Olympian Immigrants

Immigrants are in the news these days, due to some political rulings by new US President Donald Trump. These rules have recently been rescinded by the US judicial system, but for a time they appeared they would have some effect on international sports, both with foreign athletes coming to the United States, but also USA athletes traveling to foreign countries, notably to Iran last week for a wrestling World Cup event.

Further, within the last year, a refugee storm has flooded Europe from the Middle East, especially Syria, because of civil strife in that country and attacks by ISIS. Many countries have expressed concern about how they would handle this refugee crisis and provide for the many immigrants crossing their borders.

At the 2016 Rio de Janeiro Olympics, sensitive to the problems of immigrants and refugees, Thomas Bach made the bold proposal to include a team of refugee athletes who had been displaced from their homelands and had no NOC for which to compete. There were 10 refugee athletes competing in Rio, with 6 men and 4 women, and although they won no medals, they competed proudly, represented themselves well, and sent a message to the world. The only negative for the team was the 3-letter abbreviation chosen for them, ROT (Refugee Olympic Team), which has a rather negative connotation in English.

So this put us to wondering, how often have immigrants competed at the Olympics for countries other than their native one? Over the last 20 years or so, with what I have termed the Olympic diaspora, this has become much more common, or so it seems. Many African distance runners, superb athletes but unable to make their national teams, have elected to compete for other countries, often Gulf State countries such as Bahrain and Qatar.

And in further news about this development, the IAAF has recently ruled that they are studying this question, which they believe to be a problem, and such national transfers are on hold until a working committee finishes their study of this situation. FIFA also basically bans this practice as you elect one nation at the beginning of your senior international career, and you cannot change it after that.

So how often do immigrants compete at the Olympics for a nation other than their native land? This is not easy to answer. Our database includes all Olympic competitors, and it also has a field for ethnic nation, but we don’t include ethnicity for all athletes in all cases. There are a number of caveats in how we have to look at this and, as a result, some limitations on the data we are about to present. In academic papers, which I deal with daily as a medical editor, limitations to a study are usually given at the end, but we’ll present them here at the outset.

Since we can’t always tell ethnicity of every athlete in every case, the surrogate data field we chose to use is nation of birth, comparing that to nation they competed for at the Olympics. We have place of birth for over 80% of Olympic athletes, and at recent Olympics, it approaches 95% of Olympians, so we have sufficient data to use this as the surrogate.

However, that is not perfect, the best example probably being Olympic marathon champion Frank Shorter, who was born in München, Germany, where he won his 1972 Olympic gold medal, but was definitely an American national, his father having been in the military in Germany when he was born. Another similar non-Olympian was John McEnroe, also born in Germany, but of American military parents. A few of these will creep in but we don’t think it will affect the numbers much, as our datasets are large, and by the large of law numbers, this should have little effect.

Next, from our database we develop two fields “nation of birth” and “nation of competition” and pull “immigrant athletes” out whenever these two fields do not match. But a problem here arises with the break-up of the Soviet Union in 1991-92 into its constituent republics; Germany splitting into West and East Germany (FRG/GDR) in 1945 and then re-combining in 1990; and similar merging and splitting with Yugoslavia, Czechoslavia, Vietnam, and Yemen. So we then eliminated all of these possibilities. If the merged field was FRG-GDR, we deleted that record as not an immigrant athlete. Similarly for RUS-URS, YUG-SRB, etc. But you’ll see that this does not totally solve the problem. This still leaves athletes such as URS-KAZ, which are similar, but there’s a limit to what we can delete ad hoc.

When you’re dealing with over 200 nations there are always going to be anomalies. Israel is dealt with elsewhere in this article but there are others. Ireland’s proportion of 21 % of Olympians born outside its borders is also misleading as the vast majority were in fact born on the island of Ireland albeit in the British-governed north. Under a provision of the Irish constitution they had, until recently, an automatic right to Irish citizenship.

Great Britain could in fact choose any subject of the British Empire to compete at the Olympic Games until 1952 while US citizens can represent any one of its dependent territories (e.g., Guam, American Samoa) unless the NOC involved has further residency or ancestry rules.

Liechtenstein’s tally has a more practical reason. Half of all new mothers from the principality decide to travel abroad to give birth, mainly to border towns in neighboring Switzerland and Austria.

And Indonesia, formerly the Dutch East Indies, and the Netherlands are another combination that probably should be considered not to constitute an immigrant situation.

Finally, our database lists placenames as they exist today. Thus we list Moskva (Moscow) as Moskva, Russia, not Moskva, Soviet Union, even if the athlete was born in Moskva when it was part of the Soviet Union. By the previous paragraph note, this will get thrown out, but it may make some minor differences in the data.

Given all that, our numbers are pretty large. There are almost 135,000 Olympians, and we thus have places of birth for about 109,000 athletes.

The questions we want to look at are: 1) how often do immigrant athletes compete at the Olympics?; 2) which nations have had the largest number of immigrant Olympians competing for them, and which nations have sent the most immigrant Olympians to other nations?; 3) which sports have the highest percentage of immigrant Olympians; 4) is this increasing in frequency or are they any trends over time?; 5) is there any difference between Summer and Winter Olympians?; and 6) is there any difference between male and female Olympians?.

In all below IA = immigrant athletes, keeping in mind the above caveats that this means athletes born in a country other than the one they competed for; and OPB = Olympians for whom place of birth is known. Further we have absolute numbers for all the data, but the more important data are the percentages, since athletics (track & field) has had over 18,000 OPB, while synchro swimming has had only 459 OPB. And because of some small numbers for certain sports and nations, we’ll look mostly at the bigger nations and sports. With smaller numbers, the signal to noise ratio is affected much more strongly.

Here is the overall data – since 1896, there have been approximately 8,100 IA out of 109,000 OPB, or a percentage of 7.4% IA. That means about 1 of 13 Olympians can be considered an Immigrant Athlete.

Which nations have had the largest number of immigrant Olympians competing for them, and which nations have sent the most immigrant Olympians to other nations? Well, the Marshall Islands has had 7 OPB, and none of them were born there, so they have 100% IA, but again the small number of athletes affects that. Among nations with over 95 Olympians the leading nations, by percentage, are as follows:

Qatar 39 96 40.6%
Israel 110 281 39.1%
Ireland 138 634 21.8%
Liechtenstein 23 110 20.9%
Azerbaijan 25 126 19.8%
Canada 637 3978 16.0%
Cyprus 14 96 14.6%
Greece 213 1467 14.5%
New Zealand 145 1066 13.6%
Hong Kong 35 264 13.3%

But if you look at that, some of those countries have problems with our caveats. Israel gets many Jews from other nations, and Cyprus and Greece likely cross-pollinate, while Azerbaijan is still a former Soviet Republic that makes its numbers suspect, Hong Kong will have cross-over with China, and we mentioned Liechtenstein above. The most impressive country is probably Canada, with 16% of their OPB having been born in other countries, so a lot of Olympians come to Canada either to live there or to compete for them.

Qatar, however, leads, if we limit this to nations with over 95 OPBs, demonstrating how many of their athletes come from other nations, mostly Africa. Bahrain is not on the list because it does not have enough athletes to qualify, but Bahrain has had 76.9% IAs among their Olympians, a huge percentage, reflecting the way they pull athletes to compete for them from elsewhere.

Since small numbers can cause data problems, let’s look at nations with over 1,000 OPBs.

Canada 637 3978 16.0%
Greece 213 1467 14.5%
New Zealand 145 1066 13.6%
Australia 348 3132 11.1%
Austria 166 1878 8.8%
Germany 268 3504 7.6%
United States 570 7918 7.2%
Spain 136 2176 6.3%
Italy 246 4062 6.1%
Great Britain 310 5240 5.9%

Canada is again the most impressive with the most IAs competing for them. The other nations tend to be among the Western European countries, along with Australia and New Zealand, and the USA. Note that Germany is listed, and that is without counting FRG/GER, or GDR/GER doubles, as we noted above.

So which countries have sent the most Olympians to compete for other nations? Again, we’ll look at larger countries for better data.

Georgia 38 112 33.9%
Serbia 40 126 31.7%
Jamaica 78 287 27.2%
Czech Republic (Czechia) 122 518 23.6%
Ethiopia 41 186 22.0%
Côte d’Ivoire 21 102 20.6%
Malaysia 50 246 20.3%
Trinidad & Tobago 23 137 16.8%
Kenya 66 405 16.3%
Morocco 59 379 15.6%

Once again we have the problem with Georgia and the Soviet republics, Serbia and the former Yugoslavia, and Czechia/Czechoslovakia. Note that Jamaica is high on the list, as many of their sprinters go elsewhere to compete, often Canada or the United States, as it’s so difficult to make the Jamaican Olympic sprint team. Ethiopia shows up at 22.0% with many of their distance runners competing elsewhere. Not listed in the 10 nations above, but Kenya shows up at #24 with 16.3% of their native born running for other nations.

If we look at the larger nations only, with more than 1,000 OPBs, here’s what we get:

Germany 425 3504 12.1%
Russia 192 1750 11.0%
Great Britain 527 5240 10.1%
China 197 2108 9.3%
United States 604 7918 7.6%
Canada 300 3978 7.5%
Romania 110 1522 7.2%
Cuba 57 1068 5.3%
Switzerland 126 2372 5.3%
New Zealand 55 1066 5.2%

Russia and Germany again skew the data. Some of the nations are Western European, but also with Canada, USA, and New Zealand. And China, which exports so many table tennis players, also shows up high on this list.

Another other question arises – what are the most common nation combinations? Here are the top 11 of those, if we eliminate the politically connected nations:

Competed For Born In #Combos %Possible
Canada Jamaica 45 57.7%
United States Norway 35 50.0%
Greece Georgia 19 50.0%
United States Japan 14 42.4%
United States Ireland 21 37.5%
United States Mexico 9 33.3%
Turkey Bulgaria 14 28.0%
Germany Kazakhstan 9 27.3%
Bahrain Ethiopia 11 26.8%
Canada Norway 17 24.3%
Bahrain Kenya 16 24.2%

The interesting thing here is that the Canada-Jamaica nexus is the leading combination, but look at the bottom of the list with Bahrain receiving approximately 25% of its Olympians from both Ethiopia and Kenya.

So which sports have the most Immigrant Athletes by percentages? Here is that table for the larger Summer sports, those with more than 500 OPBs:

Sport IA OPBs OPB%
Table Tennis 159 648 24.5%
Tennis 131 1045 12.5%
Baseball 73 616 11.9%
Wrestling 390 4082 9.6%
Equestrian 181 1938 9.3%
Badminton 63 694 9.1%
Fencing 306 3393 9.0%
Modern Pentathlon 58 711 8.2%
Hockey 256 3160 8.1%
Basketball 222 2800 7.9%
Athletics (Track & Field) 1434 18143 7.9%

Not too many surprises here. Table tennis leads because so many Chinese go elsewhere to play internationally because the level of table tennis competition is so high in that country (see our previous post on this – http://olympstats.com/2015/04/25/one-in-every-five-table-tennis-olympians-is-chinese/). Tennis is second, which may surprise. Baseball has had a lot of IAs, and remember that in 2004 Greece, which fielded a team as host country, had to virtually import its entire team from the USA to make up a team, and that has occurred before for other host nations in team sports. Athletics is near the bottom of the list here, of those we chose to list, and is in the middle overall. But look how many athletes in that sport – over 18,000. The sheer size of the sport makes it difficult for the percentages to get that high, despite the Kenyans, Ethiopians, and Jamaicans going all over the globe to compete.

For Winter Sports, you will note that there is not much of a difference between sports:

Sport IA OPBs OPB%
Figure Skating 136 1273 10.7%
Ice Hockey 308 3147 9.8%
Skeleton 11 119 9.2%
Freestyle Skiing 44 508 8.7%
Alpine Skiing 185 2215 8.4%
Biathlon 76 919 8.3%
Bobsledding 108 1372 7.9%
Cross-Country Skiing 135 1946 6.9%
Nordic Combined 30 490 6.1%
Short-Track Speedskating 22 360 6.1%

There’s not really much to differentiate the sports. Russian figure skaters, especially in pairs and ice dance, have gone elsewhere to compete, given the high level in those disciplines in Russia. There are the famous cases of the ice dance couple, the Duchesnays, born in Canada but chose to compete for France; and in 1976 Dianne de Leeuw, a Californian with dual Dutch citizenship, chose to compete for the Netherlands when she was young, not certain if she could make the US team. She won a silver medal at the 1976 Innsbruck Winter Olympics for the Netherlands, so those fears were unfounded.

Is any of this changing over time? Here are the numbers for the Summer and Winter Games since World War II:

Summer IA OPBs %OPB
1948 206 3177 6.5%
1952 247 3896 6.3%
1956 202 2675 7.6%
1960 321 4263 7.5%
1964 265 4212 6.3%
1968 348 4613 7.5%
1972 293 5976 4.9%
1976 227 5162 4.4%
1980 96 4523 2.1%
1984 234 5914 4.0%
1988 264 7439 3.5%
1992 411 8354 4.9%
1996 580 9308 6.2%
2000 707 9690 7.3%
2004 880 9715 9.1%
2008 805 10138 7.9%
2012 921 9889 9.3%
2016 967 10632 9.1%
Winter IA OPBs %OPB
1948 34 521 6.5%
1952 30 548 5.5%
1956 34 657 5.2%
1960 41 539 7.6%
1964 46 897 5.1%
1968 54 963 5.6%
1972 30 847 3.5%
1976 36 960 3.8%
1980 36 922 3.9%
1984 48 1108 4.3%
1988 63 1253 5.0%
1992 92 1603 5.7%
1994 134 1564 8.6%
1998 165 1984 8.3%
2002 178 2210 8.1%
2006 169 2319 7.3%
2010 159 2384 6.7%
2014 150 2612 5.7%

If we look at the Winter Games first, there’s not much to offer. There has been some mild fluctuation over time, notably in the 1990s, and that is probably related to the break-up of the Soviet Union and Yugoslavia, but otherwise no real trend.

But if you look at the Summer Games, there is a trend and it is increasing. The 1990s here do not show much of a jump, but look at the 21st century. The percentages have been 9% or more at three Summer Olympics since 2000, and that had not occurred since well before 1948. So this seems to support the athletic diaspora that has started to occur in recent decades, based on African distance runners, Chinese table tennis players, Jamaican sprinters, and others.

How about differences by gender and seasons? Nothing to really speak of, as the numbers, given below, are pretty equal:

Gender Season IA OPBs OPB%
F S 1808 24330 7.4%
M S 5194 69677 7.5%
F W 288 4098 7.0%
M W 911 10967 8.3%

So those are the numbers. Some conclusions can be drawn, despite the difficulty in pulling out the wheat from the chaff in the data. The athletic diaspora at the Summer Games is definitely real and increasing since 2000. Among large nations, more immigrant athletes tend to go to Canada to compete than any other large nation. Among sports, table tennis has the biggest immigrant athlete problems, with Chinese players going all over the place.

Are these problems? Do they need solutions? That can’t be answered directly by the data, although the increased trend of more immigrant athletes in the 21st century does seem concerning. It will be interesting to see what the IAAF does with their study group and if other IFs will follow that lead.

(This post is a combined effort of Bill Mallon and Jeroen Heijmans, with input also from Hilary Evans.)



One thought on “Olympian Immigrants”

  1. Very interesting. Complicated. Had to read twice and may return to insure total understanding. I did notice that Luge was left off the Winter chart.

Leave a Reply

Your email address will not be published. Required fields are marked *