In 1990, roughly one in four players at the FIFA World Cup was based at a club outside the country his national team represented. At the 2026 edition it is nearly three in four. That is a structural shift in a labour market, and it has a surprisingly precise cause: a 1995 court ruling.
I pulled the squad data together, cleaned it into a small CSV, and put it on GitHub under CC BY 4.0 so anyone can check the trend or build on it. This post walks through what the data shows and how to reproduce the headline chart in a few lines of Python.
The headline number
Share of foreign-based players across World Cup squads:
| Year | Foreign-based share |
|---|---|
| 1990 | 26.2% |
| 2002 | 38.1% |
| 2014 | 53.9% |
| 2026 | 72.2% |
The inflection point is the Bosman ruling (European Court of Justice, 15 December 1995, Case C-415/93), which abolished EU limits on foreign players and allowed free transfers at contract end. Professional footballers became, legally, ordinary workers with the right to work anywhere in the single market.
Reproduce it
The dataset is two CSVs. Loading the time series straight from GitHub:
import pandas as pd
url = ("https://raw.githubusercontent.com/DatapulseResearch/"
"world-cup-players-abroad/main/data/"
"world_cup_squads_foreign_based_share_1990_2026.csv")
df = pd.read_csv(url)
ax = df.plot(x="year", y="foreign_based_share_percent",
marker="o", legend=False,
title="Foreign-based players in World Cup squads, 1990 to 2026")
ax.set_ylabel("% of squad based abroad")
ax.axvline(1995, linestyle="--", alpha=.4) # Bosman
One line and you can see the post-1995 climb.
The counterintuitive part
You might expect the strongest nations to export the most players. The opposite is true. The 2026 per-country table (world_cup_2026_foreign_based_by_country.csv) shows the strongest domestic leagues keep their talent at home:
| Nation | Foreign-based share, 2026 |
|---|---|
| Switzerland | 100% |
| Argentina | 94% |
| Brazil | 89% |
| France | 83% |
| Germany | 33% |
| England | 22% |
England and Germany, with the strongest leagues, retain the most. Talent follows the biggest market, not the passport.
Data and method
- "Foreign-based" = a squad member under contract at a club outside his national team's country at tournament time.
- Sources: player citizenship and club affiliation from transfermarkt.de; FIFA World Cup squad lists.
- Dataset: https://github.com/DatapulseResearch/world-cup-players-abroad
- Full analysis with all charts: https://www.finalarm.de/studies/world-cup-players-abroad
If you do something with the data (a different cut, a model, a better chart), I would love to see it in the comments.