¹ The term ‘unobtrusive measurement’ is derived from Webb, Eugene J. et al. , Unobtrusive Measures (Chicago: Rand McNally, 1966).Google Scholar But unobtrusive measurement of legislative opinion was being conducted long before the term was coined. See for example Rice, Stuart, ‘Behavior of Legislative Groups’, reprinted in Wahlke, John C. and Eulau, Heinz, eds., Legislative behavior (Glencoe, Ill.: Free Press, 1959)Google Scholar; Truner, Julius, Party and Constituency: Pressures on Congress (Baltimore: Johns Hopkins Press, 1951)Google Scholar; Truman, David B., The Congressional Party (New York: Wiley, 1959).Google Scholar

² Patterson, Samuel C., “The British House of Commons as a Focus for Political Research’, British Journal of Political Science, III (1973), 363–81, p. 376.CrossRefGoogle Scholar

³ Patterson, , ‘The British House of Commons’, p. 375.Google Scholar

⁴ Hollis, C., Can Parliament Survive? (London: Hollis and Carter, 1949)Google Scholar; Hollis, C., Has Parliament a Future?, Unservile State Papers No. i (London: Liberal Publications Department, 1961)Google Scholar; Hill, A. and Whichelow, A., What's Wrong with Parliament? (London: Penguin, 1964)Google Scholar; Foot, M., Parliament in Danger, Pall Mall Pamphlet No. 4 (London: Pall Mall Press, 1959)Google Scholar; Johnson, D., On Being an Independent MP (London: Johnson, 1964)Google Scholar; ‘The Decline of Parliament’, The Political Quarterly, xxxiv (1963), pp. 233–9Google Scholar; Crick, B., The Reform of Parliament (London: Weidenfeld and Nicolson, 1968).Google Scholar

⁵ Following the procedure developed by Stuart Rice in his ‘Behavior of Legislative Groups’. The figures on whipped votes were extracted from a paper by John Mackintosh, ‘The Influence of the Backbencher, Now and a Hundred Years Ago’, read to the Manchester Statistical Society in March 1970, an extract from which may be found in Leonard, Dick and Herman, Valentine, eds., The Backbencher and Parliament (London: Macmillan, 1972).Google Scholar

⁶ For unwhipped votes, the Labour party cohesion score fell in 1968 to only 67 per cent. The data on free votes were taken from Richards, Peter, Parliament and Conscience (London: Allen and Unwin, 1970), p. 180.Google Scholar

⁷ Away from the floor of the House, in Standing and Select Committees, party cohesion is lower. A 10 per cent sample of divisions in these committees showed cohesion scores of 95 per cent in Standing Committees and only 72 per cent in Select Committees.

⁸ Butt, R., The Power of Parliament (London: Constable, 1969)Google Scholar; Dowse, R. E. and Smith, T.. ‘Party Discipline in the House of Commons -a Comment’, Parliamentary Affairs, xvi (1963), 159–64Google Scholar; Lynskey, James J., The Role of British Backbenchers in the Modification of Government Policy (unpublished Ph.D. thesis, University of Minnesota, 1966).Google Scholar

⁹ James Christoph suggests however that division lists might be used in a systematic manner in the study of the House of Commons. Christoph, James B., ‘The Study of Voting Behavior in the British House of Commons’, Western Political Quarterly, xi (1958), 301–18.Google Scholar

¹⁰ Kornberg, Allan and Frasure, Robert C., ‘Policy Differences in British Parliamentary Parties’, American Political Science Review, LXV (1971), 694–703.CrossRefGoogle Scholar

¹¹ Finer, S. E., Berrington, H. B. and Bartholomew, D. J., Backbench Opinion in the House of Commons, 1955–59 (Oxford: Pergamon Press, 1961)Google Scholar; Jackson, Robert J., Rebels and Whips (London: Macmillan, 1968)Google Scholar: Berrington, H. B., Backbench Opinion in the House of Commons. 1945–55 (Oxford: Pergamon Press, 1973).Google Scholar

¹² Berrington, Finer and Bartholomew, , Backbench Opinion 1955–1959. p. 7.Google Scholar

¹³ This is not to say that the material lends itself to easy analysis. Indeed. Berrington has estimated that the analysis of the components of the public record which formed the basis for his study with Finer and Bartholomew took two man-years to complete. However, the availability of computer techniques for summarizing the coding materials in textual form, and for subsequent analysis of the codes allocated, has substantially changed the focus of the problem; see Franklin, Mark N., Using Computers to Analyse the Activities of Members of Parliament (University of Strathclyde, Survey Research Centre Occassional Paper No. 9. 1971).Google Scholar

¹⁴ One of the present authors has traced their development since the turn of the century. See Franklin, Mark N., Voice of the Backbench: Patterns of Behaviour in the British House of Commons (unpublished Ph.D. thesis, Cornell University, 1970), especially Chap. 2.Google Scholar

¹⁵ Franklin, , Voice of the Backbench, pp. 42–51.Google Scholar The manner in which the motions are tabled and signatures collected is detailed in pp. 51–3.

¹⁶ Jackson, , Rebels and Whips, p. 309.Google Scholar

¹⁷ Reviews of the Finer, Berrington and Bartholomew study, which was the first to make systematic use of EDMs as a data source, were almost universally hostile on this point, and the fact that EDMs have continued to be used in the manner pioneered by these authors does not in itself vindicate the procedure.

¹⁸ Interviews of a selected group of MPs were carried out in 1968 and 1969. Franklin, , Voice of the Backbench, pp. 17–18Google Scholar, gives details of the way in which these were selected. More recent communications with MPs have tended to confirm the impressions gained at the time.

¹⁹ Interview with a Labour MP, November 1968.

²⁰ Interview with a Conservative MP, November 1968.

²¹ Berrington, Finer and Bartholomew, , Backbench Opinion 1955–59, pp. 9–10.Google Scholar

²² Berrington has defended the use of EDMs in terms of the categories of MP defined in the Finer study and by comparison with voting in divisions (Backbench Opinion 1945–55, pp. 6–19).Google Scholar

²³ Berrington, Finer and Bartholomew, , Backbench Opinion 1955–59, pp. 10–11.Google Scholar

²⁴ Berrington, , Backbench Opinion 1945–1955, pp. 12–13.Google Scholar

²⁵ Interview with a Conservative MP, November 1968.

²⁶ Interview with a Labour MP, October 1968.

²⁷ In the weeks after the 1964 general election, Stokes and a special group of interviewers contacted the old and the new Members and unsuccessful parliamentary candidates from each of the eighty constituencies which had served as the primary sampling units of a survey of the electorate of Great Britain.

²⁸ For example, the coding scheme adopted in the Stokes Candidate Study for opinions regarding the Common Market was as follows: (i) should go in, for the Common Market; (2) should try to keep trading policies in line with the Six with a view to going in eventually; (3) for, right wing; (4) should go in on our own terms; (5) neutral, centre; (6) should try to get other arrangements going with the Seven and the Commonwealth in case we can't go in, shouldn't just hang around hoping; (7) should not go in while the Six looks like a rich man's trading club with reactionary right wing governments, should wait until it shows itself; (8) against; left wing; should never go in, should make permanent arrangements with the Seven or the Commonwealth and make it known that these are not just temporary makeshift anti-Common Market arrangements.

²⁹ Following the procedure outlined by Janda, Kenneth, A Conceptual Framework for the Comparative Analysis of Political Parties (Beverley Hills, Calif.: Sage Publications, 1970).Google Scholar A methodological appendix can be obtained from the authors which details the coding strategies adopted in this paper, and motions used in the analysis.

³⁰ Evidently, agreement in the assignment of policy position codes is no guarantee of correct assignment and this procedure attempts to ensure that those motions that risked being wrongly assigned were excluded. Of course, an investigator may err in the confidence with which he codes a motion as well, but with two separate estimates of confidence as well as the two separate estimates of policy position we felt that the risks of faulty coding had been reduced to a minimum. To some extent we have been able to assess independently the improvement in our coding that resulted in the use of the adequacy confidence codes, as reported in the methodological appendix referred to in fn. 29.

³¹ In this context; a ‘popular’ motion is one which proposes a policy ascribed to by a majority of signatures in the group under investigation in the issue area concerned. The extent to which such signatures should be discounted is discussed in the methodological appendix available from the authors.

³² We defined an ‘inconsistent signer’ as one who had, on at least two occasions, signed motions on both sides of an issue. Eight predictions were made as to the opinions of these individuals, of which three were in error.

³³ For a variety of reasons we preferred to measure the span in terms of the number of times a motion appeared on the Order Paper with new signatures added, rather than in terms of calendar dates.