OPINION5 September 2011

The riots: Lessons for the pollsters

Opinion

Opinion polls are an invaluable tool, says Westminster City Council’s Neil Wholey, but the surveys that followed the riots in England last month showed that pollsters could do better.

Local government research thinktank LGinsight reviewed those polls to produce a discussion paper, providing local communities, and local government in particular, with some background and a framework for any research or consultations they are carrying out.

As social researchers we are very grateful for any research put out in the public domain, but by reviewing the riot polls we have identified gaps and areas that could be usefully developed. It is clear that polling companies could do more, not only to help researchers understand what is going on in society but also to improve the service they provide to their clients.

“It’s great to provide a damning figure for the headlines, but it would have been more useful to uncover the link between the riots and the underlying concerns of the public”

We suggest that if they follow these five points, they’ll do much better next time a crisis hits.

1. Add context

It’s great to provide a damning figure for the headlines saying that x per cent believe those in charge are handling the situation badly, or that y per cent think we should lock the rioters up and throw away the key. But confirming that people are horrified by riots doesn’t really tell us anything new. The useful bit would have been to uncover the link between the riots and the underlying concerns of the public. What most pollsters missed was that, despite the riots, the economy was still the most important issue for many people. And underlying concerns like these meant that, for many Britons, the outbreak of riots wasn’t such a surprise.

The polls also managed to give the impression that the only people involved in dealing with the riots were Westminster politicians and the police. What about the fire service, local councils and health services? Being aware of the context in which public services operate may sound dull, but it would earn pollsters extra points from the sector in considering their research.

2. Be prepared

There is lots of trend data available for polling companies to track confidence in politicians and the economy, but where is this sort of data for society more generally? The government axed the Citizenship Survey which contained many questions looking at how cohesive society is. It would be great if polling companies started to use some of these questions, or others, to track how we feel about society. As it is, we do not know whether the riots have led to a decline in community cohesion or trust.

3. Challenge the public for solutions

Questions asked in polls on the causes of the riots did not follow up by asking about possible solutions. If it is a lack of respect in society or general criminality that caused this, what do the public suggest should be done? Harsh punishments are easy to get people to agree with in a survey, but what about discussing the rehabilitation of offenders, improving parenting skills or community service? There was more support, for instance, for “national citizen service” than for bringing back national military service, but what does this mean? Would people really want to be involved in this themselves or see their children signed up?

Really clever questions would have tied in possible cuts in the public sector, beyond just police numbers. What about changes to libraries, children centres and youth clubs? Also, despite wonderful community spirit being shown to clean up and deal with the riots, there were no questions asked about levels of support for this sort of action.

4. Ask about information provision

It is really useful to know in a crisis where people turn for information. Councils, the police and other public sector bodies worked hard behind the scenes to keep people informed and calm the situation. But were their emails, tweets, events and other communications any use? If riots occurred again, where should those in charge send messages so that they reach the public as quickly as possible and are trusted by the people who receive them?

5. Assess yourselves

There is a wealth of information on a range of issues available on the websites of the major polling companies, many with full computer tables showing demographic breakdowns. This is to be encouraged. But polling companies should also be looking to review the polls they conduct and assess if they could do better next time. Anyone that does could establish themselves as thought leaders on a topic, tidy up their long-term trends and put themselves in a good position to be selected for any polls that come up. Anyone can bash out a question on any topic, but few can accompany the first draft with a report or paper explaining the choice of question made.

The polling companies that had the foresight to publish their work on the riots in full have created a valuable resource for social researchers and commentators and should be commended for their openness.

But they need to take this to the next level by reviewing what they have and seeing what extra value could be added, to show us that they can go beyond providing great headlines and really help drive our understanding of an important issue.

Neil Wholey is chair of LGinsight and head of research and customer insight at Westminster City Council

3 Comments

13 years ago

The reality is that most of the key questions can't really be asked in an opinion poll but, in gaining column inches, and keeping pollsters names in the press during tough economic times, the polls served their purpose.

Like Report

13 years ago

I don't quite understand the point of this article or why Neil seems to be 'preaching' what he sees as opportunities missed and what he might have done if in the pollsters position. As Neil should know from his previous employment, many of these polls are commmissioned work and therefore pollsters do not always have the freedom to ask what he considers as the 'right' questions.

Like Report

13 years ago

Thanks Neil, a useful article. I work in a local council where there were riots and we are thinking about talking to local residents about why they think the riots happened, was it handled well here and what the impact might have been. Your article was useful as part of scoping this work.

Like Report