I return once again, with a heavy heart, to the issue of gay marriage. There is more rather one-eyed polling by the campaign against gay marriage, dutifully reported the Telegraph. The way polling commissioned by the lobby opposed to gay marriage is reported by the newspapers and those commissioning it is almost a masterclass in poor interpretation of polls.
Firstly we have the problems of polls conducting using agree/disagree statements, which risk bias in the direction of the statement asked (which I have written about at length here) and often give contradictory answers if properly balanced with statements in the other direction. For example, in this poll 62% of people agreed with the statement “marriage should continue to be definied as a life-long exclusive commitment between a man and a woman”. The campaign against gay marriage have consistently focused on this question and interpreted this as meaning that 62% of people are opposed to gay marriage. However, the same poll found people agreeing by 44% to 38% that “legalising gay marriage is important because maintaining the distinction between civil partnerships and marriage worsens public attitudes towards gay people”. One could just as easily cherry pick that question from the poll and claim that more people support than oppose gay marriage.
As I have written before, we have a multitude of polls that have actually asked directly whether or not people support the legalisation of gay marriage, and they should really be the starting point for anyone looking for polling evidence of whether people support or oppose gay marriage. I summarised the main polling on the issue earlier this month, but suffice to say, polling asking a straightforward question on whether people support or oppose gay marriage consistently finds more people support it than oppose it. Polls that offer more nuanced options, and ask if people support gay marriage, only civil partnerships, or neither gay marriage nor civil parterships, still consistently find support for gay marriage in the mid-40s, but normally find a substantial minority of people who support civil partnerships but not gay marriage, so sometimes show more people opposed to gay marriage than support it.
Secondly we have “would X make you more or less likely to vote for party Y” questions. Again, I have ranted about these at length before, but essentially it is a type of polling question which gives false prominence to and therefore greatly overstates the importance of specific policies when voting intention is actually driven by broad perceptions of parties, their competence, leaders, capability on major issues like the economy and so on.
Now we have that hoary old chestnut of a poll showing people want a referendum on gay marriage. As previously discussed, if asked in an opinion poll people want a referendum on almost everything – unsurprisingly, given that questions on referendums basically boil down to “would you like to have a say on this or should it be left to the hated politicians to decide”. It doesn’t mean there is some huge untapped demand for a referendum on that particular issue, people support a referendum on anything you ask about (the one exception I have managed to track down was a MORI poll back in 2001 that found people did not want a referendum on abolishing the monarchy)
Why do I keep coming back to this? I think its mostly the consistently credulous and one-sided reporting of polls on gay marriage in some sections of the press. Readers of some newspapers could be forgiven for thinking that the polling showed that the public were opposed to gay marriage, when any fair minded look at the broad range of polling on the issue would show that the balance of opinion is broadly positive towards it.