Experience versus Perceptions: Accounting for the NHS “Perception Gap”

Abstract

Between 2003 and 2006, annual Populus opinion polls have revealed a significant gap between personal experience and perceptions of the performance of national public services. The gap is most pronounced in the case of the NHS. In this paper, I examine whether social characteristics, party identification, newspaper readership, and variation in the levels of political awareness, issue salience, and partisan strength account for differences in evaluations of the NHS. Party identification appears to influence perceptions more than evaluations of personal experience; the evidence on newspaper readership is equivocal. Attaching importance to the NHS as an issue and strong identification with the Labour Party, along with social characteristics, party identification and newspaper readership also significantly affected the size of the gap between experience and perceptions. These findings suggest that perceptions, rather than views based on personal experience of the NHS, are more susceptible to partisan and, possibly, media influences. They also suggest that fluctuations in the NHS perception gap’s size, apparent between 2003 and 2006, are the product of changes in the intensity and direction of party support.

Acknowledgements

I would like to thank John Bartle, Chris Wlezien, and Elinor Scarbrough for their help and advice on earlier drafts of this article. I would also like to thank the three anonymous referees for their comments.

Notes

1. Populus poll for the News of the World, 24 June 2004, ⟨http://www.populuslimited.com/poll_summaries/2004_06_24_notw.htm⟩ (accessed 20 April 2006).

2. Department of Health ⟨http://www.dh.gov.uk/en/Publicationsandstatistics/Statistics/Performancedataandstatistics/HospitalWaitingTimesandListStatistics/index.htm⟩ (accessed 2 December 2008).

3. Populus poll for the Times, 5 May 2005, ⟨http://www.populuslimited.com/poll_summaries/2005_05_04_times_finalpoll.htm⟩ (accessed 20 April 2006).

4. Political evaluations can be divided into two types: simple and mediated (Fiorina, 1981, p. 106). Simple evaluations refer to conditions occurring over a specific period of time where there is no partisan reference. Mediated evaluations reflect appraisal(s) of the performance of political parties on an issue or in dealing with an event or problem.

5. The salience of an issue, which depends on interests and concerns relating to an individual’s underlying values, needs and goals (Krosnick, 1988, pp. 196–197), is widely claimed to be crucial in the formation of opinions (Himmelweit et al., 1985, p. 71; Lavine et al., 1996, p. 298; Hutchings, 2001, p. 846–848).

6. The “attitude accessibility” model claims that the more easily an attitude can be accessed from memory, the greater its influence on subsequent perceptions and behaviour. Accessibility is determined by the frequency of activation (important attitudes are frequently subjects of conscious thought), distinctiveness or extremity (attitude strength), and the extent of links between an attitude and other psychological elements (Krosnick, 1988, p. 197), for example identification with a political party.

7. This question is not strictly comparable with that used by Populus to elicit perceptions of the NHS, as it makes no reference to the government.

8. Readership of the following national newspapers was included in the analysis: Daily Mail, Daily Star, Daily Telegraph, Express, Financial Times, Guardian, Independent, Mirror, The Sun, and The Times. The Aberdeen Press and Journal, The Glasgow Herald, and The Scotsman were added to the “none most often” and “other” categories to form the base category of “other newspapers”. It would have been preferable to include measures of exposure to television and radio political coverage, as well as newspaper readership, in the analysis but these were not available in either the first or second waves of the BES. This is particularly regrettable given that most respondents indicate that they rely on, and trust, television rather more than the print media, which is often partisan.

9. Most respondents (46%) reported that they were very satisfied with recent NHS medical treatment. Just over 39% said they were fairly satisfied, 9% were a little dissatisfied, and only 6% were very dissatisfied.

10. Measures of general political knowledge are suitable alternatives – although less than perfect – for measures of specific issue knowledge, as they are closely correlated (Zaller, 1992, pp. 336–337) and generally more reliable (Bartle, 2000, pp. 471–473).