Older Workers – Ageist Language in Job Ads

We develop and implement methods to explore the role of stereotypes in hiring discrimination using the text of job ads, and apply this method to evidence on age discrimination. We make three contributions. First, we develop techniques that leverage machine learning and textual analysis to analyze the text data in job ads from a large-scale field experiment on discrimination. Second, we use these techniques to produce evidence on which age-related stereotypes that appear in job ads are associated with an experimental measure of hiring discrimination against older workers – the first evidence we know of that can establish relationships between age-related stereotypes and actual employer behavior. Third, our analysis provides evidence on whether employers with less intent to hire older workers – as captured in our experimental results – use ageist language in their ads.