GroupDifferences2011A

GroupDifferences2011A - Econ145.GroupDifferences2011A John...

Info iconThis preview shows pages 1–5. Sign up to view the full content.

View Full Document Right Arrow Icon
Econ145.GroupDifferences2011A John Pencavel GROUP WAGE AND EMPLOYMENT DIFFERENTIALS Women usually earn less than men and differences exist in the wages of workers from different ethnic or racial backgrounds. These wage differences can be measured in alternative ways and the various ways typically result in different values for the “wage gap”. Illustrate with gender. Let F i = 1 if individual i is a woman, = 0 if individual i is a man. (1) ln(w i ) = α 0 + α 1 F i + g 1 i (2) ln(w i ) = β 0 + β 1 F i + β 2 X i + 2 i where X i may represent control variables such as schooling or years since leaving school. If the values of the X s are different between men and women, then α 1 β 1 . (3) ln(w i ) = γ 0 + γ 1 X i + 3 i for women and ln(w i ) = γ 2 + γ 3 X i + 4 i for men Now even if the values of the X i ’s were the same for men and women, male- female wage differentials emerge if γ 1 γ 3 . There is no longer “a” male- female wage differential, but different wage differentials depending on the values of the X i ’s.
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
2 This is one possibility where γ 2 > γ 0 and γ 3 > γ 1 > 0 . Which of these three specifications (1), (2) or (3) is correct to report gender or racial wage differentials? If enough X variables are included (job assignment, detailed occupation, position in firm, etc.), gender & racial hourly earnings differentials are zero today, but should such variables be included? These variables themselves may reflect a narrower set of opportunities for certain groups. But do they represent different preferences? What are opportunities and what are preferences? Example: Consider the graph below that describes the relationship across occupations between the risk of fatality at work and the fraction of women workers. The unit of observation is the two-digit occupation and there are 43 occupations here. Examples of such occupations include farm workers, protective service occupations, motor vehicle operators, teachers at college & university, and computer equipment operators.
Background image of page 2
3 The horizontal axis measures the natural logarithm of the number of deaths at work per 100 full-time workers from 1992 to 1999. The central tendency of the number of deaths per 100 full-time workers is 0.002 (or 2 fatalities per 100,000 full-time workers or a fatality risk of 1 in 50,000 workers). The natural logarithm of 0.002 is -6.21. The vertical axis measures the fraction of total hours worked that are worked by women. These hours data describe individuals aged between 25 and 34 years who worked full time for the full year.
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
4 The least-squares regression line of y on x is the straight line drawn and it has a slope of -0.155 (that is significantly different from zero at less than 0.001). Why is there a negative relationship between the relative employment of women and fatality risks at work? Conjecture A: Employers are reluctant to hire women into risky jobs and this denies women the opportunity to earn the wage premium attaching to these jobs.
Background image of page 4
Image of page 5
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 21

GroupDifferences2011A - Econ145.GroupDifferences2011A John...

This preview shows document pages 1 - 5. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online