In 2003, sprinter Dwain Chambers was caught taking the banned steroid THG. He served a two-year ban, and has now turned informer to the anti-doping authorities and returned to the track. Earlier this week he ran the fastest 100m by a British athlete this year, and he intends to win a place on the British Olympic team at the trials just over a week from now. He looks to have a great chance.

The problem is that the British Olympic Association (BOA) has a policy of not selecting athletes that have failed drug tests, whether they’ve served their bans or not. No matter how well Chambers does in the trials, the BOA won’t pick him unless they’re forced to.

Chambers will challenge the legality of this policy in court, arguing that it constitutes an illegal restraint of trade. He hopes to get a ruling from the court saying that the BOA have to pick him if he does well enough at the trials.

In response to Chambers’ challenge, the British Athletes’ Commission (BAC) has taken up a petition calling for the selection policy to stay in place. Over a hundred athletes have signed, including big names such as Steve Redgrave and Kelly Holmes. The BAC’s Peter Gardner explained, “The petition carries weight because it has support from many athletes.”

He’s wrong. The petition doesn’t carry much weight at all.

The question here concerns a point of law: Does the BOA’s policy of not selecting athletes that have failed drug tests constitute an illegal restraint of trade? The petition only tells us that certain people approve of the BOA policy; it doesn’t tell us anything about whether it is legal or not. The policy may be popular, perhaps even have near unanimous support among athletes, but still break the law. This appeal to popularity shouldn’t persuade the court.

The presence of big names on the list doesn’t change this. Holmes and Redgrave have plenty of Olympic gold medals between them, but being a faster runner or rower than anyone else doesn’t give you much insight into employment law. To suggest that Chambers’ ban should be upheld because some great British Olympians say so would be an inappropriate appeal to authority.

I sympathise with the athletes that have signed the petition: I hope that the selection policy is legal and can be maintained. But the petition itself isn’t evidence that should sway the court.

A recent poll has shown that the public continue to support women having the right to choose to have an abortion. The poll was commissioned by Bpas, a charity that provides around 50,000 abortions each year, and conducted by Mori. 63% of those asked agreed with the statement, “If a woman wants an abortion, she should not have to continue with her pregnancy.” Bpas have reportedly cited the data collected as a reason to liberalise current abortion legislation.

There is always a concern with surveys commissioned by a group with a clear vested interest, such as this one: the way that the questions were framed might have distorted the results. Interestingly, some of the questions in this poll were varied, making it possible to see this phenomenon in action.

Half of those polled were asked about their attitude towards the current law which allows abortion up to 24 weeks with the consent of two doctors. The other half were asked about their attitude towards the current law which allows abortion up to 24 weeks but were not told about the requirement for doctors’ consent. 54% of the first group agreed with the current law, while 28% opposed it; 46% of the second group agreed with the current law, while 38% opposed it. This serves to illustrate how much difference the way that questions are framed can make.

The conclusion drawn from the data by Bpas is also of interest. If their argument were that the public endorsement of abortion implies that there is nothing wrong with it (call this “the moral argument”), then it would clearly commit the appeal to popularity fallacy. The moral status of an action and public opinion towards it are two different matters. If the majority of the population backed racial discrimination, for example, then that wouldn’t make it right.

If, on the other hand, Bpas’s argument is not to do with the moral status of abortion but with the legal status of abortion (call this “the legal argument”), then they may be on more solid ground. The argument, “The public backs abortion on demand, therefore the government should permit abortion on demand”, need not be understood as drawing a conclusion about whether abortion is morally right or wrong. Instead, it could be understood as drawing a conclusion about how the government should legislate.

Why is the second argument better than the first? Because it rests on a more plausible assumption.

The moral argument, in inferring the conclusion “There is nothing wrong with abortion” from the reason “Most people believe that there is nothing wrong with abortion”, makes an assumption: “Whatever most people believe is the case.” This assumption, however, is often false; often the majority opinion is incorrect. This is why appeals to popularity are fallacious.

The assumption made by the legal argument is slightly different. The legal argument infers the conclusion “The government should permit abortion on demand” from the reason “Most people want the government to permit abortion on demand.” The assumption made by this argument is that the government ought to do whatever most people want the government to do. This assumption is not obviously true–we can imagine situations in which following public opinion would lead to catastrophe–but it is more defensible than that made by the moral argument. In a democratic system, where the government is elected to rule for the people, the legal argument might just fly.

One thing that this shows is the importance of clarifying precisely what an argument is before it can be assessed. Several different conclusions are attributed to Bpas in several different places: e.g. that abortion should be available on demand, that the current legislation should be relaxed, and that it is time for the government to review abortion legislation. The stronger the conclusion they draw from the data, the weaker their argument gets.

It is quite often claimed that the ills that religion brings outweigh any good that it might do. This year, provocatively, The Guardian decided to mark Christmas by slapping precisely this claim across its front page. “Religion Does More Harm Than Good - Poll” was the headline as people across the country geared up to celebrate God becoming incarnate.

Of course, all polls have their problems. For a start, they are better as guides to public opinion than as guides to the facts (the appeal to popularity is a fallacy, remember). Then there are all of the difficulties involved in getting a representative sample of a sufficient size to be meaningful, and formulating questions that don’t have too much of a distorting effect on people’s answers.

The biggest problem with the article in The Guardian, however, is that the poll simply didn’t say what The Guardian said it did; they misrepresented the data.

The survey reported that 82% of those asked see religion as a source of tension. There is a big difference, however, between saying that a thing is a source of tension and saying that it does more harm than good. FA Cup finals are a source of tension. General elections are a source of tension. A-levels are a source of tension. That doesn’t mean that these things do more harm than good.

It may well be that most of the 82% who see religion as a source of tension also believe that tension to be trivial in comparison with the great goods that religion brings. That is, it is perfectly possible to answer “yes” to the question in the poll without endorsing The Guardian’s headline claim.

Thankfully, The Guardian didn’t make its case just on the 82% who see religion as a source of tension; they made a comparison with the 57% who said that religion is a force for good: “an overwhelming majority see religion as a cause of division and tension - greatly outnumbering the smaller majority who also believe that it can be a force for good.” (Note the subtle manipulation here; 57% said that religion is a force for good, but The Guardian weakened this to “can be a force for good.”)

The idea, then, is that because more people said that religion is a source of tension than said it is a force for good, the poll supports the idea that religion does more harm than good.

Presumably this is because to get these figures there must be some people who think that religion does some harm but no good (and so that it does more harm than good). But consider the figures carefully. What is the greatest possible number of people who did say that religion does harm but didn’t say it does good? 43%. A minority.

That means that for the poll to support the idea that religion does more harm than good, it must be assumed that a significant number of those who said that religion is both a source of tension and a force for good think that all things considered religion does more harm than good. Now that may be the case, but that’s a pretty strong assumption. Describing something as a “force for good” sounds a lot like saying “all things considered it does more good than harm”, whereas describing it as “a source of tension” sounds nothing like saying “all things considered it does more harm than good.”

The data simply don’t yield The Guardian’s headline claim. The Guardian, perhaps to try to cause a stir at Christmas, was putting words in people’s mouths.