Colleen P Cahill Econometrics II Presentation

355 views

Published on

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
355
On SlideShare
0
From Embeds
0
Number of Embeds
3
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Colleen P Cahill Econometrics II Presentation

  1. 1. A Study of How theReturn to Education and the Gender Gap Have Changed: 2000-2010 Colleen Cahill University of South Florida Econometrics II / ECO 6425 November 14, 2011 Dr. Beom S. Lee
  2. 2.  Equal Pay Act of 1963 ◦ 1960: Wage disparity approximately 60% ◦ 2010: Wage disparity approximately 77% General Expectation that More Education and Experience Equals Higher Income ◦ Women surpassed men in educational attainment in the 1990’s ◦ The ranking of U.S. education compared to other OECD countries has fallen in the past decadeMotivation for the Study
  3. 3.  Factors that Contribute to the Wage Gap ◦ Personal Choices ◦ Male-Female Differences in Skills ◦ Differences in the Treatment of Equally Qualified Men and Women Disparity in the Return to Education Potential Problems with the Use of the Basic Wage EquationIn the Literature
  4. 4.  Personal Choices Regarding Labor Force Participation ◦ Having two or more children  Human capital depreciation  Less work force experience ◦ “Sexist Family Decision Rules”  Housework time  Wives who follow their husbands to new geographic locations ◦ Choice of OccupationIn the Literature
  5. 5.  Male-Female Differences in Skills ◦ Human capital is rewarded differently for men and women  Perceived or actual differences in the quality of capital accumulated including years of schooling and experience  Initial increases in women’s labor force were associated with a declining skill level of employed women relative to menIn the Literature
  6. 6.  Differences in the Treatment of Equally Qualified Men and Women ◦ Wage growth among young women found to be less than that of young men ◦ Employers with imperfect information about potential employees  Use sex to predict future work commitment and the likelihood that a worker will quit or take time off  Women must have greater ability to be promoted  Women hold a lower proportion of high paying jobsIn the Literature
  7. 7.  Disparity in the Return to Education ◦ Individual variations in human capital imply differences in earnings power ◦ The return to education increased sharply in the 1980s  Shift to service-oriented production rather than industrialized production  The return to a college education increased more for men than for women ◦ Family decisions to invest in men’s education over women’sIn the Literature
  8. 8.  Potential Problems with the Basic Wage Equation ◦ Summary measures may be inadequate controls for work experience ◦ Failure to control for ability may lead to an upward bias in the return to schooling ◦ Endogeneity of experience and tenure controls when number of children is included in the equationIn the Literature
  9. 9.  Collected from the Current Population Survey ◦ Cross-sectional Data  Primary source of information on the labor force characteristics of the civilian non-institutional U.S. population ◦ U.S. Census Bureau for the Bureau of Labor Statistics ◦ DataFerrett Tool ◦ March 2000 – 2010 SurveysThe Data
  10. 10. The Data
  11. 11. The Data
  12. 12. The Data
  13. 13. The Data
  14. 14. The Data
  15. 15. The Data
  16. 16. The Data
  17. 17. The Data
  18. 18. The Data
  19. 19. The Models: The Estimation Method: • Ordinary Least Squares Estimation • Robust Standard Errors • StataThe Models and Methods
  20. 20.  The Return to Education ◦ Initial Samples: 2000 Return Approximately 12% ◦ Revised Samples: 2000 Return Approximately 6% ◦ Change in the Return to Education: Less than 1 percentage point for any period  Initial Samples: Statistically different from 0 at less than a 10% significance level in periods 2008-2010  Revised Samples: Statistically different from 0 at less than a 10% significance level in 2007 where it falls slightlyResults and Interpretations
  21. 21.  The Gender Gap ◦ Initial Samples: 2000 Gender Gap Approximately 42% in ln(wages) and 34% in weekly wages ◦ Revised Samples: 2000 Gender Gap Approximately 25% in ln(wages) and 22% in weekly wages ◦ Change in the Gender Gap: Approximately 5 percentage points from 2000-2010  Initial Samples: Shown to have fallen against the positive one sided alternative at less than a 10% significance level in periods 2001,2003,2005-2007, 2009-2010  Revised Samples: Shown to have fallen against the positive one sided alternative at less than a 10% significance level in periods 2003-2010Results and Interpretations
  22. 22. Results and Interpretations
  23. 23. Results and Interpretations
  24. 24.  For Each Year 2001-2010 ◦ Generated 10 Random Error Terms, Normally Distributed with Mean 0 and Variance the Square of the Mean Standard Error ◦ Generated 10 New Dependent Variables Using the Estimated Coefficients, the Existing Data for the Independent Variables and the Random Error Terms ◦ Regressed the New Dependent Variables on the Existing Data for the Independent Variables ◦ Calculated the Mean of the Estimated Coefficients for Each Year ◦ Compared the Mean Coefficient Estimates to the Original Coefficient EstimatesMonte Carlo Simulation
  25. 25. Monte Carlo Simulation
  26. 26. Monte Carlo Simulation
  27. 27.  The estimated return to education is practically small and primarily insignificant The actual gender gap is a number between those estimated ◦ The estimations using the larger, less controlled samples estimate a larger gap ◦ The estimations using the smaller, more controlled samples estimate a smaller gap For the purposes of this study, a basic wage equation seems adequate, although better data collection may lead to results closer to what has been reported in the populationConclusions

×