¹ This burst of racist voting lasted from about May 1976 (especially the District Council elections at that time) through into early 1977 or so. By May 1977 support for the NF was clearly in decline, despite the publicity given to the party's vote in the Greater London Council elections of that month. See, for example, Husbands, Christopher T., ‘The National Front Becalmed?’, Wiener Library Bulletin, XXX, New Series Nos. 43/44 (1977), 74–9Google Scholar, and Steed, Michael, ‘The National Front Vote’, Parliamentary Affairs, XXXI (1978), 282–93.CrossRefGoogle Scholar

² In fact, the Gallup Poll reported that the period of Labour's particular lack of popularity was from October 1976 to about August 1977, thus beginning almost six months after the evident increase in NF voting; see Webb, Norman and Wybrow, Robert, eds, The Gallup Report (London: Sphere Books, 1981), pp. 169–70.Google Scholar

Furthermore, some of the reason for the apparent magnitude of Labour's losses in the District Council elections of May 1976 was that the seats it was defending were won when the newly re-organized District Councils were first elected in May and June 1973 during a period of mild Conservative unpopularity.

³ The particular danger, of course, is the elision of recall of past voting into current preference; see, for example, Himmelweit, Hilde T., Biberian, Marianne Jaeger and Stockdale, Janet, ‘Memory for Past Vote: Implications of a Study of Bias in Recall’, British Journal of Political Science, VIII (1978), 365–75.CrossRefGoogle Scholar

⁴ For each constituency is given the percentage of the total vote cast for Labour in the October 1974 general election, the equivalent percentage for the NF, and the percentage swing from Labour to Conservative – as calculated on the total vote – between October 1974 and May 1979. The integer also given is the maximum possible number of opportunities that respondents resident in each place since October 1974 had to vote for the NF and/or its short-lived breakaway group the National Party (only in Deptford and Blackburn in 1976 and 1977) in local elections from October 1974 to the time of the interview. These integers vary from one constituency to another due to different arrangements for the holding of Borough/District Council elections, to the occurrence of a local by-election in Deptford, and to the absence of NF candidacies from some contests.

The study also collected data in a part-ward in the constituency of Bradford North and in an election district in the constituency of Melton, which abuts on the city of Leicester. However, the data from these two locations have not been included in the analyses because there were no NF candidacies there in the October 1974 general election. All seven constituencies whose data have been used also had candidacies from the three principal parties in October 1974.

⁵ See Husbands, Christopher T., Racial Exclusionism and the City: The Urban Support of the National Front (London: George Allen & Unwin, 1983), pp. 149–55Google Scholar, for a full description of the sampling and the weighting procedures used.

⁶ Christopher T. Husbands, addendum to end-of-grant Report of Social Science Research Council Grant HR 5472/2, ‘Social and Attitudinal Characteristics of Supporters of Right-Wing Movements in Britain’, Tables VII.3 to VII.5, pp. 528–30.Google Scholar

⁷ There were no respondents in the sample who reported an NF vote in the general election of February 1974 but not in that of the following October. However, those occasional respondents who had voted NF in October 1974 but who gave no indication at all in any of their responses to the other NF-related questions and items used to define strong NF sympathy that they held any such sympathy at the time of interview were not defined as strong NF sympathizers.

⁸ Office of Population Censuses and Surveys, Classification of Occupations, 1970 (London: HMSO, 1970), pp. x–xi.Google Scholar

⁹ Reid, Ivan, Social Class Differences in Britain: A Sourcebook (London: Open Books, 1977), p. 38.Google Scholar

¹⁰ Husbands, , Racial Exclusionism and the City, pp. 136–9.Google Scholar

¹¹ These three scales were constructed using three separate factor analyses that extracted in each case only that factor and then calculated the appropriate factor score for each respondent in the whole sample. The method of factor analysis employed was principal factoring using iterations to estimate communalities; see Harman, Harry H., Modern Factor Analysis, 2nd edn, revised (Chicago: University of Chicago Press, 1967), pp. 135–46.Google Scholar The items in each scale are: Scale of Law-and-Order Sympathy (1/4); (1) There's not enough discipline in the schools these days; (2) Britain should bring back the birch for people convicted of violent crimes; (3) Britain should bring back National Service for the young; (4) Tough sentences would deter people from breaking the law. Scale of Trade-Union and Worker Sympathy (1/2); (1) Trade unions have too much power in Britain; (2) Unofficial strikes should be banned. Scale of Belief about Detrimental Effects of Black People on Jobs and Promotion (1/2); (1) It's easier for a white person to find a job than it is for an immigrant; (2) It's much harder for white people to get promotion if they work in a job where there are many immigrant workers.

These items were asked in various Likert-scale item-grids, whose scale-values were as follows: ‘1’, Agree strongly; ‘2’, Tend to agree; ‘3’, Neither agree nor disagree; ‘4’, Tend to disagree; and ‘5’, Disagree strongly. In order that a positive score should indicate a commitment in the direction implied by the title of each scale, the initially calculated direction of the factor scores was reversed before analysis of the scales Law-and-Orderand Jobs and Promotion. The fractions next to the name of each scale give the number of items out of the total of those included in the construction of that scale for which it was permitted that ‘missing data’ be replaced by the means of the items concerned. If a case had ‘missing data’ on more items than the numerator of these fractions, then its factor score on the scale concerned was treated as incalculable. The correlation coefficients used in the computation of the factor analyses were calculated on the principle of pair-wise deletion.

¹² The questions about the specific attributes of the two major parties were asked by requesting respondents to say which of a set of descriptive words and phrases listed on a show-card they felt applied in each case. The same list of potential descriptors was used for the two major parties.

¹³ This is not to deny the ideological effect of the NF in defining the nature of the debate between the major parties on immigration; nor is it to deny that it had political effects, as reflected in certain policies of the Labour government.

Crewe and Särlvik do argue that it would not be in the interests of the Conervative party to adopt a full-blown Powellite strategy on race relations and immigration; however, they concede the effect of Mrs Thatcher's ‘swamping’ outburst in January 1978 in distinguishing the positions of the major parties on the issue of immigration. See Crewe, Ivor and Särlvik, Bo, ‘Popular Attitudes and Electoral Strategy’, in Layton-Henry, Zig, ed., Conservative Party Politics (London: Macmillan, 1980), pp. 244–75, at p. 259.Google Scholar

¹⁴ However, one should mention that Joan Lestor almost certainly held Eton and Slough only because of the intervention of an Independent Conservative whose total of 2,359 votes exceeded her majority by 1,019 and whose name was at the top of the ballot paper, while that of the official Conservative was at the bottom; even so, NF support in October 1974 was substantially lower in this constituency than in any of the six others. If the official Conservative and Independent Conservative totals are combined, the percentage swing against Labour in Eton and Slough is 9.3.