This article was first published on The Croydon Citizen on 08/03/2016
We, ‘the people’
Listen in to any Croydon Council meeting and you will hear ‘the people of Croydon’ being invoked. The Class War candidate for Croydon South was also fond of referring to ‘the people’ during the general election campaign, though in the end only 65 of them voted for him. How do they think that they know what ‘people’ think?
When pushed, politicians claim to have heard it ‘on the doorstep’. My experience is canvassing conversations have a strong tendency to be with supporters of your own party. This is selection bias: getting information from a group unrepresentative of the wider population.
Another source for what ‘people’ might think is surveys. If there’s a phrase that should set your ‘bad data’ alarm ringing, it’s ‘a survey has shown…’. The sound should become deafening if the survey was conducted online.
Young people have been asked whether they agree that more should be done to support young people’s mental health and well-being
That is not to say that surveys are not a legitimate method for acquiring information, but their results need to be treated with care. Questions you should ask are ‘who got to answer the question?’ (more selection bias) and ‘how was the question phrased?’
An amusing example of the latter was posed by Jonny Rose, he of Croydon Tech City, using the Twitter survey feature. Jonny asked his followers (a selective sample) what the best place in the UK was. There were two possible answers, Croydon and Croydon. At my last count, Croydon (option 1) led Croydon (option 2) by a ratio of six to four.
Jonny was clearly not intending to gather meaningful statistical data for analysis. Yet survey questions that are little better than Jonny’s aren’t hard to find. In a survey by the Croydon Opportunity and Fairness Commission, young people were asked whether they agree that more should be done to support young people’s mental health and well-being in Croydon.
Such surveys lead to a conclusion that more needs to be spent on everything, which is little help
The way in which the question is posed guarantees the result. This is ‘survey bias’. What young person is going to say that less should be done to support young people with mental health issues?
The purpose of the question cannot be to elicit meaningful information on young people’s priorities. It is presumably designed to give a veneer of objectivity to a question about which the commission has already made up its mind. A strap line of ‘X% of young people think more needs to be done on mental health’ awaits only filling in of ‘X’.
Information on priorities will not result from questions of this kind. Nor will posing a list of similar questions on other areas for expenditure. Inevitably, such surveys lead to a conclusion that more needs to be spent on everything, which is little help in deciding priorities.
When the going gets tough, the temptation of local politicians is to blame central government
To elicit priorities, respondents need to be asked to make a judgement that does not allow for the easy ‘we should do more’ option. A possibility is to ask for related alternatives to be put into priority order; say more spending on school sports, after school clubs, smaller class sizes or mental health. This is, to some extent, what devolving budgets to local decision-making is.
A weakness of such devolution comes when there is a need to reduce expenditure. A judgement on where to spend extra money is a lot easier than a judgement on where to cut expenditure.
When the going gets tough, the temptation of local politicians is to blame central government, especially when the ruling party is of a different stripe. The Scottish National Party has traded on this for some time. Devolution gave the Scottish parliament the ability to vary income tax. It is a power that is yet to be used. Far easier to call for more expenditure on all matters, and blame the inability to do so on underfunding from central government.
Petitions are a favourite means for opposition parties and pressure group
The Labour-led Croydon Council adopts the same line. All good things are claimed to come from them; all bad things are the result of (Tory) central government cuts and/or decisions by the previous (Tory-led) council administration.
We can also gauge the mood of the population through gathering petitions, a favourite means for opposition parties and pressure groups. Local Tories attempted to influence the council on garden waste collections by this means.
Here we will also see selection bias (signatories are more likely to be Tories), but in this case it is not about an approval rating: success is gauged by the number of signatories. Try closing your local A&E you will get a lot, for green garden waste probably less, but what’s a good number? I have no idea.
Elections are the best we have
Sadly, it would seem there are no quick, reliable and simple solutions to ascertaining the view of ‘the people’. Elections, which are just a survey where a myriad of decisions is distilled into a single choice, are the best that we have. Methods attempting to ascertain opinion on an ongoing basis are less good, but some are better than others and within each there is good and bad practice.
If there is one conclusion to be drawn, it is that when next you hear someone claim to know the view of the people, assume that they are just making it up.
By Robert Ward