I was delighted to discover this week that the Times have started an innovative new column entitled “Bad Statistics”. It seems to me to be somewhat lacking in thoroughness. I should like to submit for their consideration an article from the Sunday Times on the 14th of December.
The opening sentence is: “Public opinion has moved sharply in favour of assisted suicide, according to a poll for The Sunday Times.” This opening sentence is, I believe, incorrect.
The story is based around a youGov poll of 2,000 people. “More than two-thirds (69%) think the law should be changed…. Most strikingly, by four to one (61% to 15%) people said they would consider assisted suicide for themselves if suffering from a terminal disease.”
How do these figures compare to previous population survey data on euthanasia? Luckily my friend William Lee is an epidemiologist who also researches attitudes to physician assisted suicide in the same building as me (his colleague Annabel Price first spotted this “bad statistic”). They have conducted an extensive literature review.
Combining this with Appendix 7 from the 2005 House of Lords Select Committee report on Assisted Dying for the Terminally Ill gives a fairly broad sweep of figures to determine whether public opinion really has “moved sharply in favour of assisted suicide” with the new figure of 69%.
In fact it turns out that euthanasia and physician assisted suicide are supported by around 70-80% of the general population in the UK, with remarkably similar proportions being found whether the research is funded by supporters of a change in the law, detractors, or disinterested groups.
The earliest survey for which detailed results are available is a 1987 MORI poll of 1808 respondents, sponsored by “Doctors Who Respect Human Life” and the “Human Rights Society”. The question was a bit pokey, talking about doctors “injecting muscle relaxants so as to paralyse breathing”, but still 23% said “Euthanasia should be made legal in all cases when the patient requests it”, added to 49% who said it should be legal “when a patient who requests it is suffering from a severe illness and is in a lot of pain”, which makes 72%. Higher than 69%, in 1987, with a question somewhat inviting you to say reject euthanasia
NOP have polled around 2000 people for the Voluntary Euthanasia Society regularly for many years, with results in favour of a change in the law ranging from 69% in 1976 to 79% back in 1993. The British Social attitudes survey showed similarly high figures in favour, from 75% in 1984 to 82% in 1994. It’s a marginal, slow increase at best, and hard to see how 69% could represent a “sharp rise”.
What about other youGov polls? In 2005 a YouGov/Telegraph poll found that 87% thought the terminally ill should be allowed to ask for medicial assistance to help them die. In 2004 a YouGov poll on behalf of the Voluntary Euthanasia Society found that 80% approved of the specific proposals in the Assisted Dying for the Terminally Ill Bill.
And of course there are the crap media surveys. Viewers of the BBC’s Heaven and Earth Show voted by email, text or phone on the question: “Should assisted suicide be made legal?” 73% said yes. A Teletext poll from 2001 asked if viewers would back “a terminally ill woman’s wishes to die without her husband facing prosecution?” 90% said yes. The Independent on Sunday built their survey around the case of Diane Pretty, with motor neurone disease. “Should people have the right to die when they choose?” 85% said yes. A Sky Interactive Poll asked “Euthanasia: should it be legalised?” 84.5% said yes.
Comparing the results from different surveys is a minefield at the best of times, but overall it seems to me you would be rather hard pressed to claim that 69% shows “Public opinion has moved sharply in favour of assisted suicide, according to a poll for The Sunday Times.” Especially since it seems you may have to go back to 1976 to find a single poll that gives a result so low.
As a festive gift to me, you may wish to take up the eerily familiar request at the bottom of the “Bad Statistics” column. “Please send your bad statistics to firstname.lastname@example.org“. Merry christmas.
Please send your bad science to email@example.com