The Smoking Ban Problem - polls, libraries and predicting future behaviour

***

Many experts on polling question the effectiveness of the “what if” question:

“If politician X was leader of Party Y why would you be more or less likely to vote for them”

There are too many chances to be wrong for this to make much sense as a question – Politician X isn’t leader, you many already have a 100% likelihood of voting for Party Y, the election is in the future. In essence this type of question expects people to make a prediction of their own behaviour following a given change – a we are not especially good predictors of our own future behaviour under such circumstances. Indeed, these questions are really just a variation on “who would you prefer to be leader” or “which policy would you prefer”.

Which brings us to the way in which these questions are used to justify policy decisions. By way of benign example, let’s talk about libraries. Most of the population are not members of a public library and have at best a very occasional engagement with the library service. This isn’t to say we don’t want libraries – we’ll fight hard to protect these vital local services even though they are not services we (or anyone we know) makes use of.

Faced with this problem – declining library membership – the local council undertakes a review part of which involves surveying the public. But first the council thinks of lots of exciting things to do with or put it its libraries – computers, coffee machines, self-service issuing of books, book clubs, kids parties - you name it. And in the survey the council asks whether these things will make non-users more or less likely to use the library.

Now let’s assume that library non-users (or a majority of them) say that some or all of the changes will make them more likely to use the library. Seizing on this, the council rushes through the changes – despite that fact that current library users have said they don’t want these changes. Ah, says the Council, the policy will attract loads of new users to the library and we’ll be in clover!

So what happens? Well the current users of the library don’t like the changes (noisy kids, impersonal service, coffee stains all over the table and so on) and some decide to get their reading material from Oxfam or Waterstones. And those non-users who said they would be “more likely” to use the library? Not a sign of them. Result of the policy changes? Probably further library closures, reductions in the book fund and another review.

This is precisely what happened with banning smoking in pubs – non-users said they didn’t like pubs because of the smoke. And that they would be “more likely” to use the pub is smoking was banned. Existing regular pub users – overwhelmingly – opposed the ban preferring good extraction, separate smoking rooms and other measures. But no – banning smoking would make pubs more popular. Those non-users said so, didn’t they?

In truth those non-users said nothing of the sort – they didn’t go to pubs because they didn’t like pubs. And the smoking was just one factor – mostly they didn’t see the pleasure in sitting a drinking away from their comfortable homes. Nothing to do with smoking, nothing at all. And, following the ban, these non-pubgoers have not started going to the pub while at the same time loads of previous pubgoers now stop at home where they can smoke. With fewer regulars, the remaining hardy folk began to drift away and the local was left with three customers at 10pm on a Friday night.

The result? The pub closes. The community loses a local facility. Ordinary, harmless folk have nowhere to go for a pint and a fag – other than at home of course. Some of that much vaunted social capital is lost. The football team folds – it was a PUB team after all. The gardening club slowly declines. Other groups become ever more cliquey. Why? Because a ban brought in to satisfy folk who never used pubs resulted in those pubs closing.

....