Every time I write about experiments like the one I am going to discuss today, I have to think of the song from the musical Avenue Q that gave this post its name. The point is not that people are racist, nor that society is ‘systemically racist’. That is mostly nonsense if you ask me. But as the song so neatly explains, ‘no-one’s really colour-blind’, and that clouds their judgement in subconscious ways.
And since we are on the topic, let me clarify my stance. I think that the claims that the US or the UK are systemically racist are pretty over the top and not supported by evidence. On the other hand, I find the arguments of anti-DEI activists that we should purely judge people based on merits correct but impossible to implement in practice. Most people try to judge other people based on their deeds rather than their gender, ethnicity, etc. Otherwise, this song’s lyrics wouldn’t get such a laugh.
But there is a plethora of evidence that we are not treating people equally and independently of their gender, ethnicity, etc. We are humans and thus subject to many unconscious biases. To counteract these biases and give people of minority ethnicities a truly fair chance in job applications, career progression, etc., we need to put corrective measures in place.
Are gender or ethnic quotas the proper correctives? That is very much up for debate and should be debated rationally based on evidence and without appeals to slippery slope arguments, political slogans, or other polemic tools. Shouting slogans and absurdities at each other doesn’t help anyone and is what got us into the mess we are in today.
Thus, consider the following not as a call for ethnic or gender quotas or other affirmative action but as another piece of data to consider when considering your actions and how businesses deal with investors and other stakeholders.
A group of academics from the University of Illinois and the University of Amsterdam sent letters to the investor relations teams of Stoxx Europe 600 companies asking for ESG information about their companies. These letters were identical except that they differed in the name of the retail investor who sent the request. The letter was sent in English and the investor relations team must have perceived it as coming from the UK.
Below is the response rate of the different investor relations teams to the requests based on the investor’s name. Note that the response rates are overall quite low because the request came from a retail investor rather than an institution or a professional analyst at a bank or broker.
Response rate to retail investor requests about ESG
Source: Brown et al. (2025)
Obviously, the key insight is that requests from someone with the name of an ethnic minority tend to be answered less often than requests from someone with a name indicating a European heritage (James). The differences in response rates are statistically significant for two of the three names with an ethnic minority background and when comparing James with the average combined response rates of all other names.
Again, this is not conclusive evidence for widespread bias, but it is another data point in many studies that point in this direction. If you are curious about whether you have some unconscious bias, you can test it for yourself with Harvard University’s online implicit bias tests for a wide variety of potential biases. Having taken these tests, I know I am slightly biased in some areas, but I will not tell you which ones. That remains my secret.
