Baby-name shares and European crime: the correlation test

A cross-country comparison found positive correlations between the share of boys receiving a Mohammed name variant and several police-recorded crime rates. Here are the calculations, the sensitivity test and what this analysis can establish.

By Doggy. Adapted from my original Instagram carousel, with source notes and workings collected in one place.

Baby-name shares and European crime: the correlation test: original carousel slide 1 of 6 — Original carousel, slide 1 of 6.

What was compared?

I compared 2024 baby-name shares compiled from national statistics with 2023 Eurostat police-recorded offences per 100,000 residents. The name compilation combines Muhammad, Mohammed, Mohammad, Mohamed and Mohamad. It covers 11 countries, but England and Wales were excluded from the correlation calculation because a comparable 2023 Eurostat series was unavailable. The resulting comparison has ten countries.

The name figures are a secondary compilation, reported through the Daily Mail and discussed in The Daily Sceptic. They should not be mistaken for a harmonised measure of religion or migration background.

The Pearson correlations

Recorded offence	Ten countries	Without Poland
Robbery	+0.802	+0.780
Serious assault	+0.778	+0.818
Intentional homicide	+0.736	+0.819
Sexual violence	+0.511	+0.380
Rape	+0.503	+0.325
Theft	+0.379	+0.016
Burglary	+0.197	−0.133

Pearson’s r measures a linear association and ranges from −1 to +1. It is calculated as the covariance of the two measures divided by the product of their standard deviations. It is not a percentage increase in crime or an individual risk estimate.

Removing Poland

Poland’s displayed name share is 0.01%, making it a useful sensitivity check. Removing it leaves nine countries. Robbery remains positively correlated, while the homicide and assault correlations rise. Other relationships weaken, especially theft and burglary. Reporting the full set makes that distinction visible.

What the pattern does not prove

This is a small, country-level, exploratory comparison using different years. It does not identify offenders, establish that a name or religion causes crime, or validate a prediction for Poland’s future. Differences in reporting, legal definitions, recording practices, age structures, urbanisation and other untested factors could influence the relationship. Removing one country is a robustness check, not a causal test.