publ-mit-podpubl-mit-podTöpfer, MarinaCastagnetti, CarolinaRosti, Luisa2024-04-082024-04-082016-01-152015https://hohpublica.uni-hohenheim.de/handle/123456789/5979This paper analyzes the effect of public-contest recruitment on earnings by applying an extended version of the Oaxaca-Blinder model with double selection to microdata on Italy. We find that the gender pay gap vanishes among public-contest selected employees, and even reverses in favor of women (-17.4%) in the young sample. The reversal is because public contests are merit-based and gender-fair screening devices. They are merit-based because selected employees possess higher productive characteristics than unselected ones, both women and men. They are gender-fair because the coefficients component in the Oaxaca-Blinder decomposition is never significant among public-contest recruited employees, either with or without selection. On the contrary, among employees not hired by public contest the gender pay gap is positive and significant (7.6%), and it is entirely due to coefficients, that is to discrimination in the career path.engGender pay gapDouble sample selectionPublic-contest recruitment330LohnstrukturGeschlechterforschungWorkingPaper454061242urn:nbn:de:bsz:100-opus-11709