Blind Recruitment and Reducing Bias: What Works and What Is Theatre
Blind recruitment can reduce hiring bias, but the evidence is more complicated than the slogans suggest. One major trial even found that anonymising CVs disadvantaged some minority candidates. Here is where blind screening helps, where it does not, and how it fits into a fairer recruitment process.
In 2004, economists Marianne Bertrand and Sendhil Mullainathan sent nearly 5,000 fictitious job applications in response to real vacancies in Boston and Chicago.
The researchers randomly assigned the applications either white-sounding names, such as Emily and Greg, or African-American-sounding names, such as Lakisha and Jamal. Otherwise equivalent applicants received markedly different responses: those with white-sounding names received approximately 50% more callbacks.
Put another way, an applicant called Emily needed to send around 10 applications to receive one callback. An equally qualified applicant called Lakisha needed to send around 15.
That finding explains the appeal of blind recruitment. When a candidate's name can affect whether their experience is taken seriously, hiding it during the initial sift seems an obvious way to make the process fairer.
For recruitment agencies, this is not only an ethical issue. Every capable candidate rejected for reasons unrelated to their ability is talent that could have been placed successfully.
However, anonymisation is not a universal solution. Its effectiveness depends on what information is removed, when it is removed and what kind of assessment replaces it.
What is blind recruitment?
Blind recruitment, sometimes called anonymised recruitment, removes personal information from an application before it is reviewed.
This will usually include the candidate's name, but may also include:
- A photograph
- Age or date of birth
- Gender markers
- Home address
- Graduation dates
- The names of schools or universities
The aim is to focus the initial decision on information relevant to the job: the candidate's skills, experience, achievements and evidence of capability.
The basic logic is sound. A recruiter who cannot see a candidate's name cannot allow that name to influence the decision, consciously or otherwise.
But removing information is not the same as creating an unbiased process. It solves a particular problem at a particular stage.
The evidence for blind recruitment — and the study that complicates it
Correspondence studies conducted in different countries and labour markets have repeatedly found that personal characteristics can affect employer responses, although the size and nature of the effect varies.
This makes the initial sift a logical place to use anonymisation. It is often a high-volume process in which recruiters make quick judgements based on limited information. Removing details that are not relevant to performance can help prevent those details influencing who reaches the next stage.
However, a large French trial produced a surprising result.
In 2009 and 2010, the French public employment service ran a randomised trial involving employers that had volunteered to test anonymous CVs. Participating firms were randomly assigned to receive either standard applications or anonymised ones.
The expectation was that anonymisation would improve the prospects of applicants from minority or disadvantaged backgrounds. Instead, employers receiving anonymous CVs became less likely to interview and hire those candidates.
The researchers found that participating employers were already more likely than other employers to interview minority applicants. When those employers could see a candidate's background, they appeared to interpret potentially negative signals, such as interrupted employment, differently and sometimes less harshly.
Anonymisation removed that context. It prevented discriminatory assumptions, but it also prevented employers from considering circumstances that may have led them to give an applicant the benefit of the doubt.
The findings contributed to France abandoning plans to make anonymous CVs compulsory.
Anonymisation does not add fairness by itself. It removes information. Whether that improves the decision depends on how the information would otherwise have been used.
A practical example
Imagine a recruitment consultant reviewing 12 applications.
In the first version of the process, the consultant sees each candidate's name, photograph, address, university and employment history.
One applicant has a strong record of relevant work but also has a two-year career gap. The consultant is under time pressure and cannot immediately place the candidate's name or background. The application goes into the "maybe" pile and is never reopened.
In the anonymised version, the consultant sees the same candidate's skills, previous roles and measurable achievements without their name, photograph or address.
The career gap remains visible, but there is less personal information onto which the consultant can attach an assumption. The candidate's relevant experience receives more attention and they reach the interview stage.
That is blind recruitment working as intended.
But imagine that the consultant's next step is an informal telephone call before any structured assessment. The candidate's name, accent and personal circumstances are immediately apparent, and the conversation is judged according to the consultant's general impression rather than consistent criteria.
The anonymised sift may still have helped the candidate reach the call, which is a genuine improvement. But it has not protected the final decision from bias.
When blind recruitment becomes theatre
Blind recruitment becomes performative when an organisation can say that it anonymises applications without changing how important decisions are made.
There are three common failure points.
The CV is anonymous, but the real assessment is not
Hiding names during the initial review achieves relatively little if the first meaningful decision takes place during an informal phone call, a LinkedIn search or an unstructured conversation.
The point is not simply to anonymise the stage that is easiest to anonymise. It is to improve the stage at which candidates are selected or rejected.
The name is removed, but obvious proxies remain
A graduation date may indicate someone's likely age. A home address can suggest their socioeconomic background. Certain schools, employers and activities may also create assumptions about class, ethnicity or opportunity.
However, the French trial provides an important warning against removing every piece of context indiscriminately. Context can sometimes help a fair-minded assessor understand a non-linear career, interrupted education or unconventional route into a profession.
The aim should be to remove irrelevant triggers for bias without flattening every candidate into an identical profile.
The shortlist is blind, but the interview is unstructured
Anonymisation can help overlooked candidates reach interview. That is worthwhile.
But if the interview consists of different questions for different candidates, informal conversation and decisions based on "fit" or instinct, bias has simply been postponed.
Blind screening may create a fairer shortlist. It does not automatically create a fairer appointment.
What the evidence supports more consistently
The evidence supporting structured assessment is more consistent than the evidence supporting anonymisation on its own.
Anonymisation removes one potentially harmful input. Structure improves the quality and consistency of the decisions that follow.
The CIPD recommends several practical measures.
Use structured interviews
Ask every candidate the same job-relevant questions in the same order. Agree in advance what a strong answer should contain and score each response against consistent criteria.
This makes candidates easier to compare and reduces the opportunity for personal impressions to determine the outcome.
The CIPD cites evidence that introducing a scoring rubric increased the likelihood of Black women being selected for a role by 21%. This does not prove that structured assessment will produce the same improvement in every organisation, but it illustrates the potential effect of making decisions more consistent.
Agree the criteria before seeing candidates
Define what the job genuinely requires before reviewing any applications.
Without agreed criteria, recruiters can begin with a preference and then rationalise it afterwards. Clear criteria make it harder to move the goalposts for different candidates.
Assess skills rather than pedigree
Where possible, assess what candidates can demonstrate rather than relying on the prestige of their former employer or university.
Work samples, job-relevant exercises and evidence of previous results can provide more useful information than institutional reputation.
Use independent scoring
Where more than one person is involved, each assessor should score the candidate independently before discussing their conclusions.
This reduces the likelihood of the most senior or confident person shaping everyone else's judgement before they have formed their own view.
A visibly diverse interview panel may bring a wider range of perspectives, but diversity of the panel does not guarantee a fair decision by itself. The useful mechanism is independent assessment against clear and relevant criteria.
The UK legal context
The Equality Act 2010 makes it unlawful for employers to discriminate against job applicants because of protected characteristics.
UK government guidance states that employers must not treat applicants less favourably because of characteristics including race, sex, age, disability, religion or belief, sexual orientation, pregnancy and maternity, marriage or civil partnership, or gender reassignment.
Application forms should not ordinarily ask for information such as a date of birth unless age is genuinely relevant to carrying out the role. Equality-monitoring information should also be kept separate from the people responsible for selecting and interviewing candidates.
Blind recruitment can support a fairer process, but using it does not by itself prove that an organisation's recruitment practices are lawful or unbiased. The criteria, assessments, interviews and final decisions must also be objective and relevant to the role.
The commercial case for fairer recruitment
For a recruitment agency, a fair process is not an act of charity.
Clients increasingly expect agencies to search beyond the most familiar networks and produce credible, varied shortlists. They can often tell the difference between a genuinely broad search and a conventional shortlist presented as a diversity initiative.
More importantly, the callback gap identified by Bertrand and Mullainathan represents capable people being overlooked.
An agency that finds candidates others miss has access to a larger talent pool. It can make better placements, present stronger shortlists and build trust with clients that care about both quality and fairness.
The compliance benefit matters, but the commercial advantage is real too.
Where recruitment software fits
Technology can make anonymous screening practical at scale.
ATSpro includes anonymised screening within its UK compliance features, allowing recruiters to review relevant candidate information without manually redacting every CV.
Software can also help teams use consistent stages, criteria and assessment records rather than relying on individual memory.
But a tool cannot design a fair recruitment process on its own.
It can remove a name from a profile. It cannot decide whether the next step is a structured assessment or an unrecorded telephone conversation based on instinct. It cannot prevent a hiring manager from disregarding the agreed criteria.
Software can support good practice and make it easier to follow consistently. It cannot replace it.
The takeaway
Blind recruitment is a useful intervention, but a narrower one than it is sometimes presented as.
It can reduce the influence of identifying information during an early, hurried sift. It can help capable candidates reach a stage they might otherwise have been denied.
It can also remove useful context, and any benefit may be lost if the rest of the process relies on informal conversations, inconsistent questions or personal impressions.
The strongest approach is to combine anonymised screening with clear job criteria, structured interviews, skills-based assessment and independent scoring.
Used as part of that process, blind recruitment can help agencies identify candidates a careless system might overlook.
Used as a policy badge without changing how decisions are made, it offers more reassurance than protection.
**Sources: Marianne Bertrand and Sendhil Mullainathan, "Are Emily and Greg More Employable Than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination," *American Economic Review*, 2004; Luc Behaghel, Bruno Crépon and Thomas Le Barbanchon, "Unintended Effects of Anonymous Résumés," *American Economic Journal: Applied Economics*, 2015; CIPD, "Inclusive recruitment: Guide for employers."; UK Government, "Employers: preventing discrimination in recruitment."; per user per month.*.**