Do large employers treat racial minorities more fairly? A new analysis of Canadian field experiment data
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Analysis of amended data from a large-scale Canadian employment audit study (Oreopoulos 2011) shows that large employers with over 500 employees discriminate against applicants with Asian (Chinese, Indian or Pakistani) names in the decision to call for an interview, about half as often as smaller employers. The audit involved submission of nearly 13,000 computer-generated resumes to a sample of 3,225 jobs offered online in Toronto and Montreal in 2008 and 2009 for which university-trained applicants were requested by email submission. An organization-size difference in employer response to Asian names on the resume exists when the Asian-named applicant has all Canadian qualifications (20% disadvantage for large employers, almost 40% disadvantage for small employers) and when they have some or all foreign qualifications (35% disadvantage for large employers, over 60% disadvantage for small employers). Discrimination in smaller organizations is most pronounced in considering applicants for jobs at the highest skill levels. As well, whereas the Asian-name disadvantage is overcome in large organizations when the applicant has an additional Canadian master’s degree, this is not the case in smaller organizations. It is suggested that large organizations discriminate less frequently because they have more resources devoted to recruitment, a more professionalized human resources recruitment process, and greater experience with a diverse staff complement. Experimentation with anonymized resume review may be an inexpensive way that organizations can test their own hiring procedures for discrimination.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.013 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it