Mapping discrimination in Europe through a field experiment in amateur sport

Using mock email accounts with typical native- and foreign-sounding names, we contacted 23,020 amateur football clubs in 22 European countries, asking to participate in a training session. Response rates differed across countries and were, on average, about 10% lower for foreign-sounding names.

Like Comment
Read the paper


Finding a job, renting an apartment, and using the sharing economy are examples of activities where minority groups face a disadvantage. Through field experiments and mock applications, many studies find that individuals with foreign-sounding names receive fewer invitations for job interviews, apartment visits, and even dates, like if finding the one is not already difficult enough... Our project focuses on amateur football clubs.

In amateur football, people play for fun (I am a huge fan of Sunday football highlight videos, you should check them out!), the rules are universal and equally apply to everyone regardless of ethnicity, race, gender, sexual orientation, or religion. Additionally, amateur football is not driven by economic interests. This is why the European Union (EU) and the Union of European Football Associations (UEFA) officially recognize the potential of sport in integration and implement programs to include people who come from disadvantaged backgrounds. Amateur football is supposedly an integration tool readily accessible for everyone, but the evidence from our experiment proves otherwise.

We performed the field experiment in 22 European countries. Amateur football operates similarly in most European countries, where self-organized clubs compete in matches organized by regional federations and lack a regulated process of player admission. The task of inviting a new player normally falls to coaches, players, and other staff members, and reaching these clubs via email is a very common way of finding an opportunity to join the group.

The experiment

Identifying amateur football clubs, creating fake email accounts, buying SIM cards, selecting names, assisting researchers are exhausting but very fun tasks. In this case, we have to multiply all this fun by 22, namely: Austria (n=1840), Belgium (n=663), Croatia (n=447), Czech Republic (n=1598), Denmark (n=1135), England (n=1527), Finland (n=536), France (n=1847), Germany (n=1681), Greece (n=437), Hungary (n=345), Ireland (n=308), Italy (n=1463), Norway (n=999), Poland (n=1312), Portugal (n=791), Romania (n=493), Russia (n=1143), Serbia (n=383), Spain (n=1410), Sweden (n=1493), and the Netherlands (n=715). In total, 23,020 clubs. VERY FUN!

Figure 1. Overview of amateur football clubs in the sample.

Before contacting the clubs via email, we generated names for each group in each country: five native-sounding names and six foreign-sounding names (two names for each foreign group, representing each of the three largest foreign groups in the respective country). The language of the email coincided with the language of the homepage of the club. The email read as follows:

Subject: Trial practice

I would like to take part in a trial training session with your team. I have already played at a similar level. Could I come for a trial training session?
Many thanks


We categorized responses as (1) no response or rejection; (2) positive response (i.e., an invitation to a practice session); and (3) positive response with further inquiries (i.e., an invitation with additional questions; typical questions concerned the position, age, and previous playing experience of the applicant). To analyze the results, we combined positive responses and positive responses with additional inquiries into one group.

Long story short, individuals with foreign-sounding names have it more difficult to get a response from an amateur football club. On average, the response rate is approximately 10% lower for individuals with foreign-sounding names. We do observe differences by country. In some countries, the difference is below 4%, e.g., Ireland, France, and Portugal, and in some other countries, the difference is above 20%, e.g., Croatia, Hungary, and Austria. For most of the remaining countries, the response rate difference is between 5% and 13%. See Figure 2 below.

Figure 2. Response rate for foreign- and native-sounding names.

We hope this project lays the groundwork for other field experiments in sports and social activities. Future research can modify the design of the experiment and the text of the email to capture the effect of individual and coordinated actions (e.g., policies from governing bodies) and isolate the influence of other attributes such as language proficiency, religion, or nationality hierarchy.

Link to complete paper:

Carlos Gomez Gonzalez

Postdoc, University of Zurich