates. Provide a comparisons of the distribution of experience and education levels for black and white men. Based on a comparisons of these distributions, should black men have lower mean wages than white men? 11. In order for the human capital theory to fully explain white-black differences mean wages it would have to be the case that white-black differences in mean wages within fairly tightly defined skill (human capital) categories are small or non-existent. Within education levels (ed_level) are there differences between mean white and black wages? If so, are these differences consistent with the human capital theory?
View Full Document
Unlocking...