Are Fans More Accepting of Gay Footballers?

Sadly, Justin Fashanu is still known more for being gay than for his football.

Go to almost any match and you’re likely to hear “Player X is a poof” at some point emanating from the stands.

Or will you? Recent research conducted by Staffordshire University in England found that 93% of survey respondents felt that there was no place for homophobia in football.

Well, I call bulls**t. It’s been clear for a long time that athletes have difficulty in coming out of the closet, most often choosing to do so after their playing career has ended. As noted, the only openly gay footballer was Justin Fashanu, who was ostracized for his decision to come out and eventually committed suicide.

In addition, anecdotal evidence (including hearing it myself at matches I have attended) would suggest that homophobia is still quite alive and well amongst football supporters. Indeed, a different survey from just one year ago found that 70% of respondents had heard anti-gay verbal abuse at a match in the previous year. That same survey found that homophobia was institutionalized within English football, which only served to encourage those chanting the abuse as they felt that they would not be prosecuted or expelled from the grounds. [Full report here - PDF warning]

More importantly, at least from my perspective, is that I question the validity of the new survey. I have no particular bone to pick with Professor Ellis Cashmore (although his site is a bit commercialized for an academic), but I find the results of his efforts a bit difficult to believe given his research methods.

First, there is the matter of numbers, or what statisticians refer to as “the small ‘n’ problem” – simply put, while statistics are designed to allow us to make inferences about a population using only a sample of that population, the smaller the sample (the “n”) the more room there is for error (conversely, if you study the entire population you don’t need to make inferences at all, since you have the entire “n” available). Cashmore (and co-author Dr. Jamie Cleland, who happens to be a former Coventry City academy netminder) initially used a pilot study of 250 survey respondents and the final survey had responses from 2,000 individuals.

Considering that there are approximately 60 million people in the UK, of whom an estimated 3.6 million are gay, the response of 2,000 individuals quite possibly seriously underestimates the prevalence of homophobia. Moreover, it is impossible to determine the response rate (i.e. how many people, of those asked to complete the survey, completed the survey), although it should be noted that they are notoriously low for online surveys.

Where was the survey conducted? Did they target (either over- or under-sample) Brighton, Nottingham, or Bristol? What about London, which last year saw an increase in homophobic crime? Given that the survey was online, it was theoretically available to everyone, but it may have been more- or less-publicized in certain areas. In addition, certain individuals (e.g. football supporters) are more likely to have filled out the survey, in what statisticians refer to as a “self-selection bias.” After all, if you don’t care about football, how likely are you to fill out a survey regarding attitudes among football supporters? More importantly, those football supporters filling out the survey presumably have a vested interest in making themselves look better and respond in a manner that reflects what is referred to as “social desirability.”

Finally, what about the actual survey itself? A quick look (hey, take the survey yourself for some fun!) shows that it is composed of seven (7) substantive questions and five (5) demographic questions. If I had submitted something like that back in graduate school, my major professor would have laughed me out of the program (as an example, the last survey that I helped to design had 130 questions). It is almost impossible to gain any useful information about any sociological phenomenon on the basis of so few questions.

Small “n”, too few questions, sample bias, the likelihood of responses skewed towards social desirability? I definitely call bulls**t on these results. Cashmore even contradicts them, noting that “parroting homophobic language is a short step from bigotry” and “homophobia now [is] equivalent to the racism which blighted British soccer during the 1970s and 80s.” If those things are true, then they certainly go against his own findings. Perhaps he was trying to make the point that it is the remaining 7% of respondents who exhibit homophobic views, and that they represent a similar “loud majority” to earlier racism in football, but if so the point is made clumsily.

The fact is that homophobia (most often in the form of homophobic verbal abuse) is still very prevalent in football, and it is unlikely that we will see a player come out of the closet while still playing any time in the near future. Unfortunately, poorly-conducted research such as this only serves to gloss over that truth.