|
Activity Number:
|
173
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Business and Economic Statistics Section
|
| Abstract - #303623 |
|
Title:
|
What Differentiates Between Women and Men in the Labor Market
|
|
Author(s):
|
Edna Schechtman*+ and Shlomo Yitzhaki and Yulanda Geva
|
|
Companies:
|
Ben Gurion University of the Negev and Central Bureau of Statistics and Central Bureau of Statistics
|
|
Address:
|
Reger Street, Beer Seva, , Israel
|
|
Keywords:
|
Overlapping ; Gini ; Gender ; Classification
|
|
Abstract:
|
Assume that you are asked to classify labor market results by gender. However, the gender is not known to you. You are asked to use one of three alternative proxies: salary per month, salary per hour worked, or the number of hours worked per month. Which proxy will be the best classifier according to gender? The aim of this paper is to use the decomposition of the Gini coefficient by population subgroups in order to answer the above question. The decomposition, called ANOGI (ANalysis Of GIni) , is similar in structure to ANOVA but reveals more. It turned out that the best classifier of the three is the number of hours worked per month. Repeating the same exercise for a different year indicates that the results are robust over time although a trend of reducing the quality of the classifier is observed.
|
