Five European programmes have tested anonymous CVs, and their results differ. In the French experiment of 2009 and 2010, women's chance of an interview with a male recruiter rose from 1 in 27 to 1 in 6, while candidates with an immigrant background or from priority urban areas fell from 1 in 10 to 1 in 22.
The five programmes side by side
Public bodies in Sweden, France, Germany, the United Kingdom and Finland each chose which details of an application to hide from the people who decide whom to interview. The table follows the order in which the programmes ran, and the Sources list links the text behind every row. The results differ from one setting to the next, so each row reports them as its source states them. All figures were checked against the sources on 9 November 2026.
| Programme | What was hidden | Who was tested | What was measured | Result as the source states it |
|---|---|---|---|---|
| Sweden, Gothenburg, 2004 to 2006 | For example the name, gender and country of origin of the applicant, in the first stages | 3,529 applicants to 109 positions in two city districts, with a third district for comparison | Interviews and job offers | Women and applicants of non-Western origin were more likely to reach an interview. Women were also more likely to receive an offer, and no such effect was found for immigrants. |
| France, Pôle emploi, 2009 to 2010 | The identity block of the CV: surname, first name, address and date of birth | About 1,000 establishments in eight départements, with job offers drawn at random to receive named or anonymous CVs | Interviews, then hiring | With a male recruiter, women's chance of an interview went from 1 in 27 to 1 in 6 and men's from 1 in 5 to 1 in 13. For candidates with an immigrant background or from priority urban areas it went from 1 in 10 to 1 in 22, against 1 in 8 to 1 in 6 for the others. |
| Germany, pilot of the Federal Anti-Discrimination Agency, 2010 to 2011 | Name, gender, nationality and place of birth, disability, date of birth or age, marital status and photo | Eight organisations for 12 months each, with 8,550 applications read anonymised | Invitations to a test or an interview, and offers | 1,293 people were invited and 246 received an offer. Within the anonymised procedures, women and people with a migration background had the same chance of an invitation as the other groups. The report states that the results are not representative. |
| United Kingdom, Civil Service, 2015 | The name | Civil Service roles below Senior Civil Service level, as a commitment for all of them | Nothing reported | The press release of 26 October 2015 announces the commitment to name-blind recruitment and cites no evaluation. |
| Finland, City of Helsinki, 2020 | Name, address, age, language skills and educational institution until interview invitations were sent; the length of previous work experience was shown, the exact dates were not | 13 job titles, from February to December 2020 | Interviews, hires, job duration and wages | For applicants with a foreign-sounding name the interview probability rose by 5 percentage points relative to applicants with Finnish-sounding names; overall, 4 per cent of the former were called to an interview against 8 per cent of the latter. The effect on hiring was similar or somewhat larger, with no effect on job duration or wages. |
France: one experiment, two results
The French experiment ran from November 2009 to November 2010. Pôle emploi sent anonymous CVs to the employers behind a randomly chosen half of the participating job offers, and named CVs to the other half. The anonymous version removed the identity block and replaced it with an identifier that let the adviser identify the candidate.
The first result concerns recruiters who tend to favour people like themselves. With named CVs and a male recruiter, women had 1 chance in 27 of being invited to an interview and men 1 in 5. With anonymous CVs the gap reversed, and women had 1 chance in 6 against 1 in 13 for men. The report finds the same pattern, with smaller gaps, when the recruiter is a woman, and the reversal reaches the final hiring decision.
The second result goes the other way. Candidates with an immigrant background or living in a priority urban area had 1 chance in 10 of an interview with named CVs, against 1 in 8 for the other candidates. With anonymous CVs the gap widened to 1 in 22 against 1 in 6. The authors' paper in the American Economic Journal states that participating firms became less likely to interview and hire minority candidates when they received anonymous résumés, and explains this by the self-selection of firms into the programme and by anonymisation preventing the attenuation of negative signals. The 2011 report gives an example of that second mechanism: a gap in a CV can be read differently when the address shows a priority urban area than when the address is hidden.
The legal status followed a separate path. A law of 2006 made the anonymous CV compulsory for firms with more than 50 employees, and the implementing decree was never published. Law no. 2015-994 of 17 August 2015 changed the rule, and article L1221-7 of the Labour Code now says that the information a candidate gives in writing may be examined under conditions that preserve anonymity. The news page of the Conseil d'État on the change states that the anonymous CV is optional and no longer compulsory.
Germany: eight organisations and a list of details to hide
The Federal Anti-Discrimination Agency announced a pilot in August 2010, and it started in November 2010 with eight companies, public authorities and municipalities that joined voluntarily for 12 months each. The Institute of Labor Economics (IZA) and the Viadrina University evaluated it. Before the pilot, IZA had recommended hiding the name, gender, nationality and place of birth, disability, date of birth or age, marital status and the photo, and the organisations largely followed that list, with exceptions for gender and disability where legal promotion rules apply.
The organisations used four methods: standardised application forms that never ask for the sensitive details, blinding the details in the online application system, transferring each application into a standard table, and blacking out the details in the documents, by hand or in the PDF. The report found the standardised form the most workable. Blacking out kept the certificates and the overall impression of the application readable to the panel, and it was slow: 15 to 40 minutes on average for a conventional application, and 5 to 10 minutes when few details had to go.
On the results, the report says that within the anonymised procedures women and people with a migration background had the same chance of an invitation as the other groups. Compared with conventional procedures, women tended to have better chances. For applicants with a migration background the starting point differed: where their chances had been worse, they improved, and where no systematic difference existed, anonymisation changed nothing. The authors add that most participating organisations already promoted diversity, so the results are not representative.
Sweden and Finland: interviews, offers and hires
Gothenburg decided on its pilot in February 2004, and it ran from October 2004 to June 2006 in the city districts of Centrum and Kortedala. The district of Gunnared kept its usual procedure and served as the comparison, and the districts were not chosen at random. The researchers followed 3,529 applicants to 109 positions through each stage. In the openings where the anonymous procedure was used, gender and region of origin did not affect the probability of an interview, while they did in the comparison openings. Women also received more job offers, and no such effect was found for immigrants.
The Helsinki pilot ran from February to December 2020 for 13 job titles. The city's HR department forwarded summary files to the hiring managers with previous experience, education, degree titles and motivation letters, and it prepared them by hand: 44 of the 99 recruitment processes in the pilot were verifiably anonymous. After the invitations went out, managers saw the full files and ran the interviews as usual. The working paper reports that interview probability for applicants with a foreign-sounding name rose by 5 percentage points, that hiring moved in the same direction with a similar or somewhat larger effect, and that job duration and wages did not change. The city's news item reports that the gender of the people hired did not change, and it quotes a researcher who points out that the pilot was small and the results need careful reading.
What to hide in a CV, according to the sources
The programmes agree on one detail only, the name, and differ on the rest. The sources hid these details:
- The name was hidden in all five programmes.
- The address was hidden in France, Finland and, after the pilot, in the German recommendation.
- The date of birth or the age was hidden in France, Germany and Finland.
- The photo and the marital status are on the German list.
- Nationality, place of birth, gender and disability are on the German list, and the Swedish study names the country of origin and the gender among the hidden details.
- Language skills and the educational institution were hidden in Helsinki, which also showed the length of previous jobs without their exact dates.
Hiding the obvious fields does not remove every signal, and the sources say so. The German report notes that email addresses, telephone numbers, addresses and employment periods allow indirect conclusions about protected characteristics, and the participating organisations agreed to anonymise them as well. The French report warns that the sex and the age of a candidate can show in the body of the CV, and that languages and places of work can reveal origin or residence. The Swedish authors add that the place of education and the place of work experience must be hidden for ethnic anonymity, at least where ethnicity correlates with being foreign-born.
For the time after the pilot, the German report recommends a qualification-centred application that leaves at least the photo, name, address, date of birth, place of birth, nationality and marital status off the first page, while certificates and the usual documents stay in the file. The report adds that other details allowing indirect conclusions could be tolerated in that case.
Anonymising a batch of CVs with Redactic
Redactic replaces the personal data it detects in a CV with placeholders: names, addresses, email addresses, phone numbers, dates of birth next to a birth label and ID numbers. You can upload a ZIP of up to 200 CVs, and the same candidate keeps the same placeholder in every file, which keeps a cover letter matched to its CV. In a candidate list exported from your applicant tracking system, you mask whole columns. Nationality, gender words and profile links have no detection type in running text, so you select the word and Redactic masks every occurrence of it, and the same applies to exact job dates or the year of a degree.
Tick the option Black out all images and every image in a PDF or Word CV, a photo or a signature, becomes a black box. Text inside images is read with OCR in PDFs, Word files, scans and photos. Documents are never stored and never used to train AI models. The programmes above decided what stays visible, and with Redactic you decide it for your own selection: the HR and recruiting page shows the steps with a sample CV, and you can try it on your own document in the free trial.
If your team also uses an AI assistant on candidate files, the guide to shadow AI in HR under the GDPR covers that side. Whether a masked file counts as anonymous or pseudonymised in the GDPR sense is the subject of Pseudonymised or anonymous? EDPB 02/2026 and the SRB ruling.
Frequently asked questions
Do anonymous CVs reduce discrimination?
The sources report different results, and each programme has to be read on its own. Gothenburg and Helsinki reported more interviews for the groups they studied, the German pilot reported equal invitation chances within its anonymised procedures, and the French experiment reported a wider gap for candidates with an immigrant background. The table above gives each result as its source states it.
Is the anonymous CV mandatory?
In France it is optional. A law of 2006 made it compulsory for firms with more than 50 employees, but the decree needed to apply it was never published. Since Law no. 2015-994 of 17 August 2015, article L1221-7 of the Labour Code says that the information a candidate gives in writing may be examined under conditions that preserve anonymity. Sweden, Germany and Finland ran the procedure as public pilots, described above.
Which details should I remove from a CV?
Every programme in this article hid the name. The other details differ by programme, and the section What to hide in a CV lists them with the programme that used each one.
How long does it take to anonymise a CV by hand?
In the German pilot, blacking out a conventional application took 15 to 40 minutes on average, and 5 to 10 minutes when only a few details had to be hidden, for example in an application for a training place. The report found this method impractical for a large number of applicants.
Does Redactic remove the photo from a CV?
Yes, when you tick the option Black out all images: every image in a PDF or Word CV, such as the photo or a signature, becomes a black box in the file you download. A scanned PDF is a picture from the first page to the last, so there you draw a box over the photo in Review and edit. Details are on the HR and recruiting page.
Sources
- Behaghel, Crépon, Le Barbanchon, Évaluation de l'impact du CV anonyme, March 2011, accessed 2026-11-09.
- Behaghel, Crépon, Le Barbanchon, Unintended Effects of Anonymous Résumés, AEJ: Applied Economics 7(3), 2015, accessed 2026-11-09.
- Code du travail, article L1221-7, version of 19 August 2015, accessed 2026-11-09.
- Conseil d'État, CV anonyme, accessed 2026-11-09.
- Pilotprojekt Anonymisierte Bewerbungsverfahren, Abschlussbericht, IZA Research Report no. 44, April 2012, accessed 2026-11-09.
- Åslund, Nordström Skans, Do anonymous job application procedures level the playing field?, IFAU Working Paper 2007:31, accessed 2026-11-09.
- Åslund, Nordström Skans, Industrial and Labor Relations Review 65(1), 2012, accessed 2026-11-09.
- Kanninen, Kiviholma, Virkola, Anatomy of an Anonymous Hiring Pilot, Labore Working Paper 338, 2023, accessed 2026-11-09.
- City of Helsinki, news of 12 September 2023 on the anonymous hiring pilot, accessed 2026-11-09.
- GOV.UK, press release of 26 October 2015, accessed 2026-11-09.