Working Paper: NBER ID: w29350
Authors: David Card; Fabrizio Colella; Rafael Lalive
Abstract: In spring 2005, the Ombud for Equal Treatment in Austria launched a campaign notifying employers and newspapers that gender preferences in job ads were illegal. At the time over 40% of vacancies on the nation’s largest job board stated a gender preference; within a year the rate fell below 5%. We merge job board vacancies and employer records to study how the campaign affected hiring choices and the gender diversity of occupations and workplaces. Using pre-campaign data, we predict the use of gender preferences, then conduct a difference-indifferences analysis of hiring outcomes for vacancies with predicted male or female preferences, relative to those with no predicted preferences. The elimination of explicit gender preferences boosted the share of women hired for jobs that were likely to be targeted to men (and vice versa). At the firm-level, we find that the campaign led to a rise in the share of women at firms that were more likely to use male SGP’s, and a symmetric increase in the share of men at firms that were likely to use female SGP’s, with no effects on firm-survival, employment, or average wages.
Keywords: gender preferences; job vacancies; workplace diversity; hiring outcomes
JEL Codes: J16; J63
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| OET campaign (O30) | increase in fraction of women hired for vacancies with predicted male preference (J79) |
| OET campaign (O30) | increase in fraction of men hired for vacancies with predicted female preference (J79) |
| OET campaign (O30) | hiring of workers who would have been screened out due to gender preferences (J79) |
| OET campaign (O30) | increased gender diversity in occupations where predicted preference matched modal gender (J29) |
| OET campaign (O30) | positive contribution to workplace gender diversity (J29) |
| OET campaign (O30) | no harm to firm performance metrics (L25) |