Extended Data Table 1 Comparison of ethnic penalties with estimates from a contemporaneous correspondence study
From: Monitoring hiring discrimination through online recruitment platforms
- Our estimates of ethnic discrimination (columns 1 and 2) are compared with those reported by Zschirnt and Fibbi14 (column 3) in a previous correspondence study in Switzerland between October 2017 and October 2018. To replicate that sample as closely as possible, the ethnic penalties in columns 1 and 2 are estimated for the same ethnicities, the same gender and in the same occupations as the jobseeker profiles in the Zschirnt and Fibbi study14: sales assistants (male and female applicants), electricians (male applicants only), nurses (female applicants only) and HR clerks (male and female applicants). In the correspondence test, an ethnic background was indicated by the names of jobseekers. In addition, all minority applicants in the Zschirnt and Fibbi study14 had dual nationality (Swiss, and the country of origin of their parents). As Job-Room does not show two nationalities, we cannot replicate the profiles of the correspondence test exactly. Instead, we estimate two different models: column 1 reports the effect of ethnicity for jobseekers with a Kosovan, Turkish, and German or French citizenship and with language skills from the respective region; column 2 estimates the effect of ethnicity for Swiss citizens with a foreign-sounding name and language skills from the respective region. Standard errors are shown in parentheses and P values (two-sided t-test against the null hypothesis of no effect) in square brackets. No adjustments were made for multiple comparisons.
