Pontifications on Poison
Being some ramblings on events associated with poisonous plants.
Thursday 9th February 2012
I’ve written before about Kevin Sabet who was a senior adviser in the White House Office of National Drug Control Policy from 2009 to 2011 and now portrays himself as a rational centrist on the question of drug policy rather than an outright prohibitionist.
He came to my attention again as a result of two tweets about recent polling data on the matter of whether Cannabis sativa, marijuana, should be legal in the USA.
There was quite a lot of media interest in a recent Gallup Poll that found 50% support for legalisation. This was said to be a ‘record high’ in terms of support for an end to prohibition. Sabet’s tweets say ‘that Gallup was an anomaly’ He then cites other polls that have reported lower support and provides a link to a site that brings all the different polls together.
It is often said that there are two factors that can influence a poll result; how the question is phrased and who is asked the question. There is, though, a third factor that often gets ignored. Before coming to that I want to make a brief diversion into Scottish politics to show the effect of the first two.
For some time, there has been regular polling of whether the Scots want their country to be independent from the rest of the United Kingdom. The issue has been much more in the news recently as a result of debate about the timing of a referendum on the issue and what question should be asked.
Polls have usually indicated around 30-40% in favour of independence but then, on 29th January, the Sunday Express published a poll showing 51% in favour. What was significant about that poll was that it asked the exact question the ruling Scottish Nationalist Party wants to put on the ballot. Polling experts had criticised the wording as being biased towards a positive result and the Sunday Express poll seemed to confirm that by being several percentage points away from all the other polls. The wording of the question is, therefore, very important.
Then, only two days later the Scottish Daily Express published a poll showing 76% opposed to independence. The huge contrast between the two results was quite easily explained by the fact that the 76% figure was based on the number of people who had taken the time to phone a special number to give their view. So, the people involved in a poll can make a huge difference to its results.
But, what about that third factor I mentioned? This is the questions that have been asked ahead of the reported point. There’s a famous sequence from an episode of the BBC TV political sitcom ‘Yes (Prime) Minister’ where the civil service mandarin, Sir Humphrey, gets the politician to agree and disagree with the same question simply by changing the questions asked in advance.
With the latest American marijuana polls both Gallup and CBS News asked the exact same question ‘Do you think the use of marijuana should be made legal, or not?’ and though, clearly, they polled different individuals they used the same proven techniques for obtaining a representative sample.
How then to explain that CBS found 40% in favour of legalisation and Gallup 50%?
It’s actually quite difficult to sort through both sites to try and see if what’s gone before can be a factor. Both sites give links to ‘full poll data’ but they are only to the full data about the marijuana questions. What I did notice, however, was that the CBS poll had the marijuana question as No 15 while Gallup had it as No 23.
I was able to find other Gallup questions asked of the same size sample on the same date with lower numbers that suggest this was a wide-ranging survey covering a lot of ground. The earlier questions I found were about the upcoming elections and the state of the economy. I wasn’t able to find any of the other questions asked by CBS at the same time. That means I can only surmise that the previous questions had put respondents into a particular frame of mind when the cannabis question was asked.
It should be said that all polls, of course, have margins of error and taking the extremes for both the Gallup poll and CBS News, the results could be anything from 46% versus 43% to 54% versus 37%.
What Sabet doesn’t seem to want to draw attention to is that the CBS News poll found 77% in favour of allowing doctors to prescribe small amounts of marijuana to treat serious illnesses. Though Gallup didn’t ask that question, this time, it has previously found 70% to be in favour of permitting the medicinal use of cannabis.