Hostname: page-component-586b7cd67f-tf8b9 Total loading time: 0 Render date: 2024-11-23T20:52:54.857Z Has data issue: false hasContentIssue false
  • English
  • Français

Forecasting the British Election to the European Parliament

Published online by Cambridge University Press:  27 January 2009

Clive Payne  and
Philip J. Brown
Get access Rights & Permissions [Opens in a new window]

Extract

The first elections to the European Parliament were held on 10 June 1979 with each EEC country following its own electoral system. As in all recent elections, on-the-night forecasting was an integral part of the television coverage in Britain. This paper describes the method used by the BBC to predict the composition of the British delegation to the European Parliament. The performance of the method is discussed and some tentative conclusions are drawn about the behaviour of the British electorate.

Type
Notes and Comments
Information
British Journal of Political Science , Volume 11 , Issue 2 , April 1981 , pp. 235 - 245
Copyright
Copyright © Cambridge University Press 1981

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 We are indebted to the BBC for sponsoring this project.

2 Brown, Philip J. and Payne, Clive, ‘Election Night Forecasting (with Discussion)’, Journal of the Royal Statistical Society, Series A, 138 (1975), 463–98.CrossRefGoogle Scholar

3 A full discussion of the problems of election night forecasting is given in Brown, and Payne, , ‘Election Night Forecasting’, pp. 465–7.Google Scholar

4 The effect on coefficients estimated with high uncertainty is to shrink them towards zero whereas coefficients with low uncertainty are hardly affected. Thus the method is significant early on in the prediction process.

5 In the context of the analysis of election results see Miller, William, ‘Measures of Electoral Change Using Aggregate Data’, Journal of the Royal Statistical Society, Series A, 135 (1972), 122–42CrossRefGoogle Scholar and Miller, William, Electoral Dynamics (London: Macmillan, 1977).Google Scholar

6 See Goldstein, M. and Brown, P. J., ‘Prediction with Shrinkage Estimators’, Mathematische Operationsforschung und Statistik, Series Statistics, IX (1978), 3–7.Google Scholar

7 See for example, Miller, , ‘Measures of Electoral Change Using Aggregate Data’Google Scholar and McCarthy, C. and Ryan, T., ‘Estimates of Voter Transition Probabilities from the British General Elections of 1974’, Journal of the Royal Statistical Society, Series A, 140 (1977), 78–85.CrossRefGoogle Scholar

8 Let Z1, Z2, Z3 be the predicted shares of the electorate for the top three parties in an undeclared seat. Then the vector Z = (Z1, Z2, Z3) was assumed to have a trivariate normal distribution with a variance-covariance structure V obtained from the regressions on the set of declared results. The probability that a particular party, say party 1, will win the constituency is the probability that the random variable Z1 is greater than both the other random variables, the predicted shares Z2 and Z3. This means that Z1 – Z2 and Z1 – Z3 have both to be greater than zero and these two derived random variables have a bivariate normal distribution obtained from the trivariate distribution. The resulting calculation involves a bivariate normal integration and is done in turn for each of the three top parties.