How to anonymise CVs

Andy Babbage

12

December

2018

10

min read

|

X
Free Training: How to De-bias Your Hiring
Learn how to build a more ethical, empirical hiring process in a single session.

X

Sign up to our newsletter

Want to see more content like this, every other week?
We're on a mission to change recruitment.

Sign Me Up

FACT: CVs lead to bias, which results in minority background candidates being unfairly overlooked.

Anonymising (or blinding) job applications (CVs, resumes, cover letters) is when identifying information is removed to conceal the backgrounds of job applicants. If you want to create a fairer recruitment processes, this is an essential step. By removing irrelevant information, you're also ensuring that candidates are being assessed on what matters, and that hirers are making the best possible hiring decision.

Although it sounds simple enough, anonymising or blinding applications is sometimes easier said than done, for a variety of practical, logistical and resource availability reasons.

Below, we'll touch on the benefits of the anonymised CV, how to actually do it in a practical sense and the extra steps you can take to make your process even more effective and inclusive.

We’ve been hiring this way for 3 years now and none of us would ever go back. It’d be like going back to scheduled television after getting used to Netflix.The quantitative and qualitative feedback is clear; you will recruit people from different backgrounds who are more likely to perform well in the role than if you relied on full CVs.

Why you should anonymise CVs

Unconscious bias is pervasive in recruitment processes. Unfortunately, none of us are impervious to this (check out our article Is my recruitment process biased? if you're not yet convinced).

But this isn't anyone’s fault, it's just how the human brain processes large amounts of incomplete information to make decisions.

As easy as it would be to point to a minority of bigoted, explicitly biased hirers that, if removed, would make for fairer outcomes, this simply isn't the case.

We're all biased, no matter how well-intentioned we may be.

Bias brings unhelpful noise into complex decisions such as hiring someone to join your team. The good news is that talent acquisition professionals are at the forefront of this battle, are often very well versed on the negative impact of this bias and are well poised to combat it.

Before we get into ways of tackling bias, let's take a quick look at how it actually happens...

Our brains use two systems to make decisions.

System 1: Fast and intuitive, used for everyday decisions (like walking to work on 'autopilot')

System 2: Slow and conscious, used for bigger, more important decisions (like planning a one-off trip or a big project at work).

System 1 uses mental shortcuts and associations to help make the 1000's of micro-decisions that we encounter every day. These are the sort of decisions that if you had to think long and hard about, you’d probably never get out of bed in the morning. So, System 1 is absolutely necessary to lighten our mental load so that we can focus on the more important decisions.

However, unconscious bias arises when we default to using System 1, when we should be using System 2...

and the more rushed or fatigued we are, the more likely this is to happen.


Types of unconscious bias

The first port of call when addressing this issue is often unconscious bias training, which according to the research, has mixed results...

A meta-analysis of 426 studies (involving more than 80,000 participants) found that although there was a reduction in bias immediately after training (a very slight one), this wore off after about 8 weeks.

Effect of implicit bias training chart

Why doesn't it work? Well, given that much of bias is unconscious, awareness alone just isn't effective. You can't tell someone to stop doing something that, for the most part, they're not even aware that they're doing!

For more, read our article: Does unconscious bias training work?

The next well known method is blind hiring and the easiest-adopted approach to that is often anonymised CVs/ anonymized resumes.

There's 30 years of research which points to the dangers of including names and other identifying information on an application and blind hiring is one of the ways to mitigate these risk (check out this blog: What is blind hiring?).

Unconscious bias training doesn't work because it attempts to change human nature.

Even if this is possible, it would take an eye-watering investment if time and money.

Instead, what has been proven to work is changing processes, rather than people themselves - which is why anonymized resumes are a must.

So, we believe that anonymised CVs are an important first step to start removing bias from a process and as a result starting to make even better hiring decisions while bringing diversity of thought into your organisation.

The case for anonymises CVs - a look at the data

If we look at a 2019 study from the University of Oxford, we can see why anonymised CVs are necessary...

Identical applications were sent our to employers, with only the candiate's name being changed to reflect ethnicity.

Minority ethnic backgrounds had to send 80% more applications to get the same results as a White-British person.

The worst part? Researchers found that this rate of discrimination had barely changed since similar experiments back in 1969.

Hiring discrimination UK stats

This isn't an issue exclusive to the UK...

We tend to see similar outcomes all across the western world.

Additional CVs needed for a callback by country

Needless to say, bias at the screening stage applies to more than just race.

Candidates can be also be discriminated against on the basis of their:

For more research, check out our Recruitment Bias Report.

The types of information you should remove

A typical CV is riddled with unnecessary information that, at best, has nothing at all to do with their ability to do the job, and, at worst, sends misleading and erroneous signals that the person is not suited for the role.

We’ve broken down the types of information that you'd typically find on a CV and listed the biases that might come into play, along with an urgency that this type of information should be removed.

Unconscious bias affect on CV review


Sacre Bleu! Get these off of CVs right now

  • Photos - what you look like very rarely has anything to do with how well you can do a job (except maybe modelling - but even then beauty is subjective). A photo can subject you to affinity bias causing you to prefer people who look like you.
  • Name - someone’s name sends a lot of signals about their background. If someone’s name is like yours you are more likely to assess them positively. If someone’s name is unusual to you, then you’ll make other assumptions about them which are most likely misleading.
  • Age - age is just a weak indicator of how much someone has practiced their skills. Instead of looking at that, how about assessing skills more directly and getting a better signal?
  • Gender/religion (and other PPC) - personal protected characteristics are called this because they are protected by law. You cannot discriminate based on these characteristics and they’re not really anyone’s business in the first place.

You might want to ditch these too

  • Hobbies - you may argue that this gives an insight into the character of the applicant or whether you would like to work with them. These considerations may be important, but care has to be taken that they are not catch-all terms for ‘they weren’t like me’ or ‘I couldn’t go to the pub with him’. Also if you share a common hobby (polo or yachting for example), this might make you view their application more positively than those that are different to you and therefore could bring something different to the team
  • Professional affiliations - professional certifications I get, but otherwise I’m not entirely sure of the value of this type of information. Is it confirmation that this person comes from the same club as everyone else? If it’s not valuable to know this then bung it.

And if you want to take it up to 11

  • Educational institute names - Number of years of education is one of the least predictive forms of assessing if someone will perform well in a role (check out the seminal Schmidt study). Additionally, the name of an institute may be a proxy for intelligence, but it also signals a whole host of other socioeconomic stuff mushed in together. At the end of the day it is a proxy for a skill or ability you are trying to assess. Don’t rely on proxies, assess based on the skill itself.
  • Previous company names - you see so many job ads that say ‘must have worked in top 4 <insert industry=""> firm’. Again this is a proxy and if you rely on it you will always recruit the same type of person.</insert>
Predictive validity of assessment methods


Practical methods to anonymise CVs

Here are 3 ways of anonymising and blinding CVs, resumes and cover letters in ascending order of ease and effectiveness:

1. Manually

This is the way that many pioneering Talent Acquisition professionals have been blinding their recruitment processes for years. The approach is simple; print out the CVs and take a sharpie to the sensitive bits - you've got yourself an anonymised CV! We’ve done this before and with high numbers of applicants it isn’t fun and can get very confusing very quickly.

This is much better than an unblinded CV sift, but still a lot of work (but definitely worth it!).

Printing and manually blinding is one method
Blergh. Printing. This is soooo 1440.


2. Using an Applicant Tracking System

Applicant Tracking Systems (ATS) are a great way to track which candidate is which by helping to streamline your recruitment process. The problem is, as the name suggests, they are very one-way, kind of like attaching an ankle tracking device to someone and then forgetting they are human as they filter through your process.

Blinding is often done on these systems as an afterthought. The way it works is to first require your poor candidate to re-enter their entire professional history into a series of clunky pull-down menus and radio buttons. Once the candidate is thoroughly annoyed, the ATS then simply does not show the name of the person to the review. The problem is that the rest of the sensitive information is usually shown - why? because it’s an afterthought.

3. Using a People Decision Platform

A People Decision Platform has quality of hiring decision and diversity & inclusion at its very core, as it is centred around the person or human that is applying for the job (whilst still being a highly efficient system for everyone to use). Firstly, these systems are positioned to care about the skills and ability of a candidate, not the misleading proxies that you typically find in a CV. Secondly, they will carry out the blinding and much more to remove even more bias from a recruitment process as second nature. What’s more, they will prove that the anonymisation has made a difference by using real hiring data as a feedback loop to show the improvement. This is the whole hog, but really it’s just an enablement tool for great talent acquisition teams who have already been pushing in this direction for years. Surprise, surprise, Applied happens to be a People Decision Platform :).

Beyond anonymised CV and blinding

Anonymising CVs is just the first step in structuring your applicant data in order to quickly get to the substance and skills of an applicant. There are dozens of other behavioural science methodologies that can be woven throughout a process to improve its effectiveness. Have a read of our article on How do I remove bias from my recruitment process? for more information.

Beyond CVs and Resumes - Work Sample Questions

The days are numbered for CVs and resumes, after all they have been in use for hundreds of years (far too long). They’re riddled with bias, full of misleading proxies and are actually one of the least effective predictors of job performance. There are lots of alternative methods to CVs out there, the one we recommend is simply using the most predictive forms of assessment upfront.

For most roles this is ‘work sample questions’ which challenge the applicant to think through a realistic challenge that they could meet on the job.

Work samples test for skills directly instead of relying on proxies like education and experience.

Work samples are created by identifying which skills are needed for the job and then building questions that will test these skills.

Here’s a work sample we used for a recent Community Lead role:

You’ve been invited to be on a panel on hiring & recruitment. You’re the only D&I expert (possibly the only one that thinks it’s important there) in the room. What are your opening lines to the audience to convince and engage them on the subject?

Skills tested: D&I Knowledge, Communication

The best work samples rare those which mirror real-life tasks/scenarios most closely.

At Applied, candidates submit 3-5 of these work samples anonymously instead of a CV.

Hiring Bundle


What next?

Hopefully this article gave you pause for thought about how to go about anonymising your process and much more. Check out our resource centre for a whole bunch of guides and best practices on this. If you're interested in learning more about our anonymous recruitment software, request a demo of the Applied platform. Happy hiring!