Estimating the Gender Wage Gap with Gender-Specific Measurement Errors
Early Career Research Award
It is common in household surveys that a household member, usually the head, is selected as the representative respondent and that person provides data on all the members. This procedure is preferred by survey designers because it is cost efficient and saves time. The share of male representative respondents has decreased substantially in the last two decades, from about 90% to 55%. The author studies this significant trend and also examines whether the change of the gender of household representative respondent biases the estimate of the trend of the gender wage gap.
Does it matter WHO Responded to the Survey? Trends in the U.S. Gender Earnings Gap Revisited
ILR Review 65(1)(January 2012): 148-160
Does it Matter Who Responded to the Survey? Trends in the U.S. Gender Earnings Gap Revisited,
IZA Discussion Paper No. 5512, 2011