Historically French Canadians have been underrepresented on English Canadian NHL teams, relative to their representation on U.S.-based teams.
The underrepresentation of people from Quebec is particularly dramatic behind the bench, in the coaching staff.
Here is a xlsx spreadsheet and here is a csv one listing everyone who has ever coached in the NHL, their place of birth, and their coaching statistics. (The coaching statistics were drawn from hockey-reference.com.) The most difficult part was deciding who to classify as francophone. For example, does "the Brian Mulroney of hockey", Pat Burns, count as francophone? (I excluded him). What about Jacques Martin, who was born in Ontario and has spent most of his career coaching English-Canadian or US teams? (I included him). For the most part I classified someone as Francophone if they were born in Quebec and had a traditionally French-Canadian name; for people outside of Quebec I looked for a French Canadian first name and last name, plus some time spent coaching in Quebec. Please let me know if you think I have misclassified any coaches.
What is immediately obvious from the data is that NHL coaching staff is overwhelmingly drawn from English Canada, or "Rest of Canada". Of the 244 North American born coaches whose careers began in the post-expansion era (1968 or later) 12% were born in the US, 14% in Quebec, and 75% in the Rest of Canada (the numbers don't add to 100 because of rounding). The 5:1 Rest-of-Canada to Quebec ratio is far in excess of what one would expect based on, for example, the approximately 3:1 Rest-of-Canada to Quebec ratio for the population as a whole. Moreover, a substantial proportion of those Quebec-born coaches – for example Scotty Bowman and Bryan Murray – are Anglophone Quebeckers.
One explanation of the underrepresentation of people from Quebec is employer or customer discrimination. For example, a Saskatchewan-born player like Tiger Williams might sell more tickets than a Montreal-born player like Roberto Luongo. Alternatively, it might be that people born in Quebec are different in some way – size, training, attitude, linguistic ability – and the differences that we observe between the Rest of Canada and Quebec simply reflect those unobservable factors.
The standard way of testing between these two explanations is to compare the on-ice performance of the two groups of players. If there is discrimination against French Canadians, then a French Canadian player will have to be better than an English Canadian to earn a place on a team’s roster. A study carried out in the mid-1980s by University of Ottawa economists Marc Lavoie, Gilles Grenier and Serge Coulombe found exactly that: French Canadian and European forwards and defensemen typically performed better than their English Canadian and American counterparts, although subsequent research by Neil Longley has disputed those findings.
But what about coaching? If Francophone or Quebec coaches face discrimination in the hockey world, then they would have to be better than their rest-of-Canada counterparts to earn a place behind the bench.
A proper analysis of coaching success would look at numerous factors – for example, the quality of the team coached, length of time with the team, and so on. I haven't done any of these things. I've just calculated some average play-off stats – the playoff win-loss percentage, and the average number of Stanley Cup wins over the course of the coach's career.
| Quebec-born | Quebec-born francophone | Francophone | Rest of Canada | US | |
| Playoff win-loss percentage |
42.5 (3.6) |
46.3 (3.8) |
46.4* (3.3) |
37.9 (1.6) |
43.1 (2.9) |
| Career Stanley Cups (average) | 0.42 | 0.18 | 0.17 | 0.12 | 0.14 |
The number that I find most telling is the playoff win-loss percentage. The career Stanley Cup numbers are somewhat distorted by Scotty Bowman's 9 cup wins, which brings the Quebec-born coach numbers up substantially. Still, the figures presented here are far from conclusive. The standard errors shown in percentages indicate that the differences are, for the most part, not statistically significant – except for francophones coaches, we cannot rule out the possibility that this is just random variation.
In tonight's game we have a rare sight: the match-up of two teams, both of whom have Francophone coaches. Will the series go to Vancouver coach and Quebec City native Alain Vigneault? Or to the Boston coach, Blind River Ontario native Claud Julien?
Worthwhile Canadian Initiative 
Is the 5:1 Rest-of-Canada to Quebec ratio far in excess of what one would expect relative to the population of youth hockey players?
Also, a very easy adjustment to make would be to inflate the stats of coaches who were behind the bench when the playoffs were a much shorter experience – a maximum of 21 games in the Original Six era with a winning coach accumulating 8 wins versus 105/16. Right now I would say you’re overweighting modern coaches in the sample due to the expansion of the league. That won’t unbias you, but it will jump up the variance a lot.
Andrew, it’s a win/loss percentage so it shouldn’t be affected by the length of the playoffs or the number of years coaching.
The population of youth hockey players and the population of potential NHLers is different – there are lots of youth hockey players in cities, but NHLers come overwhelmingly from small towns with abundant outdoor ice and restricted non-hockey job opportunities. There’s lots of natural ice in Quebec, so if anything I’d think the overall population ratio under-estimates the extent of the skewedness. Though having said that, the population, demographic and ethnic structure of Quebec/R of C has been changing throughout the post-expansion period. And of course the pool of coaches in the post-expansion period reflects to some extent the pool of players in the pre-expansion period, and there are a whole number of issues to deal with there, for example, Montreal’s preferential rights to draft Quebec players (I don’t remember when that ended), and some of the really nasty stuff that went on back then. (I was reading some Hockey News article where, I think, Brian Kilrea was reminiscing about an incident early in his career. A coach, who decided that his goalie was going down too early, tied a noose around the goalie’s neck and put him in net – a good way of breaking the going-down-too-early habit…)
But, yes, it would be an excellent idea to get stats on the pool of potential NHL coaches/players.
I did a couple of papers on hockey card prices a while back (unpublished as yet), and if I remember correctly there was something along the lines Frances is suggesting. I didn’t compare the numbers to population, but there did seem to be a selectivity thing going on – you had to be better than the English to make it. I think Mark Lavoie referenced it in one of his papers.
Out of curiousity, how is it that the average playoff win/loss percentage is less than 50%? That would seem to suggest that we should be hiring coaches from outside of Quebec, Canada, the US and not francophones, since someone is beating these guys 55-65% of the time!
Bob, good point. A couple of things:
At one point there were ties allowed in the playoffs. But I don’t think that’s it.
I excluded everyone whose coaching career began pre-1968, just because things were so much different back then. My best guess is that this group is the ones beating up on the coaches in my sample – the established coaches of the original-six coaches who were wiping the floor with the expansion franchises back in the 1970s.
It’s an arbitrary cut-off – and of course it would be much better to work with year-level data so the kind of asymmetry that you spotted doesn’t arise – but there were so few francophone coaches in the pre-quiet revolution, pre-expansion years – the only ones I identified were Pit Lepine and Percy LeSueur – that it just didn’t make any sense to include that period.
I dug through the numbers a bit more. There’s 11 coaches who have playoff win/loss percentages of zero. At first I thought these were people I’d miscoded, but then I checked – these are people who coached for one season and then were wiped out in the first round of the playoffs, so ended up with a zero win/loss percentage.
So that’s the explanation – each coach gets the same weight, whether they coach for one season or twenty.
Imagine one team has the same coach for 10 seasons, and that coach wins every play-off game. Another team has a different coach every season, and that coach loses every play-off game. If we look at the coaches, what we’ll see is 11 coaches, 1 with a w-l% of 1.0, and 10 with a w-l % of 0.0, so the average w-l% is around 9%.
Frances, the same size is a bit small but the question is interesting. After all, Justin Wolfers got a QJE for showing racial discrimination in the NBA.
As a next step,
did you think of looking at race team composition (depending on the coach) and minutes played. Maybe you can find also referee discrimination like wolfers…
I think the coach / player interraction is interesting to show discrimination… A bunch of subquestions can be answered. eg: How much can you sacrifice team success to discrimination?
Could it simply be the language barrier kicking in? For a Quebecer to coach a team, he has to have an at least decent English. The same goes for Europeans whose native language isn’t English.
BorrorO – yup, I think language is crucial.
John, yes, game-level and team-level data would be much better than what I have right now. The link under “historically” suggests that the underrepresentation of francophones is most pronounced in English-Canadian teams.
Can I nitpick a bit about your identification scheme for discrimination?
” If Francophone or Quebec coaches face discrimination in the hockey world, then they would have to be better than their rest-of-Canada counterparts to earn a place behind the bench”
Really? Suppose they were statistically just as good yet still underepresented relative to their representation in the population. Isn’t that evidence of discrimination? After all, the idea wouldn’t be that we expect everyone to discrimintate, just that it happens on average.
For example, I don’t think you need to show that all the best engineers are female to say there may be some discrimination going on. There are so few that all you really need to show is that they are competent.
Furthermore, I think you sort of reverse cause and effect. Given that discrimination is happening you’d expect the under represented to group to outperform since those hired come from the top of the distribution for that group. I don’t think that is further evidence of discrimination.
Adam P, “Given that discrimination is happening you’d expect the under represented to group to outperform since those hired come from the top of the distribution for that group.”
Yes, that’s the point. If one finds no difference, then it’s pretty hard to argue that there is, in fact, discrimination going on.
Your point, however, about the fact that I don’t have information on the pool of potential NHL coaches is a good one. It might be, for example, that Quebec-born NHL players have more lucrative off-ice job opportunities than Rest of Canada born players. I’m not too worried about this, however, because the numbers are just so far out of line with what one would expect based on population.
I need to go out and buy Bob Sirios’s book in which he gives all sorts of stats that he claims prove discrimination against francophones in the NHL. And troll through Gaberiel Desjardin’s web site, which is a trove of hockey trivia.
Frances: last time the subject came out
http://www.sportsnet.ca/hockey/2010/10/25/nhl_discrimination/
http://www.barakabooks.com/catalogue/discrimination-in-the-nhl/
(the book by Bob Sirois updating Lavoie, Grenier, Coulombe ( Serge is a high-school , cegep and university buddy, so he must be good)
on the CBC, Peter Mansbridge merely said “Usual sour grapes from Québécers”. There are ladies on this blog so I won’t quote the national affront to color coordination…but you can see a lot of threads on this google search.
http://www.google.ca/search?q=nhl+discrimination&rls=com.microsoft:fr-ca:IE-SearchBox&ie=UTF-8&oe=UTF-8&sourceid=ie7&rlz=1I7GGLD&redir_esc=&ei=F1z2Tby_OoXZgAfn-oDyCw
Jacques Rene, thanks. It’s interesting that there’s been so much more attention paid to this on the player side rather than the coaching side.
I think that part of the problem is that, as Adam P points out, even when one can find some evidence of differences between francophone and Rest of Canada coaches, one can’t conclude that it’s discrimination. The thing is, a coach has to be able to speak English, so francophone coaches are also bilingual coaches. Their superior performance might be because they’re bilingual, not because they’ve had to overcome discrimination.
B.t.w., Jacques Rene, if you spot any problems with my coding of francophones, please let me know. I know that some would consider Pat Burns francophone, for example.
I just read in Le Devoir
http://www.ledevoir.com/politique/canada/325107/bilinguisme-a-sens-unique-au-npd
that francophone NDP staff in francophone ridings must be bilingual. Of course, no anglos need sweat about needing to speak french. So what about their support for bilingual judges on the Supreme Court?
Frances, yes your right if we think the distribution of talent is the same in both groups.
However, suppose that the minority group has the same mean ability but a (much) lower variance? (Minority and majority aren’t to be taken to literaly here.)
In the absence of discrimination should they be under-represented? I’m thinking no, seems to me they’d simply displace those from the majority group who were at the bottom of the distribution. Nonetheless, in this case those that were employed would still tend to underperform because they’d still be inferior to the best from the majority group.
In the presence of discrimination they would be under-represented but still, those that were employed would not tend to be the top performers. It would just be the case that the best of the minority group was still inferior to the very best of the majority group.
Adam P “However, suppose that the minority group has the same mean ability but a (much) lower variance?…In the absence of discrimination should they be under-represented?”
Ah, the Larry Summers argument – men and women have the same mean ability, but because women have lower variance, they are under-represented in the extreme tails, i.e. science and math at Ivy League schools.
Getting into the NHL is incredibly difficult, IIRC one of the articles I’ve read today said that there were 45 (francophones? Quebecois?) playing in the NHL right now – that’s out of a talent pool that, by any reckoning, would be in the six to seven figure range. So a lower variance among Quebec players could explain their underrepresentation.
People arguing against the presence of discrimination in the hockey league point to size differentials (those 6’4″ Saskatchewan farm boys who are as broad as they are tall.) Arguably there is more variation in size among the Rest of Canada population than among the francophone population because of greater ethnic diversity.
On the other hand, those arguing that there is discrimination point to the fact that, when ability can be reasonably objectively measured, i.e. for goalies, the francophone disadvantage disappears.
Jacques Rene – Let’s not go there.
One point that is interesting on language, however, is that in European soccer there doesn’t seem to be this idea that coaches must speak the same language as their team – the coach market, as well as the player market, is increasingly global.
I think an important side point is that almost all of the francophone coaches got their start with the Montreal Canadiens and went on to find bigger success elsewhere. Not a bad place to start if you want to build up a temper quickly. The francophone coaches were all positively discriminated for their ability to speak the language.
Quite clearly, no other organization “takes a chance” with Francophone coaches, but I wouldn’t be surprised to find that the Canadiens are forced to take more chances than other teams because of the smaller recruitment pool. Let’s say it’s extremely rare that a coach makes a comeback to the same city and when all the Francophone coaches were minted in your team… We could verify this by comparing the number of first time NHL coaches hired by different teams.
As for the inevitable judge comparison, the Montreal Gazette has (of course) been there before: http://www.montrealgazette.com/opinion/language+sports+translates+courts/4807097/story.html
(I won’t comment, judge the article yourself)
A good book about the Jewish experience in sports
Plus ça change…
Guillaume: “Quite clearly, no other organization “takes a chance” with Francophone coaches,”
Actually, Jacques Demers got his start coaching in the US, Alain Vigneault came from the QJMHL to the Ottawa Senators, Jacques Martin started with the St Louis Blues.
Bob Hartley, who is a Franco-Ontarian (and not actually on my list of Francophones, though if he had been he would have brought the francophone numbers up with a 0.58 win/loss playoff percentage) got his start in the US also.
Yes, if there were more Quebec franchises, there might well be more francophone coaches. But it’s interesting that some of the research on francophone underrepresentation has found that it’s more pronounced in the English Canadian than the American teams.
I remember in my youth ( that is early professionnal career) going from YUL to LGA on Eastern . Noting to the head steward that I had better non-english services than on on AC he replied ” But you are a valued foreign customer.”
Now , of course , every customer is badly treated,whether on AC or any americam carrier.
To an american, internal canadian squabbles are a quaint habit. Same as their attitudes towards race seems to us.
It would be really interesting to look at the role of mentoring in all of this, and the tendency of coaches to hire and train assistant coaches who are like themselves. Does anyone know if anyone has looked at this?
Frances, I was channeling Larry alright, I didn’t mention him because I couldn’t remember exactly what he was talking about:)
By the way, for a good analysis of NHL players by place of birth:
http://www.behindthenethockey.com/2009/10/22/1096299/on-french-canadians-in-the-nhl.
Thanks for all of the input, here’s my Globe and Mail post with a (hopefully slightly more refined) look at the issue. The headline is not mine:
http://www.theglobeandmail.com/report-on-business/economy/economy-lab/frances-woolley/is-winning-the-stanley-cup-all-in-a-coachs-genes/article2060069/.