Intended for healthcare professionals


Association between voting patterns and mortality remains

BMJ 1997; 315 doi: (Published 16 August 1997) Cite this as: BMJ 1997;315:430
  1. George Davey Smith, Professor of clinical epidemiologya,
  2. Danny Dorling, Lecturer in geographya
  1. a University of Bristol, Department of Social Medicine, Bristol BS8 2PR

    Editor—We reported on the association between voting patterns and mortality in England and Wales during the 1983, 1987, and 1992 British general elections.1 There was a strong negative association between voting Conservative and mortality and a strong positive association between voting Labour and mortality; there was a weaker negative association between voting Liberal Democrat and mortality and a weaker positive association between abstention and mortality. The recent general election, in which a Labour government was elected with a large majority, has been interpreted as reflecting a breakdown of traditional voting loyalties.2 3 We have analysed the results of the 1997 general election in England and Wales with the exception of two seats: Tatton, in which an independent candidate stood against a Conservative alleged to have taken money in return for asking questions in parliament, and West Bromich West, where the speaker of the House of Commons was unopposed. We used mortality data for 1990-2; these are the most recent data from the 1991 census for which estimates of the population at risk of death are available. In the 1 have added our findings for the 1997 election to the results from previous elections.

    Correlation between party voted for and mortality, 1983-97

    View this table:

    The correlations between voting Conservative, Labour, or Liberal Democrat and mortality have changed little. There is a weak negative correlation between voting for the Referendum party and mortality; the Referendum party fielded candidates in 547 constituencies and won 2.7% of the vote. The main change over time is the continuing strengthening of the positive correlation between abstaining and mortality. The 60% variance in mortality explained by the voting data is virtually identical with the 61% explained by voting in the 1992 election. Correlations in the analysis for the 1997 election may be weakened by the fact that the mortality data are taken from six years before the election. The main change is that for the 1997 election, voting for Labour and abstaining are the most important variables in explaining the variance in mortality in multiple regression analyses, while in the 1992 election Labour and Conservative voting contributed most to the analyses.

    Frank Dobson, the new health secretary, has stated that “Labour will tackle the underlying causes of bad health and if we succeed we will be able to use the slogan 'vote Labour, live longer.' “4 Increasingly it is young adults living in places where few people vote who die youngest. Perhaps the new Labour government should concentrate on socially excluded groups, who may not contribute to electoral success but nevertheless require the most from an administration that offered to lead us away from 18 years of selfishness under the Conservative government.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    View Abstract