EC507, Spring 2021, Prof. Jordi Jaumandreu

Midterm 1, February 25

        Part B. Practical exercise (40 min, 50 pts.)

        The data set naics2011.dta reports several variables for US Manufacturing broken down in 473 6-digit industries for the year 2011. The variables in the file are: NAICS (6-digit), EMP (employment in thousands), PAY (payroll in million $), PRODE (production workers in thousands), PRODH (production worker hours in thousands), VSHIP (value of sales in million $), MATCOST (cost of materials in million $) and CAP (capital stock in million $).

        Using Stata (or equivalent software) perform the following numbered questions, writing the answers and results in a sheet. Please use three decimal digits everywhere (except when the number is an integer):

        1. (8 pts) Generate the variables wage=PAY/EMP, bluep=PRODE/EMP, hours=PRODH/PRODE, pcm=(VSHIP-PAY-MATCOST)/VSHIP, kperw=CAP/EMP. (bluep stands for propor-tion of bue collars, pcm for price cost margin and kperw for capital per worker). Report the means and standard deviations. Write a line giving the meaning of each variable.

        2. (8 pts.) Categorize the variable wage in four intervals (low, medium-low, medium-high and high) using the quartiles. Report the quartiles and the number of observations in each category.

        3. (5 pts.) Provide the average of variable bluep for each category of wage. Are wage and bluep related? Why is so?

        4. (5 pts.) Construct a new variable lwage taking the natural logarithm of wage. Sketch and comment the difference between the histograms of both variables.

        5. (6 pts.) Now compute the histogram for variable hours. Compare with any of the histograms for wages (obviate the fact that wage and hours refer to sightly different samples of workers: EMP and PRODE). You may want to compute the coeffients of variation to support your asessment. What would you answer to an economist that suggests that the dispersion in wages across industries may be due to the dispersion of hours of work?

        6. (7 pts.) Compute the log of variable kperw and draw a scatter relating lwage and lkperw. How is the relationship between the two variables? Compute the correlation coefficient and say if it reinforces your visual conclusion.

        7. (6 pts.) Compute the average of pcm for each category of wage. Can the average be increasing in wage when PAY shows up subtracting the numerator of pcm? Interpret the result.

        8. (5 pts) Summarize in a sentence you findings about the association of wages with other variables.