Before and after

Does giving voters the facts on issues like the NHS change their views? Up to a point
April 16, 2005

In the run-up to the general election, surveys of British opinion have been showing a "perception gap" over public services—a divergence between the personal experience people report and their views of the trend in the quality of services generally. While they are likely to be satisfied with their local school or hospital, they regard the education system and the NHS as a whole more negatively.

This may be partly because of optimism—people (and companies) regularly expect their own economic prospects to be better than those for the country as a whole. But there is evidence to suggest that people have more negative views about broader social change than is justified by the facts. This may, in turn, be a function of the "bad news" bias of news values, at a time when more people depend exclusively on the news media for political information and are less likely to use other sources—such as trade unions, political parties or churches.

There is an identifiable correlation between on the one hand the amount and accuracy of the information available to people on political issues, and on the other their opinions about those issues. A recent example was the debate on student tuition fees last year, when surveys showed that those who were most familiar with the detail of the proposal were more likely to support it. Of those aware that there were no up-front fees, that interest rates on loans were very low and that students from poor families would not miss out, a large majority supported the scheme, compared with only three in ten overall.

However, successive polls by Mori in the run-up to last year's referendum on an elected regional assembly in the northeast illustrate an opposite relationship between information and opinion. In February 2003, 59 per cent of residents said they supported the idea. By the summer of 2004, this had dropped to 39 per cent. By October those in favour had fallen to 29 per cent. And the actual poll in November showed only 22 per cent support. All along, opposition was strongest among those who claimed to be well informed on the issue—and that group could only have grown as the campaign and the publicity developed. The government is hoping that momentum in the referendum on the European constitution will go the other way, given that more knowledgeable people are at the moment more likely to support it.

Last September, we at Mori designed a three-stage poll of 1,975 people across Britain to throw some light on this "perception gap." In the first stage, we asked people for their opinions on controversial subjects—the NHS, crime, asylum and Britain's EU membership. We then gave them a mini-quiz on these topics, in the form of a series of factual questions. Finally, we gave them the correct answers to the factual questions and asked if this knowledge changed their opinions.

In the first stage, 21 per cent of people said they thought the NHS was getting better, and 36 per cent that it was deteriorating. Views of the police service were very similar: 20 per cent thought it was improving and 37 per cent deteriorating. On asylum, more than four times as many thought government policies were unsuccessful (74 to 18 per cent), while EU membership was favoured by 47 to 38 per cent.

In the second stage, respondents were asked a series of questions on these topics—for example, how many NHS nurses are there compared to a year ago?—and given three options from which to choose their answers. It turned out that those who gave the correct answer on nursing numbers—that there are 14,000 more than a year ago—were most likely to think the NHS was improving, while those believing it had worsened thought by three to one that the figures had dropped or remained the same. The pattern was similar on the question of NHS waiting lists, which have fallen by 100,000 over the past year.

A majority (54 per cent) of those who saw improvements in police services thought that there were more officers (there are 11,000 more than four years ago), while over a third who saw a worsening believed numbers had fallen. On asylum, 67 per cent of those taking a positive view of government policy thought that benefits paid to asylum seekers were less or no more than the level for UK residents (in reality, they are considerably less). However, 76 per cent of those dissatisfied with policy believed that they got at least as much—and over half thought they even received more.

Of course, being aware of the facts does not automatically make one more positive about these issues. It may simply be that those with good personal experience, or those who support the government, are more likely to believe the positive case. But there is some indication that political sympathies are less important than factual knowledge—for example, while 35 per cent of those who knew that police numbers had been rising thought the service is improving, only 26 per cent of Labour supporters took this view.

In the final stage, informants were given the correct answers to the factual questions they had been asked, and were asked if that knowledge changed their views. Respondents were split into two groups: one was told that the figures came from government sources, and the other that they were independent.

In general, around a fifth of those who were wrongly informed were prepared to change their views on being given the correct figures. As might be expected, government sources were somewhat less trusted, with 16 per cent on all topics except Europe (19 per cent) shown government figures considering a revised opinion, while possible switchers based on independent sources varied between 18 per cent (Europe) to 23 per cent (NHS).

Taking at face value these statements of willingness to change views, just how significant are they? Strikingly, if the full suggested switching occurred, the negative views of the NHS and police services would be reversed. On the NHS, the initial negative balance of 36 per cent thinking it was getting worse as against 21 per cent of people thinking it was getting better becomes a positive one of 37 per cent better and 34 per cent worse (see chart above). Similarly, while policing was seen to be deteriorating by 37 per cent compared to 20 per cent thinking it was getting better, this changes to a close but positive balance of 37 per cent to 36 per cent when people are told that police numbers have risen.

The sample's view on asylum policy is not reversed by the information provided. But a ratio of four to one (74 per cent with negative views to 18 per cent with positive) is reduced to two to one (61 per cent against 31 per cent) when people are told how little benefit asylum seekers receive.

The most interesting shift is on attitudes to the EU. Because 61 per cent of supporters underestimated the cost to Britain of membership (about £2.6bn a year), supporters were reduced from 47 per cent to 38 per cent after being given the figure, with opponents increasing from 38 per cent to 47 per cent.

That only 20 per cent of respondents were prepared to shift their opinions when presented with facts that seemed to undermine their views might be regarded as trifling. On the other hand, the figure may have been depressed by the psychology of the interview situation, with people not wanting to lose face or to admit that their views were based, in part, on false information. (Only a third of those who were not prepared to reconsider their views challenged the figures given to them or said they "didn't trust" the government. Another third of those maintaining their original views justified this by referring to conditions being "different" locally.)

On the other hand, one could argue that one in five is quite a big shift in opinion on the basis of one or two statistics, and the finding does suggest that a wider spread of accurate knowledge is likely to have an impact on assessments of recent change.

Presumably nobody would argue against the idea that a better informed electorate is desirable. If the popular belief is that crime is soaring, nurses are flooding out of the NHS and asylum seekers receive benefits at the same level as British citizens, policies on these subjects will either fail to reflect public opinion or be distorted to fit faulty perceptions.

This, of course, leads back to the role of the media in providing information and influencing opinions. Another recent Mori study suggests that their effectiveness varies greatly from issue to issue, but is particularly high where asylum and crime are concerned. Much of the coverage of these topics by the press is demonstrably inaccurate and misleading. The widespread misperception of the facts identified by our survey cannot be unrelated to this relentless barrage of misinformation. And yet only 11 per cent of those not prepared to consider changing their views on asylum acknowledged that this was because they preferred to rely on the media. In practice, of course, many of those who said they did not believe the figures they had been given, or who did not trust the government, would have been influenced by the press. It is quite common for people to be reluctant to admit that the media affect their views, and just as common for them not to know where their opinions come from at all.

David Marquand claimed (Prospect, March 2005) that the next election will be a "populist auction," and that the "people" can be wrong. While the people still seem to have enough collective wisdom for their decisions to sustain a working democracy, improved information can make perceptions more realistic. It remains to ensure that people can distinguish between competing sources and develop trust in those that demonstrate credibility.

Full poll results are available from Mori: