The Magic Sound of Figures

What do the people of the UK think about animal research? Last year Dr Sarah Bailey of Bath University wrote an article for the Guardian newspaper in which she claimed that the “vast majority of the UK public supports the use of animals in medical research where there is no alternative.” She cited successive Ipsos MORI polls to prove it. Of course that wouldn’t in itself make the practice right, though it might give its practitioners a boost and make their work more politically manageable. But is it true anyway? The figure being relied on in that Guardian article was 65%, as reported by Ipsos MORI in its survey for 2016. Does that represent, to a scientist, a “vast majority”? Is it even (as more modestly she goes on to describe it) “over two-thirds”? I make it just under two-thirds, but unlike Dr Bailey I’m not professionally bound to get my maths right. Anyway, whatever its mathematical status, the number was not an accurate précis of the 2016 results as a whole. In fact a study of that survey for this blog suggested that 1% was the more conclusive figure (see

Anyway, now we can start again, because the latest of Ipsos MORI’s surveys has just been published, revealing what the British people thought about animal research in 2018.

On the face of it, they haven’t changed their thinking much. The very same 65% of them re-appears answering in the affirmative that basic question about medical research. However, as the survey’s editor Michael Clemence says, “exploring public attitudes in detail reveals a more complex picture”. He picks out a few of the responses which differ most noticeably from those in the two previous surveys of 2014 and 2016. There is less confidence this time in the usefulness of animal research: only 41% think it “important to human health”, as opposed to 46% in 2016 and 43% in 2014. On the other hand, there is more concern about the animals themselves: only 15% now agree with the statement “It does not bother me if animals are used in scientific research”, compared to 22% in 2016 and 19% in 2014. This, says the editor, is “the lowest score recorded not only in this current survey [by which he means those undertaken since 2012, when the wording of the questions was altered] but also in the longer-term trend since 2002.” But perhaps most strikingly, the proportion who believe, “on animal welfare grounds”, that animals should not be used for research at all, “has increased from 31% in 2014 to 35% in 2016 and 38% this year.”

Wait a minute: this 38% opposition (to all research) and the 65% approval (for specifically medical research) surely add up to more than 100. Can that be right?

Let’s try to understand the way this survey works. The selected people – that is, the “1,011 adults aged 15+ from across Great Britain” chosen to reflect the “population profile” of the nation as a whole – were interviewed “face-to-face and in-home”: a personal setting, but completely without any context for the subject in hand. No information, no examples or pictures, were there to modify the abstract nature of these questions. (The front cover of the report this year is appropriately a plain blue blank, a sort of blindfold.) In such a situation, the wording of questions becomes more than usually persuasive. Thus when the key term ‘animal research’ replaced the more graphic ‘animal experimentation’ for the first time in the Ipsos MORI survey of 2014 – as requested by the Department of Business, Innovation and Skills which commissioned it – the downward trend of approval since 2010 was abruptly corrected. The journal Science headlined its news report on that survey ‘How much does the public support animal research? Depends on the question.’ So in 2018, when offered “an outright ban of all forms of animal research”, 27% supported it, but when the same idea was put in another way, dropping the word ‘ban’ but introducing that phrase ‘animal welfare’, then, as we’ve noticed above, 38% supported it.

Presumably the phrase ‘animal welfare’ gave respondents some slight hint of the experience of real beings, a nudge towards reality (whereas the change from ‘experimentation’ to ‘research’ had taken them in the opposite direction, towards the abstract). How much more must this have been so when these people were invited to think about actual species, and asked to say yes or no to the use for research of a pageant of visualizable animals. Unsurprisingly, their permission was most willingly given for rats and mice, then with decreasing enthusiasm through fish, pigs, rabbits, and birds, to monkeys, dogs, and the great apes. This set of preferences, so the editor remarks (just as he did in 2016), is “aligned with the government statistics on procedures involving animals.” But no, it isn’t: for even the rats are approved for experimentation by only 47% of respondents, fish (increasingly common in labs) by only 21%, birds (such as the sparrows in the previous post of this blog) 16%, dogs 14%, “others” (insects? sloths? whales?), 1%. So far from aligning with official practice, then, the answers show no majority in favour of any animal research at all.

These percentages just quoted refer to research at its most photogenic, or most advertised anyway: that is, medical research “to benefit people”, the sort of research which had elsewhere in the survey scored that 65%. For “environmental research” (water quality, safety of chemicals, etc.), the results are even less favourable. Rather puzzlingly, slightly more favour is shown towards research where “animal health” is the aim: for instance, 21% rather than 16% in the case of dogs. Perhaps the feeling was that such same-species research might at least be effective (recall that only 41% thought that animal research could be informative for human health). But again we need to remember that this is not a seminar or even a focus group where information or discussion might make more clear to the interviewees the implications of a question. It’s likely that the phrase “animal health” (often a thoroughly fraudulent one, as this blog has argued) was being at least partly confused with ‘animal welfare’, rather than understood in its more usual meaning: that is, keeping animals up to the mark for commercial purposes.

How much these at-home people really know about the research they’re being asked about is obviously an important factor in their answers (we haven’t got to the end of the 65% problem yet, but it will come). Some questions in the survey specifically deal with it. The basic question here was not ‘how well-informed are you about animal research in the UK?’ but ‘how well-informed do you feel?’ Only 35% felt well-informed (probably deceiving themselves). 26% candidly felt “not at all informed”, and answers to some other questions suggested that they were quite right to feel that (for instance, 10% believed that all animal research was illegal). Only 1% didn’t know how informed they felt, and yet that’s surely a very proper condition of mind to be in when the subject is itself both covert and contentious.

Where to get information from, then? When people were asked about “animal research organisations”, the most frequently selected characteristic was “secretive” (41%): so not from there, evidently. A question about the reliability of various other sources showed that only 7% of respondents trusted those organisations like Understanding Animal Research dedicated to showing the value of animal research. Animal protection groups scored much better at 33%. But “the most trusted source on animal research”, says the editor, are the vets who work in the laboratories.

Most trusted, yes (whether wisely trusted is another matter), but that does not mean ‘trusted by most’, for in fact even the vets score only 45%. And here we come to a most telling characteristic of this Ipsos MORI document. We’re told during the editor’s commentary that “mice and rats are the most acceptable animals for research”; or that public trust in scientists “is higher than trust in the system”; or that confidence in the strictness of the regulations and in their effective application “has been maintained at reasonably high levels.” Yet in only one case here (strictness of the rules) has a majority been achieved (the rules get 51%). Can even that be called “reasonably high”? This way of characterizing the results is not, I would say, a form of partisanship (I found only one involuntary hint in the text of pro-research views on the part of the editor). It’s rather that the expectations, after twenty years of quizzing people on the subject, are very low; no public enthusiasm for animal research, or confidence in its practice or its apologists, are looked for. Reluctance and suspicion are the standard.

And that brings us back to that 38% who opposed all animal research on ‘welfare grounds’. For that does not mean that 62% (still less 65%) approved it. Some respondents were undecided or said they didn’t know; they made up about 26% of the total. That leaves 36% backing it (or perhaps opposing it on some other than animal welfare grounds). Using Dr Bailey’s rough and ready maths, one might say that only one third of people support the practice. In short, the 2018 survey, like the one before it, records a public vote against vivisection.

To repeat: that isn’t what makes the practice wrong, or for that matter right. I recall from my childhood a school debate in which a room-full of boys (age, 12 or so) argued the motion ‘A woman’s place is in the home’. A period comedy! I don’t recollect the result, but whatever it was, I do know its exact relevance to the chosen subject. You see the connection, I hope. It’s true that we can’t ask the animals what their attitudes are, but then we also are animals, so we shouldn’t have to. Asking each other what we think over and over again is just another such boys’ debate about women, and a good deal less forgivable.


Notes and References:

The title is a phrase used sardonically in Graham Greene’s novel The Quiet American (Vintage Classics edition, 2004, p.17).

Dr Bailey’s article appeared in the Guardian on 6 April 2018. The Ipsos MORI survey, Public Attitudes to Animal Research in 2018, commissioned by the Department of Business, Energy and Industrial Strategy, was published on 24 May this year. It can be read here:

The article in Science was published in the issue for 5 September 2014, and can be read here: