THE DECLINE OF THE BRmSH LEBERAL PARTY: A COMPARATIVE AND THEORETICAL ANALYSIS Thesis for the Degxee of Ph. D. MICHlGAN STATE UNIVERSITY Thomas W. Cassfevens 1966 Vimrsvs M ichigm State University This is to certify that the thesis entitled THE DECLINE OF THE BRITISH LIBERAL PARTY: A COMPARATIVE THEORETICAL ANALYSIS presented by Thomas W. Casstevens has been accepted towards fulfillment t of the requirements for PhD Political Science degree in .— luv ’ [1/ CA ,4 gl/I'é" “a. 7’ A Major professorJ ‘ t I. Date February 8 , 1966 0-169 ABSTRAC T THE DECLINE OF THE BRITISH LIBERAL PARTY: A COMPARATIVE AND THEORETICAL ANALYSIS by Thomas W. Casstevens This dissertation seeks a satisfactory answer to the question: Why did the British Liberal Party decline in the twentieth century? The problem is analyzed from an historical and from a com- parative point of view. The historical perspective focuses upon the British elections from 1900 through 1931, upon the Liberal Party's schism (and the concomitant upsurge of the Labour Party) from 1916 to 1923, and upon the Liberal Party's leadership selection (and deposition) process from William Gladstone to Lloyd George. Exten- sive use is made of British statistics on the condition of the economy from 1885 through 1914, on voting in parliamentary elections from 1900 through 1931, and on voting in the House of Commons from 1919 through 1922. The comparative perspective focuses upon the schisms in the Canadian Liberal Party, the Australian Labor Party, and the United States Republican Party, as well as upon the subsequent rise of the Canadian Pregressive, Australian Country, and United States (Bull Moose) Progressive parties. The magnitudes of these schisms, 2 Thomas W. Casstevens and the degree of success of the rising parties, are estimated and compared with the schism in the British Liberal Party and the rise of the British Labour Party. Without a theoretical decision about what constituted the Liberal Party, the question ' Why did the British Liberal Party de- cline? ' cannot be answered unequivocally. The decision to exclude or include Lloyd George's followers, from 1918 to 1923, is especially critical. This study justifies the (theoretical) exclusion of Lloyd George's group from the Liberal Party, and as a result, the decline of the Liberal Party was rather obviously a consequence of the 1918 schism in the Party. From a comparative point of view, the Liberal Party's schism was unusually severe, and the Labour Party was unusually successful. From the perspective of the Liberal Party's traditional leadership selection process, Herbert Asquith erred by not resigning as Leader of the Party in 1916. Asquith's error probably accentuated the magnitude of the Liberal Party's 1918 schism, and the magnitude of that schism caused the decline of the Liberal Party. THE DECLINE OF THE BRITISH LIBERAL PARTY: A COMPARATIVE AND THEORETICAL ANALYSIS BY Thomas W. Cas stevens A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOC TOR OF PHILOSOPHY Department of Political Science 1966 ACKNOWLEDGEMENTS The author wishes to thank Professor Richard Jones for intro- ducing him to British political history, Professor Richard Rudner for introducing him to the philosoPhy of science, and Professor Herbert Garfinkel for pervasive encouragement and advice. Professors Garfinkel, Joseph Schlesinger and Iwao Ishino have given the author the benefit of their critical readings of the manuscript. Professor Schlesinger especially is responsible for the author's decision to include the study of the leadership selection process in the British Liberal Party. Jeanne S. Casstevens and Harold T. Casstevens, II have facilitated, in diverse ways, the completion of this study. The author himself, of course, accepts sole responsibility for any short- comings in the present work. ii TABLE OF CON TENTS Chapter I. INTRODUCTION . . Stability and Change . Historical Background . . . . . The Structure of This Study . . II. RATIONAL VOTERS AND THE PARTY SYSTEM, 1900-1914 . . . . . . ..... Introduction. ..... A Critique of the Center Party Thesis. An Empirical Critique of the Electoral System Thesis Conclusions. III. THE DECLINE OF THE BRITISH LIBERAL PARTY . . Introduction. ........ The State of the Parties in 1914. The Liberal Party . The Irish Nationalist Party . The Labour Party. . . . The Unionist Party ........ The Origins of the Liberal Schism, 1916-1918 The Reconstruction of the Labour Party and the Decline of the Liberal Party. . The Liberal Party as a Third Party, 1918-1931. Comparative Reflections on Major Party Schisms. . The Rise of the Canadian Progressive Party. . . The Rise of the Australian Country Party . The Rise of the American (Bull Moose) Pro- gressive Party. . Conclusions. iii Page [—1 ll 11 12 40 58 60 6O 61 61 63 63 65 71 83 96 127 128 130 131 135 Chapter Page IV. A THEORY OF HOMEOSTATIC TWO- PARTY SYSTEMS.......................136 Introduction. . . . . . . . . . . . . . . . . . . . 136 Conceptual Foundations. . . . . . . . . . . . . . . . 137 Two- Party System. . . . . . . . . . . . 137 Major Parties and Political Parties. . . ..... 150 The Theory. . . . . . . 158 The Rise of a New or Third Major Party . . . . 159 The Decline of a Major Party ...... . . . . 165 Conclusions and a Formal Model of the Theory . . . 173 V. THE LEADERSHIP OF THE BRITISH LIBERAL PARTY, 1868-1917. . . . . . . . . . . . . . . . . . 175 Introduction. . . ..... . . . 175 The Selection of Liberal Leaders, 1868- 1908. . . . 175 The 1916-1918 Crisis as a Crisis in the Liberal Leadership Selection Process. ..... . . . . . 206 Conclusions. . . . . . . . . . . . . . . . . . . . . . 221 VI. A SUMMARY, CONCLUSION AND PREDICTION . . . . 223 BIBLIOGRAPHY......................... 241 iv Table 2 . 1 LIST OF TAB LES Percentage of Unemployed in Certain British Trade Unions, 1888-1914 . Money Wages, Cost of Living and Real Wages in Britain, 1885-1914. . . . . . . . ..... . . Strikes and Lock-outs in Britain, by Number of Labour Disputes, WorkpeOple Directly and Indirectly Involved, and Aggregate Duration in Working Days, 1893-1914 .......... Number and Membership of Trade Unions in Britain, and Number and Membership of Trade Unions Affiliated with the Labour Party, 1900-1914 . . Summary of Voting in Three-Cornered Contests in British Elections, by Party, 1900-1914 Voting in Three-Cornered Races, by Period and by Party Percentages . .......... Voting in Three-Cornered Races, by Period and byPartyAverage . . . . . . . . . . . ..... Percentage Results of British General Elections, 1900-1910, for Two Coalition Parties . . ..... Percentage Results of British General Elections, 1900-1910, for Five Parties . . .......... Number of Candidates in the 1910 General Elections, by Party .......... . . . ..... Aggregate Results of the 1910 General Elections, byParty............. ..... Page 25 26 28 29 33 36 37 46 47 67 68 Table 3 . 3 Percentage Results of the 1910 General Elections, by Party . . Total Aggregate and Percentage Results of Contested By-Elections in 1911-1914, by Party. Voting on Asquith's Motion for a Select Committee to Investigate General Maurice's Allegations, by Party .......... Number of Candidates in the 1918 General Election, byParty........ ..... ..... Aggregate and Percentage Results of the 1918 General Election ..... ..... Arithmetic Means of the Coefficients of Cohesion for the Coalition, Labour and Liberal Parties in the Chief Divisions of the House of Commons, 1919- 1922 . . . . . . . ............. Arithmetic Means of the Coefficients of Likeness of All Possible Parties in the Chief Divisions of the House of Commons, 1919-1922 ....... Total Aggregate and Percentage Results of Contested By-Elections in 1919-1922, by Party. ..... Number of Candidates in the 1922 General Election, by Party . . Aggregate and Percentage Results of the 1922 General Election, by Party. . . . . ...... Number of Candidates in the 1923 General Election, byParty. . . . . ........... Aggregate and Percentage Results of the 1923 General Election, by Party ..... Number of Candidates in the 1924 General Election, by Party . . vi Page 69 70 82 91 95 101 102 103 110 111 115 116 119 Table 3. 3. 16 17 .18 .19 .20 .21 .22 .23 Aggregate and Percentage Results in the 1924 General Election, by Party . . . Total Aggregate and Percentage Results of Contested By-Elections in 1924-1929, by Party. Number of Candidates in the 1929 General Election, by Party . . Aggregate and Percentage Results of the 1929 General Election, by Party. . . Total Aggregate and Percentage Results of Contested By-Elections in 1929-1931, by Party . Number of Candidates in the 1931 General Election, by Party . . Aggregate and Percentage Results of the 1931 General Election, by Party. . . Five Major Schisms, by Party, Magnitude, and Third Party Beneficiary . . . . . ........... vii Page 119 121 121 122 123 125 126 134 LIST OF APPENDICES Appendix Page 1 The Rational Voter in a Single-Member District........................229 2 British Districts With More Than One Purely Liberal- Labour-Unionist Contest, 1900-1914. . . . . . . . . 236 viii CHAPTER I INTRODUCTION 1. Stability and Change Since the era of William Gladstone and Benjamin Disraeli, the British political arena has usually been dominated by a pair of political parties: Her Majesty's Government (Party) and Her Majesty's Opposition (Party). These leviathans, buttressed by rival organiza- tions in hundreds of constituencies, have virtually monopolized parliamentary and ministerial positions. Independents and minor parties, although not totally devoid of influence, have typically been dwarfed by two major parties.1 A two-party battle occurred in the general election of 1880, the last general election that featured both Gladstone and Disraeli (the latter as Lord Beaconsfield), and in the general election of 1964, the most recent general election (at the time this is written). But although the number of leading roles remained the same, the cast of actors changed between 1880 and 1964. In 1880 Gladstone led the l"Britain may fairly be called the classic home of two-party government. " Leslie Lipson, ”The Two—Party System in British Politics," American Political Science Review, Vol. XLVII, No. 2, (June, 1953), p. 337. 2 Liberal Party to victory over the governing Conservative Party; in 1964 Harold Wilson led the Labour Party to victory over the govern- ing Conservative Party. The displacement of the Liberal Party by the Labour Party, as the alternative to a Conservative Government, is especially strik- ingz since the Liberal Party seemed to have been well established. 3 This decline of the Liberal Party is the focal point of the present study. 11. Historical Background A study of so controversial a subject as the decline of the Liberal Party can hardly expound the accepted ideas about its subject for there scarcely are any. Economists, sociologists, political scientists, historians, politicians and journalists have written about the subject , with different points of view. Commentaries have ranged 2”The most striking change since the turn of the century has been the virtual elimination of the Liberal Party as a political force [ . . ] . " David C. Marsh, The Changing Social Structure of Ejland and Wales, 1871-1951 (London: Routledge 8: Kegan Paul, Ltd., 1958), p. 180. 3"Here was a party that had every advantage--money, pres- tige, able leadership , a glorious history, and a large body of devoted followers--yet it was strangled [ . . . ] . " E. E. Schattschneider, Party Government (New York: Rinehart 8: Company, Inc. , l942),p.79. 4This section is a background sketch, neither a history nor an historiography. from personal apologia to abstract theory. The subject has been approached from an historical perspective focusing upon the British political tradition and from a comparative perspective focusing upon elements shared with other parties in other countries. The pertinent literature is vast, diverse, growing, and neither a definitive history nor a generally accepted theory has been published. Perhaps the only premise shared by all, or nearly all, writers is that the decline of the Liberal Party was somehow intimately linked to the rise of the Labour Party. After it became widely recognized that the Liberal Party not only had declined but also was unlikely to recover, in the late 19208 and early 19305, a trio of explanations of the decline rapidly became conventional. Although none of these interpretations has been dis- tinguished by general acceptance, the trio have remained conventional. These explanations, as a matter of convenience in this study, are called the center party thesis, the electoral system thesis, and the split the sis . The center party thesis was advanced in 1935 by the his- w torian, George Dangerfield,5 and has recently been given a 5George Dangerfield, The Strange Death of Liberal England, 1910-1914 (New York: Capricorn Books, 1961). This volume was originally published in 1935. 4 comparative-theoretical formulation by the economist, Anthony Downs.6 The center party doctrine is to the effect that the Liberal Party declined because its policy position was between the Labour Party's on the Left and the Conservative Party's on the Right. This has been called the ”conventional explanation".7 of the decline of the Liberal Party. The electoral system thesis was prOposed in 1930 by the historian Ramsay Muir, 8 and in recent years has been given a com- parative-theoretical deve10pment by the political scientist, Maurice Duverger.9 The electoral system doctrine is to the effect that the Liberal Party declined because, after it became a third party, the electoral system Operated to its disadvantage. This has been said to be the ”usual argument”10 accounting for the decline of the Liberal Pa rty . 6Anthony Downs, An Economic Theory of Democracy (New York: Harper 8: Brothers, Publishers, 1957). 7Henry R. Winkler, Great Britain in the Twentieth Century, Publication Number 28 of the Service Center for Teachers of History (Washington: American Historical Association, 1960), p. 14. 8Ramsay Muir, How Britain is Governed (New York: Richard R. Smith, Inc. , 1930). 9Maurice Duverger, Political Parties, trans. Barbara and Robert North (New York: John Wiley 8: Sons, Inc. , 1955). 10The original quotation literally reads ”usually argued. " Harry H. Eckstein, ”The British Political System, " in Samuel Beer The split thesis was adumbrated in 1926 by the Leader of the Liberal Party, Herbert Asquith, 11 and has more recently been endorsed by the historian, Charles Mowat,12 but apparently has not been investigated from a comparative-theoretical perspective. The split doctrine contends, roughly speaking, that the split between Asquith and David Lloyd George, dating from 1916, was fatal for the Liberal Party. An analysis of these explanations is deferred to later chapters, but it should be noted at this point that 'the decline of the Liberal Party' does not have the same meaning for each doctrine. The Liberal Party's transition from a position as first or second party to a position as third13 party was in some sense a decline, but although this sense may be encompassed by the center party and split theses, it is excluded by the electoral system thesis. This conceptual problem is complicated by the Liberal Party's kaleidosc0pic history and Adam Ulam (eds. ), Patterns of Government (New York: Random House, 1958), p. 132. 11Earl of Oxford and Asquith, Memories and Reflections (Boston: Little, Brown, and Company, 1928), II. The reference is Asquith's letter of resignation in 1926. 12Charles Loch Mowat, Britain Between the Wars, 1918- 1940 (Chicago: The University of Chicago Press, 1955). 13That is, the party that ranks third in terms of the total national vote . of alliances and schisms. The most significant of these alliances and schisms were the following: (1) After the general election of 1880, the Liberal M.P. 5 contained a contingent of Irish Home Rulers who became followers of Charles Parnell, the Leader of the nascent Irish Nationalist Party, and began to harass the Liberal Government in the House of Commons. Parliamentary harassment became outright Opposition in the general election of 1885, but with Gladstone's espousal of Home Rule for Ireland, a working alliance deve10ped between the Irish Nationalist Party and Liberal Party in 1886. This alliance was durable, 14 and the Irish Nationalist M. P. 8 gave crucial support to minority Liberal Governments in 1892-1895 and 1910-1914. The Irish Nationalist Party was destroyed in the 1918 general election. (2) In 1886 two groups of Liberal M. P. s, the Whig followers of Lord Hartington and the Radical followers of Joseph Chamberlain, combined with the Conservatives to defeat Gladstone's Home Rule Bill in the House of Commons and to defeat the (Gladstonian) Liberal Party—Irish Nationalist Party alliance in the general election. Al- though a few Liberal Unionists (as the followers of Hartington and l4The cohesiveness of, and distribution of power within, the alliance is perhaps be st indicated by the fact that Gladstone was able (successfully) to insist upon Parnell's ouster as Leader of the Irish Nationalist Party in 1891. 7 Chamberlain became known) returned to the (Gladstonian) Liberal Party, most Liberal Unionists remained allied with the Conservative Party. 15 This alliance became an official merger of the Liberal Unionist and Conservative parties in 1912. (3) In 1903 the leaders of the Liberal Party and the Labour Party16 negotiated a secret alliance. Under the terms of this agree- ment, contests between the two parties were minimized in the gen- eral election of 1906. Although not covered by the terms of the agreement, Labour Party M. P. s tended to support the Liberal Government between 1906 and 1910. The agreement was substantially perpetuated17 in the general elections of 1910, and the Labour Party M. P. 8 gave support to the minority Liberal Governments of 1910- 1914. The Labour Party terminated this alliance in 1918. (4) After Lloyd George replaced Asquith as Prime Minister in 1916, the Liberal supporters of L10yd George and the Liberal followers of Asquith became estranged. In 1918 the Lloyd Georgian Liberals, calling themselves Coalition Liberals, fought the general 1l-r’Liberal Unionists and Conservatives shared ministerial Posts in 1886-1892 and 1895-1905. 16Strictly speaking, the Labour Representation Committee, founded in 1900. In 1906 it was baptized as ”the Labour Party. " l7Apparently, no official agreement existed in 1910 but rather the arrangements of 1906 were continued as a matter of con- ven1ence, in a number of constituencies. election in harness with the Conservatives, in Opposition to the (Asquithian) Liberal Party. After the Coalition was dissolved in 1 922, the Lloyd Georgian Liberals fought the 1922 general election a. 8 National Liberals but merged with the (Asquithian) Liberal Party for the 1923 and subsequent general elections. (5) During the events associated with the formation of the National Ministry and the general election of 1931, the Liberal Party divided into three fragments, viz. , a few followers of Lloyd George who refused to support the National Ministry, the Liberal Party under Sir Herbert Samuel who withdrew from the National Ministry in 1932, and Liberal Nationals under Sir John Simon who remained in the National Ministry and allied with the Conservative Party. Thus the meaning of the fundamental concept 'Liberal Party' depends upon the interpretation of a complicated history of alliances and schisms. An interpretation is crucial for explaining the decline of the Liberal Party. The electoral system thesis, for example, PiVOts about when the Liberal Party became a third party. If we accept Asquith's view that the Coalition Liberals were not part of the Liberal Party, 18 then the Liberal Party became a third party in 18Earl of Oxford and Asquith, op. cit. , II, pp. 284-288. Asquith also excludes the National Liberals of 1922-1923 from the Liberal Party. 9 1 9 .l 8, but if we accept Duverger's view that the Coalition Liberals we re part of the Liberal Party, 19 then the Liberal Party became a third party in 1922. Despite these significantly divergent views, a systematic explication of the concept 'Liberal Party' has not pre- v i ously been published. III. The Structure of This Study This study stresses the conceptual and comparative-theo- retical aspects of the decline of the Liberal Party, and in essence, the broad argument of the study is a version Of the split thesis. Chapter II clears the decks for the construction of the argument, With an empirical critique of the center party and electoral system theses, 20 and establishes the critical fact that the Liberal Party manifested no signs of declining before the First World War. Chapter III is an historical exposition of the split thesis, stressing the Pivotal nature of the dispute between Asquith and Lloyd George as Well as suggesting historical parallels with other party systems. Chapters IV and V amplify two facets of the split thesis: Chapter IV presents the split thesis as a comparative theory of the rise and 19Duverger, op. cit., pp. 225, 227. Duverger also in- cludes the National Liberals of 1922-1923 in the Liberal Party. 20A rational critique of the two theses is presented in Appendix 1. 10 decline of major parties in two-party systems. Chapter V reconsid- e rs the roles of Asquith and Lloyd George, from an historial per- spe ctive. The conclusions of the study are presented in the final chapter. It should be noted that this structure entails that some information is repeated in the course of the exposition, but in such instances, the initial presentation is usually somewhat lengthy and subsequent references are rather brief. CHAPTER II RATIONAL VOTERS AND THE PARTY SYSTEM, 1900-1914 I. Introduction "[T]he death of Liberalism was pronounced [when] it was no longer the Left. "1 ”The Liberal party [. . .] was destroyed by the statistical tendency of the single-member district system."2 Those Statements are exceptionally pithy formulations, respectively, of the center party and electoral system theses, conventional explan- ations of the decline of the Liberal Party. 3 Despite the widespread acceptance of these explanations, as this essay seeks to show, the explanations are neither very clear nor very plausible. A close consideration of each doctrine, with especial reference to the period 1900-1914, not only illustrates the shortcomings 1George Dangerfield, The Strange Death of Liberal Eflland, 1i10-1914 (New York: Capricorn Books, 1961), p. 10. 2E. E. Schattschneider, Party Government (New York: Rinehart 8: Company, Inc., 1942), p. 79. 3See Henry R. Winkler, Great Britain in the Twentieth My, Publication Number 28 of the Service Center for Teachers of History (Washington: American Historical Association, 1960), p. 14: Harry H. Eckstein, ”The British Political System, " in Samuel Beer and Adam Ulam (eds. ), Patterns of Government (New York: Random House, 1958), p. 132. 11 12 of the explanations but also illuminates the condition of the Liberal Party on the eve of the First World War. II. A Critique Of the Center Party Thesis The center party thesis is fundamentally a theory about economic classes and politics, a theory with an ancient pedigree and eno rmous variations. When applied to the decline Of the Liberal Party , it has been given a Syndicalist,4 a Marxist, 5 and a Uti‘l.it:a.rian6 guise. The Syndicalist and Marxist versions are closely allied to the view that the decline of the Liberal Party was inevitable;7 that the Liberal Party performed its historic role (whatever that may be) and then perished.8 From this perspective, the replacement of the B0 urgeois Liberal Party by the Socialist Labour Party was the \— 4Dangerfield, op. cit., Part 11, Chapter IV. Rob r5Maurice Duverger, Political Parties, trans. Barbara and e rt North (New York: John Wiley 8: Sons, Inc. , 1955), pp. 213—214, Yo 1- 6Anthony Downs, An Economic Theory of Democracy (New k: Harper 8: Brothers, Publishers, 1957), pp. 48-49, 128-129. Ge 0 7"There have been a number of attempts, notably in Mr. 1‘ ge Dangerfield's brilliant work The Strange Death of Liberal M, to present the death of the Liberal Party as having been an he - p0 Vltable political develoPment. " Colin Cross, The Liberals in W 19 3e) 1‘ (1905-1914) (London: Barrie and Rockliffwith Pall Mall Press, a p. 188. 8The Liberal Party "has been killed by success. " Sir Henry Sle s S '3 r, ”The Liberal Party: History, " in Sydney D. Bailev (ed. ), H., k. k. 13 result of the natural evolution of the sociO-economic-political sys tem.9 These deterministic interpretations, although stimulating, are essentially metaphysical, incapable of confirmation or refuta- tion - 10 The present chapter is concerned with testable interpretations of the center party doctrine, and since there is no canonical version, seve ral formulations are considered. Perhaps the classic version of the center party the sis has bee :1 advanced by George Dangerfield: By 1910, [according to Dangerfield] the Liberals had reach- ed a point where they could no longer advance; before them stood a barrier of Capital which they dared not attack. [. . .] The 1906 elections [. . . had] resulted in a Liberal ¥ Thewitish Party System (2nd ed.; London: The Hansard Society, 19 5 3 ) a p. 84. ecl - c)In Britain ”the two-party system suffered a period of cenltPSe [ . . .] to be reborn later in a new guise approximately in 0p olrmity with the class-struggle pattern of Marxist doctrine: citposltion between a Bourgeois and a Socialist party. " Duverger, pp. Et’u’ p . 213. For an acute analysis of Duverger's use of the ideas of re and natural evolution, see Aaron B. Wildavsky, "A Method- 01 - p O glc221.1 Critique of Duverger's Political Parties, " Journal of W Vol. XXI (1959). pp. 303-313. ine - 10The argument that the decline of the Liberal party was h v 1table seems to be a particular application of what Karl R. POpper a e in _Qalled "historicism" and of what W. H. Walsh has called "mean- Pa. 8 1‘51 history;" roughly speaking, of the view that human history 89 8 through pre-determined stages to a pre-determined end. See 1n b R. Popper, "Prediction and Prophecy in the Social Sciences, " Pr a-l‘.rick Gardiner (ed. ), Theories of Histog (Glencoe: The Free 3‘s, 1959), pp. 276-285; w. H. Walsh, '“Meaning‘ in History," w. , pp. 296-307. 14 landslide. But the Liberal Party which came back to West- minister with an overwhelming majority was already doomed. It was like an army protected at all points except for one vital position on its flank. With the election of fifty-three Labor representatives, the death of Liberalism was pro- nounced; it was no longer the Left. The Conservatives might have consoled themselves with the fact that they represented a logical Right [ . . . ]. 11 A slightly different, but equally brief, version of the doc- trine has been articulated by Maurice Duverger: [T]he Liberals [Duverger contended . . .] had realized the essentials of their programme and so found themselves constrained to adopt a conservative attitude: the appearance of a Socialist party naturally took from them a section of their left-wing support, whilst fear of the 'Reds' threw another section into the arms of the Conservatives [. . .].12 The most complete formulation of the center party the sis, appa re ntly, has recently been presented by Anthony Downs: Before 1900, [according to Downs] there were two major British parties, the Liberals [. . .] and the Tories [. . .]. They were under the usual two-party pressure to converge. However, the enfranchisement of the working class in the late nineteenth century had shifted the center of voter dis- tribution far to the left of its old position. And the Liberal Party, even after it moved to the left, was to the right of the new center of gravity, although it was the more left of the two parties. The founders of the Labour Party correctly guessed that they could outflank the Liberals by forming a new party [ . . .] to the left of the latter, which they did. This trapped the Liberals between the two modes of the electorate, and their support rapidly diminished to \— “Dangerfield, op. cit., pp. 8, 10. 12Duverger, op. cit., p. 214. 15 insignificant size. The crucial factor in this case was the shift of the electorate's distribution along the political scale as the result of the extension of the suffrage to a vast number of new voters, many of whom were near the extreme left. Whenever such a radical change in the distribution of voters occurs, existent parties will probably be unable to adjust rapidly because they are ideologically immobile .13 From the preceding three formulations, the center party thesis may be characterized as follows: The Reform Acts of 1867 and 1884 enfranchised workingmen in sufficient numbers to constitute a majority of the electors, but even after these Acts, neither the Lib- eral Party nor the Unionist Party14 represented the interests of the working class, although the reformist Liberal Party was more representative of workingmen than the conservative Unionist Party. Furthermore, even after the Labour Party was formed to represent the interests of the working class, neither the Liberal Party nor the Unionist Party pre-empted the prOgram of the Labour Party. Thus workingmen naturally transferred their votes to the Labour Party as the only party representing their class interests, and since most h. l3Downs, op. cit. , pp. 128-129. "This is roughly what -_, happened in the case of the Labour Party. " Ibid. , p. 129. 14The Unionist Party is called the Conservative Party or Tory Party by most writers--e. g. , apparently by Dangerfield, §:Verger, and Downs--but the distinction between the Unionist e1 rty and the Conservative Party is of some significance for the ectoral system thesis. This point is discussed below in Section of the present chapter. 16 workingmen had previously voted for the Liberal Party, the Liberal Party declined. That decline was accentuated by the appearance of revolutionary socialism, with the 1917 Bolshevik Revolution in Russia, which frightened some more voters away from the reformist Liberal Party to the avowedly conservative Unionist Party. This argument is sometimes supplemented with additional considerations, some of which are discussed below, but the core of the argument is the presumption that most working class electors voted for their class interests. The center party thesis, as interpreted above, is testable in principle, but in practice, a decisive test is difficult (perhaps impossible) since the extant historical data are limited. For example, the thesis is "in principle" susceptible to testing by survey methods, but pertinent survey data are non-existent.15 Nevertheless, since Downs has develOped the doctrine as a general theory that is suppos- edly applicable to other times and other places, an imaginary survey test deserves consideration before turning to the task of testing the thesis with actual data. This hypothetical survey elucidates the logic \— d 1'-"’The history of scientific election surveys is customarily Aated from the Gallup Poll of 1936 in the United States. See Robert - Alford, Party and Society (Chicago: Rand McNally 8: Co.. 1963). 17 of the thesis and, perhaps, may prove suggestive for comparative tests of the general theory. 16 An actual questionnaire, of course, should be extensively pre-tested and the results should be subjected to tests for statistical significance. But for an imaginary questionnaire to have been used in a survey (let us say) in 1910, it is sufficient to sketch the general intent of the key questions and the broad nature of the analysis. The crucial questions would be of the following type: 1. PeOple often Speak of the working class, the middle class, the upper class, and so forth. To which class do you belong? 2. (a) Which party do you believe best represents the inter- ests of the class? (b) Which party do you believe least represents the interests of the class? 3. Did you vote in yesterday's election? (a) For which party did you vote? [AND IF APPROPRIATE.] (b) The party did not run a candidate in your par- liamentary district in yesterday's election. If the party had run a candidate, would you have voted for that candidate? A questionnaire of this type pivots about the self-definition of the respondent and might be criticized by a Marxist or a Freudian for not elucidating the real class or true class interests of the -\——— 16For a critical analysis of the general theory in the light 3f recent survey data from the United States, see Donald E. Stokes, Spaftial Models of Party Competition, " American Political Science w, Vol. LVII, No. 2 (June, 1963), pp. 368-377. 18 respondent. 17 However, to the extent that the center party thesis is about the voter's subjective rather than objective situation, this type Of questionnaire, administered to a representative sample of the total British electorate in 1910, would have been fairly decisive. If the center party thesis is valid, then the following sort of results would have been expected, i.e. , would have confirmed the thesis: (1) Persons who answer "working class” to question (1) should respond "Labour Party" to question (2a) and "Union- ist Party" to question (2b). Persons who answer ”upper class" to question (1) should respond ”Unionist Party" to question (2a) and "Labour Party" to question (2b). Persons who answer "middle class” to question (1) should respond .( less solidly) "Liberal Party" to quest ion (2a) and be divid- ed between "Labour Party" and "Unionist Party" in respond- ing to question (2b). These results would establish the Left- Right cortinuirn for the parties that is assumed by the center party thesis and also would establish the postulated con- nection between the parties' continuum and the class struc- ture. (2) Persons who answer "Labour Party” to question (2a) should respond "Labour Party" to question (3a) if the Labmr Party ran a candidate in their district; otherwise, they should respond "yes" to question (3b). An analo- gously consistent result should occur for persons who answer ”Liberal Party" or "Unionist Party" to question (2a). The response given to question (2b) should never be given to question (3a); these results would establish that the electors voted for their class interests as assumed by the center party thesis. _____~‘_______. 17The Marxist doctrine of false class consciousness is well- kn?“’n. Dangerfield, op. cit., p. 235, repeatedly referred to uncon- :Clous impulses and instinct. These views often seem to function as %x machina. it 19 If some of these results did not appear in the hypothetical survey analysis, then the center party thesis would be suspect. The center party thesis may be more narrowly and precisely stated in terms of the hypothetical questionnaire: Persons who would have answered ”working class" to question (1), would also have answered "Labour Party” to questions (2a)and(3a) or ”yes” to(3b). (Ques- ticn(3b) is especially important for the imaginary 1910 survey since in 1910 the Labour Party fielded candidates for only about ten percent of the seats in the House of Commons.) When stated in these terms, the center party thesis is perhaps less plausible than usual, but this imaginary survey is hardly persuasive. Although in practice the validity of the center party thesis cannot be directly examined by survey methods, the doctrine can be assessed by other, less direct and less conclusive, means. The nub 0f the center party explanation of the Liberal Party's decline is the working class elector who, when given the Opportunity, ”naturally ”18 transferred his vote from the Liberal Party to the Labour Party. That vote transfer was at least accelerated and at most caused, so the explanation runs, by the Liberal Government's failure to ”steal \— l8Duverger, op. cit., p. 214. all "P ll. 6.! f1; rL’ 20 the thunde r" of the Labour Partylg and by the deterioration Of the conditions of working class life under the Liberal Government from 1906 to 1914.20 These supplementary hypotheses Of the doctrine are examined before turning to a consideration of the putative vote switching . The social reform program Of the last Liberal Government has frequently been both under-estimated and Over-estimated by preponents of the center party explanation: Its magnitude has been under-estimated; its doctrinal content has been over-estimated. The volume of social legislation enacted was immense, and the content, rather than forming a completed doctrinaire program, was quite pragmatic, dependent upon the personalities in the Liberal Cabinetz1 ”That the Liberals did move to the left in the years 1906- 1914 cannot be denied."22 19The Liberal Government "dared not attack [. . .] Capital, " was "constrained to adept a conservative attitude, " and was "ideologically immobile." Respectively, Dangerfield, op. cit. , P- 8; Duverger, op. cit., p. 214; Downs, op. cit., p. 129. . 20Marxist doctrine, of course, postulates such a deteriora- tlon- Dangerfield, Op. cit. , Part II, Chapter IV, argued that it actually occurred. 21For a general history, emphasizing the social legislation of the Liberal Government, see Cross, op. cit. A ZZAnthony King, "The Decline of the British Liberal Party," paper read at the American Political Science Meeting. 1962. p. 5 - "The acts passed by Liberals in power since the end of 1905 21 This was the period of what Professor Beer has called the Liberal 'New Deal.’ The years before 1914 saw the intro- duction of old age pensions and medical and unemplOyment insurance, the extension of workman's compensation, the e stablishment Of minimum wages and maximum hours in a number of industries, and the imposition of hitherto unheard of rates of direct taxation. 23 When the Liberal Chancellor of the Exchequer, Lloyd George, intro- duced the 1909 Budget "[ i]t was at once remarked by [Phillip] Snowden and others, that a great many features of the budget seemed to have been 'cribbed' from a pamphlet by Snowden entitled The ficialist Budget. ”24 The Budget and its author were praised by S Orne Soc ialists: We shall be quite frank about Mr. Lloyd George's Budget; it is splendid. Two minor defects apart, the Budget is not only more than we had dared to h0pe [ . . .] but almost as \— five 118 moderate experiments in social legislation, but they gave a aeclsIVe impulse to state intervention and may perhaps be regarded s the foundations of the Welfare State. ” Thomas Jones, Lloyd W (London: Oxford University Press, 1951), p. 39. E 23King, op. cit., pp. 5-6. The reference is to Samuel H. eer, "Great Britain: From Governing Elite to Organized Mass 1: . (grfiles, " in Sigmund Neumann (ed. ), Modern Political Parties i hula-go: University of Chicago Press, 1956), p. 36. (The above st of enactments is not exhaustive.) 24D. C. Somervell, British Politics Since 1900 (New York: 0 xford University Press, 1950), p. 66. Snowden was a Labour Party . ° 13- in 1909 and became Chancellor of the Exchequer in the minor- 11: y Labour Governments of 1924 and 1929-1931. 22 much as we could have expected from a Socialist Chan- cellor in his first year of office [. . ..] We cannot deny, and we have no intention Of denying that the author of the present Budget is good enough statesman for a Socialist to support during the next five or ten years at any rate. 25 "There was no period in the long history of Conservative Govern- ments [declared Arthur Henderson, a Labour Party M.P. , in 1909] which, in the output of beneficient social legislation, would compare with the last three years.”26 These legislative achievements have been recognized but discounted by some center party theorists, notably by Dangerfield: [ The Liberal Government had come forward] with a shining procession of social reforms--a Workmen's Compensation Act, an Old Age Pension Act, a Miners' Eight Hours Act, a Trade Boards Act [ .] And yet--what could be the matter? --wages never went up. In fact, they continued to fall. [. . .] [T]he workers of England ['8 . . .] political barometer was wages and nothing more [. . . ] . 27 This statement, if about the course of money wages. is simply in errOr. 28 From 1885 to 1914, there was a secular trend of rising \— E . 2'5New Age (Socialist), May 6, 1909. Quoted in s. Macoby, wRadicalism 1886-1914 (London: George Allen 8: Unwin Ltd. , 53). p. 502p. 26Speech at Wigan, April 6, 1909. Quoted in The Liberal M . p 3‘ azine, 1909 (London: The Liberal Publication Department, 1910), '. 2'55. ”The Labour Party had become a sort of admonishing left fig-lgffl the Government, supporting its sick policies with all the ezlégy of a slightly cantankerous nurse." Dangerfield, Op. cit. , 27Ibid., pp. 225, 228. 23see below Table 2. 2. In fairness to Dangerfield, it should 23 money wages. In each successive year, from 1910 to 1914, 29 the index for money wages increased. Since the statistics for money wages, given in Table 2. 2, relate only to employed persons, the trend in money wages should be considered in connection with the trend in employment. A rough index for the direction of changes in unemployment is presented in Table 2.1. These figures indicate that unemployment was relatively high for 1905-1909 but was rela- tively low for 1910-1914.30 A stronger inference is probably not Justifiable, given the nature of the unemployment data, but the direction of change for 1905-1914 is significant: During the latter half 0f the Liberal Government's tenure, the level of unemplOyment deClined . The course of real wages (i.e. , money wages adjusted for the COSt of living) is another, perhaps better, indicator of the c:ha‘nging conditions of the working class. 31 With respect to real \—- b e noted that the standard work on the course of wages--A. L. Owle‘)’. Wages and Income since 1860 (Cambridge: Cambridge nlversity Press, 1937)--was published after the original publication 0 f 13E”lgerfield's book. 29Dangerfield's book focused on 1910—1914. a1 30see below Table 2.1; compare 1905-1909 with 1892-1896; 8° compare 1910-1914 with 1897-1901. 31see below Table 2. 2. 24 wages, Dangerfield was probably correct (although the data have a large margin of error) that ”in 1910 the English worker wasa poorer man than he was in 1900.”32 However, the years 1899-1900 are misleading bases for a comparison: During the Boer War, real wages reached their highest point in the entire period 1885—1914. 33 Real wage s, so far as can be reconstructed, were rather stable during the years 1905-1914. 3‘ZDangerfield, op. cit. , p. 218. See also below Table 2.2. 33See below Table 2. 2. Four series--money wages with and without an adjustment for unemployment and real wages with and Without an adjustment for unemployment--for 1850 to 1902, con- 8tructed by G. H. Wood, also show 1900 as the peak year for the ent_ire period, for each series. See B. R. Mitchell, Abstract of Wstorical Statistics (Cambridge: Cambridge University Press, 1962), pp. 343-344. This selection of an extreme as a base for a Comparison is only one of many examples that could be cited Concerning Dangerfield's highly selective and quite misleading use of statistics; this particular flaw also vitiates the comments on feal Wages by G. D. H. Cole, British Working:Class Politics 1832- & (London: George Routledge 8: Sons, Ltd., 1941), pp. 210ff. 25 TABLE 2 .l PERCENTAGE OF UNEMPLOYED IN CERTAIN BRITISH TRADE UNIONS, 1888-1914* Year Percentage Year Percentage 1888 4.9 1901 3.3 1889 2.1 1902 4.0 1890 2.1 1903 4.7 1891 3.5 1904 6.0 1892 6.3 1905 5.0 1893 7.5 1906 3. 6 1894 6.9 1907 3.7 1895 5.8 1908 7.8 1896 3.3 1909 7.7 1897 3.3 1910 4.7 1898 2.8 1911 3. 0 1899 2.0 1912 3.2 1900 2.5 1913 2.1 ————______. 1914 3.3 (a) *Mitchell, op. cit., pp. 64-65. ar About these statistics, the source remarks: ”Before 1922. there Whe' no go neral statistics of unemployment. But certain trade unions to 3:1“ Pald unemployment benefit to their members, did make returns indue lfabour Department, and these returns cover a wide variety of ret Strles back to 1888. [. . .] The percentages computed from these the urns have been commonly used as a general indicator of unemploy- is :b. _ In view of the small number of workers covered by them this of 8kY1 M. P. 5. Sir Donald Maclean has sta'rking described the Party meeting 0f the twenty-three89 Liberal M- P- S "on the day before the House [of Commons] met after the 1918 election:" We were all sore and disheartened, blaming everybody and everything, and we were united only on one thing, and that was on the moral iniquity Of the election. Francis Acland took the chair, and somebody prOposed that 'We do not form a party. ' After this hopeful and cheerful beginning, which stimulated everybody--(laughter)- -we had a long and heated discussion. After that had lasted about three or four hours, I pointed out that if we went out of the door without passing a resolution to form a party, Liberalism would go under for a generation. Thereupon the resolution to form a party was carried by a large majority. By this time everyone was so exhausted by the debate that someone seized the opportunity to prOpose me as chairman, and it was carried nem con. The next day we faced a House hostile and, indeed, derisive90 This small band Of Liberal M. P. s countered the larger Labour Party's claim to be the Official Opposition, and within the House Of Commons, 7 88From Sir Donald Maclean's remarks of April 26, 1920 at a dinner honoring him for his service as Chairman Of the Parliamentary Liberal Party, 1919-1920. Quoted in the Liberal Magazine, 1920, p. 221 . Maclean had previously served as a junior Liberal Whip and then as Deputy Chairman of Ways and Means. 89Maclean spoke of "[a]bout twenty-eight [ . . .]." Ibid. , p. 220. However, only twenty-three names are listed in the Liberal Migazine, 1919, p. 89. 90Liberal Magazine, 1920, p. 220. The fact that the resolution to select Liberal Whips was carried only "by a large majority" rather than by acclamation is noteworthy. 97 Maclean was soon unofficially recognized as the Leader of the Opposition. 91 The chief divisions in the House Of Commons found the Liberal and Labour parties united in Opposition to the Coalition Government from 1919 to 1922,92 but this cooperation did not extend to the con- stituencies. Liberal and Labour contests in by-elections, more Often wOn by Labour, were relatively frequent. In January-February, 1920 Lord Haldane, the former Liberal Lord Chancellor and future Labour Lord Chancellor, publicly addressed himself to the relationship Of Liberalism and Labour: "There are some who believe that Liberalism and Labour might be got to work together, but that is plainly impos- sible. [. . .] And it is with Labour that the hope lies tomorrow. "93 Haldane took "the numerical representation in the House of Commons at the moment and the success of Labour in recent by-elections as his justification, "94 but his remarks were contested by Asquith: 91See Bonar Law's remark in the House Of Commons on May 14, 1919. Liberal Magazine, 1919, pp. 283-284. This was partially a result of Maclean's skill as a parliamentarian and partially a result of the fact that, due to the defeat of Labour's most distinguished par— liamentarians in 1918, the Chairman of the Labour M.P. s was "the Obviously incompetent Adamson [ . . . ]." Pelling, Op. cit. , p. 49. 925ee below Tables 3.8 and 3.9. 93Newspaper interview of February 5, 1920. Liberal Mag- azine, 1920, p. 86. 94Reference to Haldane's speeches Of January 13, 1920 at the 98 What does Lord Haldane say? What is his advice to the young politician? He tells you [ . . .] that Labour has captured the heights and that we Liberals are on the plains below. [. . .] Let me say [ . . .] that the battle of politics is carried on, not on the heights, but in the plains.95 Edinburgh Liberal Club. Ibid. , p. 23. In those speeches, Haldane said that the Labour Party had "out-sped" the Liberal Party, that the Liberal Party "was only the third party in the State, " and that teh "Liberalism of the future was Liberalism of the spirit. " I__ip_id. pp. 23- 24. Emphasis added. 95Speech at Cambridge, January 23, 1920. _I_b_i_d., p. 25. Asquith's reply, a play upon Haldane's metaphor of the "heights and plains," could hardly be rationally persuasive. Indeed, when due allowance is made for Asquith's position as Leader of the Liberal Party, Asquith had already said much the same thing as Haldane: "I am quite aware--evidence reaches me from many different quar- ters--that some among the more ardent of ouryounger spirits are tempted, under present conditions, to doff their Old uniforms, and_tp migrate to the Labour camp. The Labour Party and ourselves have many aims in common, and in not a few of the controversies, actual or imminent, of the immediate future, we shall, I am certain, find in them, as they find in us, associates, and more than associates, allies; but I desire to say, with the utmost clearness and emphasis, that I Le_gard the maintenance, with its sgparate identity and its separate traditions, of the Liberal Party, as of the first importance to the State. We are not, and we never have been, a party Of iron- clad dogmas or exclusive tests. We have always been ready, and I hOpe we always shall be, to profit by the teaching of experience, both international and domestic, in this, which is an ever-changing and ever-develOping world." Asquith's speech to the Yorkshire Liberal Federation, June 19, 1919. Liberal Magazine, 1919, p. 362. Empha- sis added. In a speech Of November 15, 1919 to the Oxford University Liberal Club, Asquith stated that "[ s] ome peOple seemed to think, quite honestly, that the future Of this country would lie between two combinations--Labour upon one side, and for want of a better term, what he might call a Bourgeois Party on the other, and that this was the only choice before us. [. . .] 'If one were disposed to be a pes- simist [recent events in the Liberal Party] would fill one with de- spair. And if I am not pessimistic--as I am not--it is because I 99 Despite Asquith's protestations, the by-election results supported Haldane's conclusion. When the results of the contested by-elections of1919-1922 are compared with the results of the 1918 general election, the Liberal Party's prOportion of the total poll increased slightly from twelve to sixteen percent, but the Labour Party's pro- portion of the total poll rose sharply from twenty-two to thirty-three percent.96 The Coalition Government was not seriously threatened by either the Liberal Opposition or the Labour Opposition.97 Within Parliament, the Coalition Liberals and Unionist Party not only had a have faith and hOpe in the new generation of which this club is going to be to this ancient university the centre and rallying point. My advice to them and my final word to them at this their inaugural meet- ing is very simple and plain, but, as I think, very urgent. Keep your Liberalism clean, firm, resolute, buoyant, and be assured that so long as you do that you have with you, whatever may be the passing accidents of the hour, the promise of the future. "' Ibid. , pp. 625- 626. (Oxford University, of course, was the traditiO-n-al source of young Liberal politicians.) Perhaps Asquith was not literally pes- simistic, but he certainly was discouraged, as he himself later admitted. See Earl of Oxford and Asquith, Op. cit. , II, p. 284. Haldane's remark about the impossibility of Liberal and Labour fusion was also consistent with Asquith's earlier statement that "we are not [ . . . ] , and we do not intend to be, a wing either of the Tory or Labour Parties. " Speech to the National Liberal Federation, November 28, 1919. Liberal Magazine, 1919, p. 628. 965ee above Table 3.7 and below Table 3.10 97There was no single official Opposition, hence the unusual capitalization. 100 substantial majority but also were quite cohesive, as was reflected by the endemic talk Of permanent fusion98 and by the chief divisions in the House Of Commons.99 Furthermore, the Coalitions, propor- tion of the total poll, when the 1919-1922 by-elections are compared with the 1918 general election, declined only from forty-eight to £0 rty—four percent. 100 Nevertheless, the Coalition Party101 had difficulties with constituency organizations. 98Leaders of both wings of the Coalition favored closer co- operation. For the point of view Of the Coalition Liberal leadership, see the Liberal Magazine, 1919, pp. 427-429; Liberal Magazine, 1920, pp. 147-148; Liberal Magazine, 1921, pp. 87-89; Liberal Magazine, 1922, pp. 33, 42-43. For the point of view of the Unionist Party leadership, see R. T. McKenzie, British Political Parties (New York: St. Martin's Press, Inc., 1955), pp. 83-109. 99see below Tables 3. 8 and 3.9. 100See above Table 3.7 and below Table 3.10. 101The Coalition was quite commonly regarded as a single party. For example, see Austen Chamberlain's remark, quoted in McKenzie, op. cit., pp. 92-93. See also the Liberal Magazine, 1922, p. 212. 101 TABLE 3 . 8 ARITHMETIC MEANS OF THE COEFFICIENTS OF COHESION FOR THE COALITION, LABOUR AND LIBERAL PARTIES IN THE CHIEF DIVISIONS OF- THE HOUSE OF COMMONS, 1919-1922* Political 1919 1920 1921 1922 Party N=20 N=25 N=32 N=46 Coalition . 769 . 776 . 827 . 832 Unionist . 870 . 801 . 868 . 850 Coalition Liberal . 632 . 805 . 679 . 760 Labour .925 .940 .956 .961 Liberal .870 .873 .907 .896 *For the method of calculating coefficients of cohesion and of likeness, see Stuart A. Rice, Quantitative Methods in Politics (New York: Alfred A. Kn0pf, 1928). The voting breakdowns were drawn from the regular monthly section, "The Divisions of the Month, " in the Liberal Magazine, 1919-1922, beginning with the July 1919 issue and ending with the September 1922 issue. The Liberal Magazine, an official publication of the Liberal Party, called these divisions "the more important divisions," "the chief divisions in the House Of Commons. " (In the Liberal Magazine, 1919, the curi- osity on page 390 was omitted and the breakdowns on pages 337-341 were included. ) 102 TABLE 3. 9 ARITHMETIC MEANS OF THE COEFFICIENTS OF LIKENESS OF ALL POSSIBLE PARTIES IN THE CHIEF DIVISIONS OF THE HOUSE OF COMMONS, 1919-1922* Political 1919 1920 1921 1922 Party N=20 N=25 N=32 N=46 Liberal— Labour . 903 . 847 . 936 .946 Unionist-Coalition Liberal . 818 .856 . 869 . 873 Labour-Coalition Liberal .. . 344 . 262 . 248 . 231 Liberal-Coalition Liberal . 297 . 249 . 271 . 271 Labour-Coalition .235 . 206 .157 .153 Liberal-Coalition . 212 . 198 . 180 .194 Labour-Unionist . 207 .179 .131 .130 Liberal- Unionist .175 .182 .135 .172 a a *See the note to above Table 3.8. The 'Coalition' in these tables is the source's 'Unionists' combined with the source's 'Coalition Liberals. ' 103 TABLE 3 .10 TOTAL AGGREGATE AND PERCENTAGE RESULTS OF CONTESTED BY-ELECTIONS IN 1919-1922, BY PARTY* Political Aggregate Percentage Party Results Results Coalition Votes 737, 828 43. 8 Seats 42 55. 3 Liberal Votes 271,186 16.1 Seats 7 9 . 2 Labour Votes 554,484 32. 9 Seats 21 27 . 6 Other Votes 122, 509 7. 3 Seats 6 7. 9 Totals Votes 1,686,007 100.0 Seats 76 100. 0 *Liberal Year Book, 1923, p. 164. 104 The Coalition Liberals attempted tO capture the constituency units (local Liberal Associations) of the National Liberal Federation. This effort was contested since the National Liberal Federation adopt- ed the policy that every Liberal Association should jealously guard its independence, and should, without loss of time, cause it to be clearly understood that at the next general election its countenance and support will only be given to a Liberal can- didate who is independent Of other political ties and claims .102 In Wales, where Lloyd George was President of the Welsh National Liberal Council, the Coalition Liberals succeeded and excluded every (Asquithian) Liberal from the Executive Committee Of the Welsh National Liberal Council. The Liberal Party countered by setting up a rival organization, the Welsh Liberal Federation. 103 Elsewhere, with isolated exceptions, the Coalition Liberals were repulsed by the (Asquithian) Liberals. 104 The Coalition Liberals, expelled from the lozProceedirigg of the Thirty-Sixth Annual Meeting of the National Liberal Federation, 1919, p. 16. 103'Liberal Magdzine, 1921, pp. 20—25. For Wales. see 3180 above footnote 42. 104For an early example, see above footnote 46. The Spen Valley by-election of December 20, 1919 provided another relatively early and especially noteworthy case. The by-election resulted from the death Of Sir Thomas Whittaker, M. P. , a Coalition Liberal. By a majority vote the Spen Valley Liberal Association decided not to nominate a Coalition Liberal and, subsequently, by an unanimous vote nominated Sir John Simon. The Coalition Liberal Whips found a can- didate to stand as a Coalition Liberal, and this action provoked a widespread protest from Liberal Associations and Clubs. The Labour 105 Liberal Party, 105 formed a National Liberal Council in January 1922, with Lloyd George as President and Churchill as Vice President. Since the die-hard Unionists were virulently criticizing the Coalition by this time, the National Liberal Council had a dual purpose, in- dicated by Sir Gordon Hewart, M. P.: [I]t is of vital importance that we should have a strong Liberal organization in existence so long as the Coalition continues. [ . . .] [I]t is no less vitally important we should have a strong Liberal organization in being when the Coalition comes to an end. 106 Austen Chamberlain, who had succeeded Bonar Law as Unionist leader in the House of Commons, and the Unionist members of the Cabinet, with the exception of Stanley Baldwin, we re loyal to Party candidate won the by-election; the Coalition Liberal finished third and last. See the Liberal Magazine, 1919, pp. 712-715. 105"Although the statement that the Coalitionists have been 'expelled' is not literally true, we may as well make them a present of this point. What is literally true is that they and their actions during the past three years have been repudiated by general Liberal Opinion. The local Liberal Associations refuse to nominate them for another election unless they come out of the Coalition." Liberal Magazine, 1922, p. 28. "Mr. Lloyd George's 'unforgivable sin' is plainly visible to everyone. It is the fact that he has supported Tories ggainst Liberals, that he has done so deliberately, repeatedly, and consistently for more than three years, and still continues to do so. " ib_i_d. , p. 31 . 106Ibid. , p. 31. For a detailed description of the founding Conference of the National Liberal Council, see Ibid. , pp. 28-43. 106 Lloyd George, 107 but die-hard Unionists, critical of the Coalition, were powerful in the Unionist constituency organizations and the National Unionist Federation. Chamberlain agreed to meet his critics at a meeting of the Unionist M. P. s, scheduled for October 19, 1922. When, contrary to Chamberlain's advice, the Party meeting voted--187 to 87--to fight the next general election as an independent Party, the Coalition Government re signed. Bonar Law, who had come out of retirement to address the Party meeting and who favoredindependence, became Prime Minister and Leader of the Unionist Party. 108 The general election of November 1922 was contested by four significant parties: The Unionist Party, Liberal Party, Labour Party, and National Liberal Party. 109 Since some prominent Unionists were 107"Chamberlain's willingness to consider fusion seems to have been much influenced by his early experience of fusion between the Conservatives and Liberal Unionists. " McKenzie, Op. cit. , p. 92n. The earlier parallel was also noticed by contemporary Libe rals: "It is in vain that Coalitionists who speak for Mr. Lloyd George's half Of that lop-sided body profess and call themselves Liberals. They are precisely in the same position as the 'Liberal' Unionists of 1886 onwards--lost to the Party to which they once owned alle- giance, and in process of being absorbed by another. " Liberal Magazine, 1921, p. 87. 108For an excellent analysis of these events, see McKenzie, Op. cit. , pp. 83-109. Although the term 'Conservative' had replaced the term 'Unionist' by this time, the latter term is consistently used in this chapter. 109With the independence of Southern Ireland, a significant 107 not only benevolently aloof from the Unionist Government but also were benevolently inclined towards LlOyd George, Bonar law did not sanction Official Unionist Opponents for the National Liberals, but Lord Beaverbrook helped launch some fifty- six unofficial Unionists against Lloyd George's partisans. 110 The Liberal and Labour parties, encumbered by neither Official nor unofficial alliances, were sturdily independent. There was little pretense that the Liberal Party could win the general election of 1922. Reginald McKenna, the former Liberal Chancellor of the Exchequer and future Unionist Chancellor-dc signate, publicly endorsed Bonar Law's Government on October 24, 1922 and averred "that there was no alternative Government except a Labour Government [. . ..] "111 Viscount Grey, the former Liberal Foreign Irish party was not involved for the first time since 1885. 110Lord Beaverbrook, The Decline and Fall of Lloyd George (New York: Duell, Sloan and Pearce, 1963), pp. 213-214. Lord Beave rbrook also noted that, when faced with Unionist Opponents, "only two of the Lloyd George candidates were successful." lipid. , p. 214. lllLiberal Mgazine, 1922, p. 754. The quotation is not direct. In a public speech, during the campaign, Asquith indirectly conceded the accuracy of McKenna's comment: "He [McKenna] told them [the Unionists] some things that were very true--in fact I know of no thiflgthat were not true--but nothing that was very new. " li>_i_c_1. , p. 761. Emphasis added. Presumably, Asquith had mainly in mind McKenna's views about fiscal responsibility, but at a minimum, the emphasized phrase is a revealing slip of the tongue. 108 Secretary and a loyal follower of Asquith, speaking at a public meet- ing on October 24, 1922, was also quite candid about the Liberal Party's prospects: [A] Liberal Goverment is what I would prefer [. . .] Well, people say to me, 'Oh you know that is not really practical politics; the Liberal Party has so few seats and has to gain so many before it could get a majority that it cannot really have a practical prospect of getting an independent majority and making an Independent Liberal Government such as we used to have. ' My answer to that is that we need not stop to think about the chances. [ . . .] If we are to get--as I hope we shall and think we must get--a better House of Commons as the result of this election, one which is more thoroughly representative Of feeling in the country, it is essential that in that House Of Commons there should be a very largely increased Liberal representation. That is enough for us at the moment, and it is enough for every Liberal candidate. HZ The number of candidates run by each party, given in Table 3. 11, supports McKenna's conclusion and emphasizes Grey's modest hopes. The Liberal Party needed to elect virtually all Of its candidates to Obtain a working parliamentary majority; only the Unionist and Labour parties fielded enough candidates to reasonably hOpe to Ob- tain a majority in the new House of Commons. The results Of the general election, given in Table 3.12, further confirmed McKenna's view. The Liberal Party polled only nineteen percent of the total vote. The National Liberal Party received ten percent Of the total vote. The Unionist Party, with thirty-nine percent of the poll, 112Ibid., p. 744. Emphasis added. 109 continued to form the Government. The Labour Party, with thirty percent Of the vote, became the official Opposition in the new House of Commons . 110 TABLE 3.11 NUIVIBER OF CANDIDATES IN THE 1922 GENERAL ELECTION, BY PARTY* Political Number of Party Candidate 3 Unionist 491 Liberal 331 Labour 411 National Liberal 139 Others 47 Total 1419 *Liberal Year Book, 1923, p. 182. 111 TABLE 3.12 AGGREGATE AND PERCENTAGE RESULTS OF THE 1922 GENERAL ELECTION, BY PARTY* Political Party Aggregate Percentage Party Results Results Unionist Votes 5,565,843 38.7 Seats 347 56.4 Liberal Votes 2,683,371 18.7 Seats 64 10.4 Labour Votes 4,251,011 29.6 Seats 142 23.1 National Liberal Votes 1,429,641 9.9 Seats 53 8. 6 Other Votes 446,492 3.1 Seats 9 l. 5 Totals Votes 14,376,358 100.0 Seats 615 100. 0 *Liberal Year Book, 1923, p. 184. (a) "There were 253 con- tests in which more than two candidates competed for one seat. Some of these were three-cornered, and some four- and even five-corner- ed. " 139 won by Unionists; 76 won by Labourite s; 20 won by Liberals; 18 won by National Liberals. Ibid. , p. 184. (b) Forty-four percent (61 /139) of the National Liberal Party candidates were Opposed by Liberal Party candidates; this includes 17 straight fights between the two parties. Computed from Ibid., pp. 190-292. Nevertheless, some scholars have added the National Liberal vote and the Liberal vote, treating the result as the vote received by a single party. See, for example, Lakeman and Lambert, op. cit., p. 26; Duverger, Op. cit., 1 12 The Unionist Party majority in the House Of Commons was substantial after the 1922 general election, but "a number of the best minds in the party were outside the Government [. . .]."113 When Baldwin succeeded Bonar Law as Prime Minister and Leader of the Unionist Party in May 1923, the Party was far from united since the pro-Coalition Unionists, such as Austen Chamberlain, had not broken their informal political alliance with Lloyd George's National Liberals. Baldwin's decision to fight a general election in 1923 on the issue of Protection drove a wedge between the pro- Coalition Unionists and the National Liberals, reuniting the Unionist Party: Austen Chamberlain rallied to the cause Of his father and enthusiastically supported Baldwin's appeal. Lloyd George and the 114 merged with the National Liberals, with individual exceptions, Liberal Party and fought the general election as Free Traders. The Labour Party also challenged the Unionist Government's tariff policy and denied Baldwin's claim that tariffs were a remedy for unemploy- ment. 115 pp. 225, 227. The Lakeman-Lambert classification, in the Opinion of the present writer, is a gross error. (c) The Constitutional Year Book, 1923 gives the following figures for the Liberal Party: 2,530,822 votes, 53 seats. $1.9 , pp. 254, 248. 113'McKenzie, op. cit., p. 110. 114Winston Churchill was the most distinguished exception. 115$ee the Labour Party's Manifesto, reprinted in the Liberal 113 The Liberal Party, for the first time since the second general election of 1910, made a serious bid for power in the 1923 general election. 116 "After some preliminary conversations, Mr. Asquith, Mr. Lloyd George, Sir Alfred Mond, and Sir John Simon held a con- sultation on November 13th, and came to a complete agreement for united action. "117 The following Official statement was issued to the Press: Arrangements are now completed for all Liberals to fight the coming election as a united party. Both in the constituencies and at headquarters all candidates will be adopted and des- cribed as Liberals, and will be supported by the whole strengh Of the party without regard to any past differences. Mr. Asquith and Mr. Lloyd George at their meeting this morning settled plans for the campaign in common, and it is already certain that Liberal candidates will go to the poll in such numbers as to make united Liberalism a practical alternative to the present Government. [. . . ]11 The results Of this united action represent an astonishing achieve- ment for a third party in a two-party system: The Liberal Party's Magazine, 1923, pp. 719—721; Liberal Year Book, 1924, pp. 173— 175. Perhaps, the Labour Party's confidence in 1923 is be st illus- trated by the Manifesto's phrase: "When Labour rules [. . .]." M” p. 175. 116For the first time since 1910, the Liberal Party ran candi- dates for more than fifty-four percent of the seats in the House of Commons. Indeed, for the first time since 1910, the Liberal Party fielded more candidates than the Labour Party. See above Tables 3.1, 3.6, 3.11 and below Table 3.13. 117Libera1 Magazine, 1923, p. 710; Liberal Year Book, 1924. p. 168. 6 118Liberal Magazine, 1923, p. 710; Liberal Year Book, 1924. p. 1 8. 114 prOportion of the total poll rose markedly from nineteen percent in 1922 to thirty percent in 1923.119 This achievement, rather Obvious- ly, was made possible by the 1922 split in the Coalition Party which led Lloyd George and his National Liberal Party to fuse with the Liberal Party. 120 119See above Table 3.12 and below Table 3.14. l‘ZOFor the comparative significance of such splits, see below Section VI. 115 TABLE 3.13 NUMBER OF CANDIDATES IN THE 1923 GENERAL ELECTION, BY PARTY* Political Number of Party Candidate 8 Unionist 536 Labour 423 Liberal 454 ' Other 31 Totals 1444 *Liberal Year Book, 1924, p. 176. 116 TABLE 3.14 AGGREGATE AND PERCENTAGE RESULTS OF THE 1923 GENERAL ELECTION, BY PARTY* Political Aggregate Percentage Partj Results Results Unionist Votes 5,544,540 38.1 Seats 258 42.0 Labour Votes 4,508,504 31.0 Seats 191 31.1 Liberal Votes 4,314,202 29.6 Seats 158 25.7 Other Votes 196,789 1.4 Seats 8 1. 3 Totals Votes 14,564,035 100.0 Seats 615 100.0 *Liberal Year Book, 1924, p. 179. 117 Despite the Liberals' bid for power, the Labour Party also increased its percentage of the total vote between 1922 and 1923. The Unionist Party's proportion Of the poll declined only slightly, but its prOportion of seats fell sharply. The verdict of the electorate was against Protection, and as the. more successful Free Trade party, the Labour Party formed its first (minority) Government. 121 Thereafter the Liberal Party's fortunes fell (albeit not uniformly) until the Party had no reasonable hope of returning to power after 1931 . That history may be briefly recounted, augmented by the statistics in Tables 3.15 through 3. 22. The general election of 1924 was tragic for the Liberal Party: The number of Liberal candidates fell sharply to three hun- ,. dred forty-two, and the Party's prOportion of the total poll dropped to eighteen percent. 122 "On the other hand, though the (Labourite) 121For the results of the 1923 general election, see above Table 3.14. lzzFor the 1924 general election, see below Tables 3.15 and 3.16. "In 222 of the contested seats [. . .] there was unfortunately no Liberal candidate; and. in these constituencies there is no doubt that the Labour pOll was considerably swelled by the votes of Lib- erals who had no Liberal candidate to vote for [ . . .] But why did 80 many peOple, in the 300 constituencies where there was a choice, vote Conservative rather than Liberal? [Partly] because they per- ceived that the Liberals had not put enough candidates in the field to ensure the formation of an independent Government. " Liberal ~{gear Book, 1925, p. 175. (See also the Liberal Magdzine, 1924, 118 strength in the House of Commons had been reduced, their electoral strength in the country had increased by more than a million. It was plain that Labour would come again. "123 p. 634). "The situation [. . .] was attributed by Asquithians to Lloyd George's stinginess in allocating the rich resources of the Party Fund." Jones, Op. cit. , p. 211. "Lloyd George insisted that the reason for contesting so few seats in the Election had not been shortage of funds but of 'eligible candidates. "' Frank Owen, Tempestuous Journey, Lloyd George His Life and Times (New York: McGraw Hill Book Company, 1955), p. 685. l?-3Ibid., p. 684. 119 TABLE 3. 15 NUMBER OF CANDIDATES IN THE 1924 GENERAL ELECTION, BY PARTY* Political Number Of Party . Candidate 8 Unionist 534 Labour 516 Liberal 342 Other 29 Totals 1421 *Liberal Year Book, 1325, p. 176. TABLE 3.16 AGGREGATE AND PERCENTAGE RESULTS OF THE 1924 GENERAL ELECTION, BY PARTY* Political Aggregate Percentage Part_y . Results Results Unionist Votes 7, 838, 225 47.8 Seats 415 67. 5 Labour Votes 5,423,589 33.1 Seats 152 24.7 Liberal Votes 2,925,142 17.9 Seats 42 6. 8 Other Votes 197,673 1.2 Seats 6 TOtals Votes 16,384,629 100.0 Seats 615 100. 0 *Liberal Year Book, 1925, p. 178. 120 Despite the dispute between Asquith and Lloyd George, during the course of the General Strike in 1926,124 the Liberal Party's per- centage of the total vote in the contested by-elections of 1924-1929 increased to twenty- seven percent. 125 But in the general election Of 1929, although the Party ran 512 candidates, the Liberal prOpor- tion Of the total poll dropped to twenty-four percent. 126 Once again the Liberal Party held the balance of power in the House Of Commons, with a Labour Government in Office. The Liberal Party sought to have the Labour Government enact prOportional representation, '27 in exchange for support, but before this arrangement could be con- summated, the Labour Government broke up under the onslaught of the Great Depression. Prime Minister MacDonald formed a National Ministry, including Liberals and Unionists, and was repudiated by the Labour Party. 128 124This occasioned Asquith's resignation as Leader of the Liberal Party. See Earl Of Oxford and Asquith, op. cit. , II, Chap- ter XXIV. 125See below Table 3.17. 126For the 1929 General Election, see below Tables 3.18 and 3.19. 127After the 1918 general election, the Liberal Party was intensely interested in the enactment of prOportional representation. lZBMacDonald carried only 13 of 262 Labour M. P. s with him. Arthur Henderson became Leader of the Labour Party. 121 TABLE 3.17 TOTAL AGGREGATE AND PERCENTAGE RESULTS OF CONTESTED BY-ELEC TIONS IN 1924-1929, BY PARTY* Political Aggregate Percentage Party . - Results Results Unionist Votes 630,795 38.1 Seats 29 47 . 5 Labour Votes 567,925 34.3 Seats 22 36 . 1 Liberal Votes 447,834 27. 0 Seats 10 16. 4 Other Votes 10,101 0.6 Seats 0 0. 0 Totals Votes 1,656,655 100.0 Seats 61 100. 0 *Liberal Year Book, 1930, pp. 180—181. TABLE 3.18 NUMBER OF CANDIDATES IN THE 1929 GENERAL ELECTION, BY PARTY* Political Number of Party Candidate 3 Unionist 590 Labour 570 Liberal 51 2 Other 57 Total 1729 *leeral Year Book, 1930, p. 178. 122 TABLE 3.19 AGGREGATE AND PERCENTAGE RESULTS OF THE 1929 GENERAL ELECTION, BY PARTY* Political Aggregate Percentage Party Results Results Unionist Votes 8,664,243 38.3 Seats 260 42. 3 Labour Votes 8,379,978 37.0 Seats 288 46.8 Liberal Votes 5,301,127 23.4 Seats 59 9. 6 Other Votes 293, 949 1. 3 Seats 8 1. 3 Totals Votes 22,639,297 100.0 Seats 615 100. 0 *Liberal Year Book, 1930, p. 176. 123 TABLE 3. 20 TOTAL AGGREGATE AND PERCENTAGE RESULTS OF CONTESTED BY-ELEC TIONS 1N 1929-1931, BY PARTY* Political Aggregate Percentage Party Results Results Unionist Votes 470, 793 43.4 Seats 16 48. 5 Liberal Votes 113,293 10.4 Seats 0 0. 0 Labour Votes 430, 776 39. 7 Seats 16 48.5 Other Votes 69, 309 6.4 Seats 1 3.0 Totals Votes 1,084,171 100.0 Seats 33 100. 0 *Liberal Year Book, 1932, pp. 185-186. 124 When the National Ministe ry appealed to the country in the general election of 1931, the Liberal Party was split into three quite distinct segments: The Independent Liberal Party, virtually Lloyd George and his family, totally refused to support the National Government. The regular Liberal Party, under Sir Herbert Samuel, pledged qualified support to the Government. The Liberal National- ists, under the leadership of Sir John Simon, who had repudiated the Liberal Whip, 129 gave categorical support to the National Ministry. The general election of 1931 was a disaster for all parties, save for the Unionists and factions who supported the National Ministry. '30 The regular Liberal Party was reduced to a small shadow Of its former self; it has not since recovered. 12c("Sir John Simon, Sir Robert Hutchison, and Mr. Ernest Brown have informed the Chief Liberal Whip that they prOpose in future severally to act in independence of the Party, and that they therefore do not desire to receive the Liberal Whip. The letters in which the three Members stated their views and Sir Archibald Sinclair's replies were published in the Press on June 29th. " Liberal Magzine, 1931, p. 334. 130For the 1931 general election, see below Tables 3.21 and 3. 22. 125 TABLE 3. 21 NUMBER OF CANDIDATES IN THE 1931 GENERAL ELECTION, BY PARTY* Political Numbe r of Party Candidate 8 National Unionist 517 Libe ral 42 Labour 21 Libe ral 11 1 Labour 51 3 Independent Liberal 7 Other 75 Total 1286 *Liberal Year Book, 1932, p. 184 and Constitutional Year Book, 1932, pp. 183-246. (a) The numbers of "National Liberal, " "Liberal" and "Independent Liberal" candidates were obtained by a constituency-by-constituency count in the Constitutional Year Book, 1932; the sum total (160) was confirmed by the Liberal Year Book, 1932. (b) The Unionist Party's differential relationships with National Liberals, Liberals and Independent Liberals are indicated by the following statistics: Twelve percent (5/42) of the National Liberal candidates had a Unionist Party Opponent. Seventy percent (78/111) Of the Liberal candidates had a Unionist Party Opponent. Eighty-six percent (6/7) Of the Independent Liberal candidates had a Unionist Party Opponent . Computed from Constitutional Year Book, 1932, pp. 183-246. (c). The Liberal Nationalists are designated National Liberals in this and the following table, for convenience Of classification, but should not be confused with the National Liberals of 1922. 126 TABLE 3 . 22 AGGREGATE AND PERCENTAGE RESULTS OF THE 1931 GENERAL ELECTION, BY PARTY* Political Aggregate Percentage Party Results Results National Unionist Votes 11,907,870 55.1 Seats 470 76.4 Liberal Votes 809,102 3.7 Seats 35 5. 7 Labour Votes 335,704 1. Seats 13 2.1 Liberal Votes 1,405,759 6. 5 Seats 33 5.4 Labour Votes 6,618,316 30.6 Seats 52 8. 5 Independent Liberal Votes 106,106 0.5 Seats 4 0. 7 Other Votes 430,938 2. 0 Seats 8 l. 3 Totals Votes 21,613,795 100.0 Seats 615 100. 0 *Constitutional Year Book, 1932, pp. 181, 265. (a) The vote listed under "Other" includes: "[Independent Nationalists" with 62,820 votes. The eight seats listed under " ther" include: "Inde- pendent Nationalists" with 5 seats. Ibid. , pp. 181, 265. (b) The total vote for all types of Liberals, in wig—table, is 2, 320,967. The Liberal Year Book, 1932 gives the total vote for all types of Liberals as 2, 320, 310; this Is a difference Of only 657 votes. Ibid. , p. 184. 127 VI. Comparative Reflections on Major Party Schisms The parochial (purely British) case for the split thesis rests upon the account that has been given or upon a substantially similar account. In addition, since the break-up of the Coalition Party in 1922 and the upsurge of the Liberal Party in 1923 is suggestively parallel tO the split in the Liberal Party and the rise of the Labour Party in 1918, the preceding account has also provided a comparative instance, consistent with the split the sis. In the twentieth century, comparative examples can also be drawn from the two-party systems of Canada, Australia, and the United States. The Canadian Progressive Party, Australian Country Party, and American (Bull Moose) Progressive Party have been the most successful third parties, in their respective countries, in the twentieth century. 131 The upsurge of each of these parties--in 1921, 1919, and 1912, respectively--was preceded by a schism in one of the country's two major parties. Indeed, the antecedent splits in the Canadian Liberal Party, Australian Labor Party, and the American Republican Party were the most severe schisms that have occurred, in contem- porary major parties, in these nations. 132 A summary description 131Degree of success, in the present context, is méasured by a party's percentage of the total vote in a general election for the national legislature. 132For estimates of the magnitudes of these splits, see below Table 3.23. 128 of the culmination Of each of these schisms is given in the following paragraphs. The Rise of the Canadian Progressive Party. The Canadian Liberal Party split on the issue of conscription in 1917. 133 When voting on the Conservative Government's Military Service Bill, twenty- six Liberal M. P. 8, led by G. P. Graham, the leading Ontario Liberal, supported the Bill's compulsory features. 134 The Opposition to the Bill, led by Sir Wilfrid Laurier, the Leader of the Liberal Party, consisted mainly Of Liberal M. P. s from Quebec. "Laurier [. . .] with consummate skill had made the vote an open one. He hOped still to re-unite the party for the forthcoming election. "135 However, "a Union Government of Conservatives and dissident Liberals was 133’The connection between this schism and the rise of the Pro- gressive Party has been frequently noted by scholars. For general background, see W. L. Morton, The Progressive Party in Canada (Toronto: Toronto University Press, 1950), pp. 49-60; H. McD. Clokie, Canadian Government and Politics (Toronto: Longmans, Green 8: Company, 1944), pp. 79, 82, 84-85; Robert MacGregor Dawson, The Government Of Canada (Toronto: The University of Toronto Press, 1947), pp. 511-513. 134The number is drawn from J. Castell HOpkins, The Cana- dian Annual Review of Public Affairs, 1917 (Toronto: The Annual Review Publishing Company, Limited, 1918), p. 345. A total of eighty- seven Liberal M. P. s had been elected in the preceding gen- eral election. See Howard A. Scarrow, Canada Votes (New Orleans: The Hauser Press, 1962), p. 27. For Graham's vacillation in the subsequent general election, see HOpkins, op. cit., pp. 567, 619. 135Morton, Op. cit., p. 51. 129 formed. "136 The Union Government appealed to the country as a coalition and was victorious, sweeping the we stern provinces, in the general election Of 1917. "[ T]he fact that We stern Canada (normally Liberal) [shook] off its Old party allegiance over the conscription issue greatly lessened the difficulty Of making a change [. . . ] ."137 "A succession of provincial elections between 1919 and 1923 brought into office insurgent parties variously called United Farmers and Progressives in Ontario and the prairie provinces. "138 The national Progressive Party was formed in 1920 and, in the general election of 1921, captured twenty-five percent of the total vote and twenty- eight of the seats in the Dominion House of Commons. 139 From this pinnacle, the Progressive Party gradually declined. 136Clokie, Op. cit., p. 82. 137Dawson, op. cit., p. 511. Morton remarked that "the conscription issue of 1917 disrupted the national Liberal party and freed the agrarian we stern wing from the alliance with the conser- vative Liberals of Quebec." Morton, Op. cit. , p. 49. l38Clokie, op. cit., p. 84. 139William Paterson, "The Progressive Political Movement 1919--1930," (University of Toronto, M.A. thesis, 1940), Appendix B. For slightly (but not significantly) different figures, see Scarrow, 02. cit. , p. 35. The Progressive Party polled fewer votes, but won n'IOre seats, than the Conservative Party; the Liberal Party finished first in both votes and seats. 130 The_Rise of the Australian Countrj Party.140 The Australian Labo r Party split also on the issue of conscription, in 1916. When the L abor Party refused to countenance compulsory service, the Labo r Prime Minister, William Hughes, and thirteen Labor Repre- sentatives left the Labor Caucus. 141 Frank Tudor, the retiring Labor Minister for Trade and Customs, succeeded Hughes as Leader of the Labor Party. The ex-Laborites merged with the Liberal Party, forming a Nationalist Party, with Hughes as Prime Minister. The Nationalist Government was returned victorious, against Labor Opposition, in the general election of 1917. "From January 1918 on, a group of ex-Liberal Nationalists numbering at the most nine were sufficiently dissatisfied with the government to constitute the begin- nings of a corner party. "142 These ex-Liberals were allied with conservative agrarian organizations which had traditionally been Sympathetic to the Liberal Party but which found Hughes, with his urban brand of Nationalism, to be objectionable. 140The literature on the Country Party is not voluminous. For general background, see Louise Overacker, The Australian Party System (New Haven: Yale University Press, 1952). 141Just before the split Labor held forty seats in the House of Representatives. Both numbers are found in Geoffrey Sawer, wralian Federal Politics and Law, 1901-1929 (Victoria: Melbourne University Press, 1956), p. 130. 142Ibid., p. 130. 131 At the election [ . . .] held on 13 December 1919, [. . .] [t]he farmers' organizations in New South Wales, Victoria and Western Australia endorsed a number of candidates standing solely in their interest, and also gave a separate endorsement to many candidates who carried the endorsement of the Nationalist Party; from members so nominated emerged the Country Party. 143 After the 1919 general election, with a precarious parliamentary majority, Hughes' Nationalist Government continued in office until 1923. The Country Party held the balance of power after the 1922 general election, with thirteen percent of the electorate's first preferences and nineteen percent Of the seats in the House Of Repre— sentatives,144 and insisted that Hughes resign. A Nationalist-Country Government was formed, after Hughes' resignation, and the Country Party became "a regional and autonomous extension Of the main non- Labour Party. "145 The Ripe of the American (Bull Moose) Progressive Party. The American Republican Party split over its presidential nominee in 1912.146 When the Republican National Convention renominated 143Ibid., p. 183. 144Australia adOpted the preferential system Of election in 1918. The election statistics are drawn from L. C. Webb, "The Australian Party System,".in S. R. Davis, W. McMahon, A. A. Calwell and L. C. Webb, The Australian Political Party System (Sydney: Angus and Robertson, 1954), p. 97. 1451bid., p. 103. 146The connection between this split and the rise of the Progressive Party is so Obvious as to require little comment. For general background, see George Mowry, Theodore Roosevelt and the Progressive Movement (New York: Hill and Wang, 1960), Chapters 132 President William Howard Taft, Theodore Roosevelt, the Republican ex-President, and many of his four hundred fifty-seven delegates bolted the Republican Party and formed the Progressive (Bull Moose) Party. 147 With Roosevelt as its presidential nominee, the Progres- sive Party out-polled the Republican Party, but lost to the Demo- cratic Party, in the 1912 presidential election. However, in the 1912 congressional elections, the Progressive Party was less successful, polling somewhat less than twenty-one percent of the total vote and winning about four percent of the seats in the House of Representa- tives.148 From this peak, the PrOgressive Party rapidly declined. Although these five schisms differ in detail, 149 the broad pattern is the same: A major split in a major party was followed (or accompanied) by the upsurge of a third party. Nevertheless, the British case of 1918 is still unique in one salient respect: The third l‘47There were one thousand and seventy-eight accredited delegates. Both numbers can be found in Victor Rosewater, Back Stage in 1912 (Philadelphia: Dorrance 8: Company, Inc. , 1932), Chapter 16.WSee also Mowry, Op. cit. , pp. 252-253. 148These figures are probably more closely comparable than the presidential results, for institutional reasons, to the British, Canadian and Australian data. The figures are drawn from U. 5. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1957 (Washington: U. S. Government Printing Office, 1960). 149A rigorous comparison of these cases presupposes a systematic conceptual framework. For an experimental framework, possibly suitable for this purpose, see below Chapter IV. 133 party (Labour) surpassed an existing major party (Liberal).15O Since this feature is critical and unique, further comparative con- clusions must be somewhat tentative. There is, however, another unusual feature of the 1918 British case that should be noticed: The Liberal Party's broad leadership corps, when compared with the other cases, split most severely, the dissident faction was prOpor- tionally the largest. 151 This unusual magnitude, perhaps, accounts for the third party's unique success. f 150The measure, in the present context, is a party's per- centage of the total poll in a general election for the national legis- lature. See above footnotes 131, 139, and 148. 151See below Table 3.23. 134 TABLE 3 . 23 FIVE MAJOR SCHISMS, BY PARTY, MAGNITUDE, AND THIRD PARTY BENEFICIARY* Major Party and Magnitude Third Party Date of Schism Regulars Dissidents Beneficiary British Liberal Party, 1918 50-% 50+% Labour British Coalition Party, 1922 57 43 Liberal U.S. Republican Party, 1912 58 42 Progressive Australian Lab or Party, 1916 65 35 Country Canadian Liberal Party, 1917 70 30 Progressive *For the derivation of this table, see the following notes: (a) For the British Liberal Party in 1918, the regulars are the Asquithian M. P. s and the dissidents are the Lloyd Georgian M. P. s. The percentage division is estimated from Asquith's and Churchill's remarks, quoted above in connection with footnotes 3 and 76. (b) For the British Coalition Party in 1922, the regulars are the Coalition Unionist M. P. 8 who favored independence and the dissidents are the Coalition (87 lhionist. all Liberal) M. P. s favoring fusion. (c) For the U.S. Republican Party in 1912, the dissidents are Roosevelt's delegates. The percentage division is estimated from the information in above footnote 147 and related text. (d) For the Australian Labor Party in 1916, the dissidents are Hughes and the other thirteen La bO r Representatives who left the Labor Caucus. For the basis of the percentage division, see above footnote 141 and related text. (e) For the Canadian Liberal Party in 1917, the dissidents are the twenty-six Liberal M.P. s who Opposed Laurier on conscription. For the basis of the percentage estimates, see above footnote 134 and related text. 135 VII. Conclusions The split thesis argues, roughly speaking, that the British Liberal Party declined because "[t]he bulk of the Old Liberal parli— amentary party deserted to the Coalition. "152 This study has essayed to Show that the Split thesis fits not only the decline of the Liberal Party (and the upsurge of the Labour Party) but also the rise Of major third parties in comparable two-party systems. From a com- parative perspective, the schism in the British Liberal Party was unusually severe and the British Labour Party was unusually success- ful. The association of this pair of unique features, in the 1918 British case, seems hardly accidental. The validity Of the Split thesis has not been widely recognized by scholars, 153 perhaps because the magnitude of the 1918 debacle for the British Liberal Party has Often been under-estimated and because the existence of comparable cases has frequently been Over looked. Nevertheless, the split thesis is a viable explanation of the decline of the British Liberal Party, 152Earl Of Oxford and Asquith, op. cit., II, p. 284. 153Curiously enough, contemporary politicians saw the sit- uation with greater accuracy. This has been illustrated above, but for another illustration, see Snowden, Op. cit., I, pp. 319, 495. CHAPTER IV A THEORY OF HOMEOSTATIC TWO-PARTY SYSTEMS I. Introduction A description Of the decline Of the British Liberal Party could scarcely avoid scholarly controversy since 'decline' and 'British Liberal Party' have no conventionally fixed meanings.l And without a description Of the decline Of the British Liberal Party, a general explanation2 of the decline could hardly be proposed since the phenom- enon to be explained would not be specified. Since theory and descrip- tion are wedded by shared concepts, any systematic description of the decline of the British Liberal Party implicitly has theoretical content. The preceeding chapter has essayed a systematic description of the decline of the British Liberal Party, and the present chapter constructs a theory Of homeostatic two-party systems which is 1See above Chapters 1, II and 111 2By 'general explanation' the author means 'explanation by deductive subsumption under generalizations'. This is commonly equated with 'scientific explanation'. See Carl G. Hempel, "Deductive- NomOlOgical vs. Statistical Explanation, " in Herbert Feigl and Grover Maxwell (eds. ), Minnesota Studies in the Philospphy of Science (Minneapolis: University of Minnesota Press, 1962), III, pp. 98-169. 136 137 consistent with that description. 3 Theories about political parties can be constructed in many different ways and on many different conceptual bases. The present theory is not advanced as a complete theory of two-party systems; much less is it a complete theory of political parties; at most it is one way to elucidate the homeostatic character of stable two-party systems. 11. Conceptual Foundations The conceptual foundations for the theory are three concepts, viz. , 'two-party system', 'major party', and 'political party'. Since these concepts are fundamental, explicit definitions are prOposed after selective reviews of existing scholarly usage. Two-Party System. The distinctions between two—party and single-party systems and between two-party and multi-party systems are commonly judged to be fundamental4 but are not uniformly drawn 3From the present perspective, it can be seen that Chapter II represented a "clearing of the decks" for the constructions in Chapters III and IV. 4"The contrast between the multi-party and the single-party systems has become a commonplace in discussions [. . .] the con- trast between the two-party and the multi-party systems is of much less importance [. . .] however, it is undeniably a fundamental distinction." Maurice Duverger, Political Parties, trans. Barbara and Robert North (New York: John Wiley 8: Sons, Inc. , 1955), p. 206. "One-, Two-, and multi-party systems [. . .] the classic trichotomy of party classification." Charles W. Anderson, "Central American Political Parties: A Functional Approach, " Western Political Quar— terly, Vol. XV (1962), p. 127. 138 in scholarly analyses of party systems. A cursory survey of the literature shows that the Turkish party system has been dubbed a 'two-party system'5 and a 'multi-party system';6 the Japanese party system has been labeled a 'two-party system' and a one-and-a-half party system';7 the Australian party system has been termed a 'two- party system', 8 a 'duet in function and a trio in form9 and a 'three—party _ 5Duverger, Op. cit., pp. 211, 218; H. B. Sharabi, Govern- ments and Politics in the Middle East in the Twentieth Century (Princeton: D. Van Nostrand Company, Inc., 1962), pp. 54-61. Un- less otherwise indicated, in the present essay, all comments about Turkey refer to the period between the end of the Second World War and the 1960 Coup. 6Kemal H. Karpat, Turkey's Politics: The Transition to a Multi-party System (Princeton: Princeton University Press, 1959); Kemal H. Karpat, "The Turkish Elections Of 1957, " Western Political Quarterly, Vol. XIV (1961), pp. 436-459. 7"In formal terms perhaps, Japan has a two-party system. In broad political terms, that system is really a one-and-one-half party system. But in the most basic functional sense, it is a system of federations--a system of one dominant federation and one minority federation, with each having multiparties constantly in flux. " Robert A. Scalapino and Junnosuke Masumi, Parties and Politics in Contem— porary Japan (Berkeley and Los Angele 5: University of California Press, 1962), p. 81. "Japan may well be what Professor Robert Scalapino has called a one and one-half party state or what we call a six quarter-parties state. " Ardath W. Burks, The Government of Japan (New York: Thomas Y. Crowell CO. , 1961), p. 76. 8Gwendolen M. Carter, "The Commonwealth Overseas: Var- iations on a British Theme, " in Sigmund Neumann (ed. ), Modern Political Parties (University of Chicago Press, 1956), p. 58. 9Leslie Lipson, "Party Systems in the United Kingdom and the Older Commonwealth: Causes, Resemblances, and Variations," Political Studies, Vol. VII(1959), pp. 17, 27. 139 system'; 10 the British party system has been called a 'two-party system',11 a 'two-and-a-half party system '12 and a 'three-party system'.13 This variegated usage is paralleled by numerous defini- tions Of 'two-party System'. A few writers have had a very literal conception of the two- party system, viz.: "A two-party system is a party system with two and only two political parties. "14 Perhaps from a narrow linguistic perspective, this definition has the merit of simplicity, but from an empirical perspective, the definition is simultaneously too broad and too narrow. The definition is too broad because it encompasses some 10Duverger, op. cit.., p. 235. 11"As an example of the two-party system, the United Kingdom is probably most Often quoted. " Gunnar Heckscher, The Study of Comgarative Government and Politics (New York: The Macmillan Company, 1957), p. 153. 12Carl J. Friedrich, Constitutional Government and Democrag (rev. ed.; Boston: Ginn and Company, 1950). p. 414. 13The other citations refer rather generally to the British party system in the twentieth century, but this one refers to the period from 1918 to 1931. Leslie Lipson, "The Two-Party System in British Politics," American Political Science Review, Vol. XLVII, No. 2(June, 1953), p. 339. 14"With regard to the number of parties, there was a definite trend towards a two-party system, although attempts were made to introduce other parties." George Lenczowski, The Middle East in World Affairs (2nd ed.; Ithaca, New York: Cornell University Press, 1956), pp. 151-152. The reference is to Turkey and is related to above footnotes 5 and 6. 140 single-party systems. 15 A system with a single perpetually dominant party and with a single minor Opposition party, would be called a 'two-party system', Thus "Mexico's one-party system, "16 for ex- ample, would be classed as a two-party system. The definition is too narrow because it excludes some two-party systems. A system with two and only two major parties that alternate in power, but with transient minor parties, would not be called a 'two-party system', Thus "the two-party system in the United States,"17 for example, would not be classed as a two-party system. Several writers have adopted the view that the number Of major parties is the distinguishing feature of two-party systems, viz: "A two-party system is a party system with two and only two major parties. "18 This definition has the advantage of not grossly conflicting 15This type of statement is frequently made in the present essay and is an appeal to intuitive adequacy. Definitions, of course, may be insisted upon despite such appeals. See Carl G. Hempel, Fundamentals of Concept Formation in Empirical Science (Chicago: University of Chicago Press, 1952), pp. 11-12. 16L. Vincent Padgett, "Mexico's One-Party System: A Re- Evaluation, " American Political Science Review, Vol. LI (1957),pp. 955-975. 17William Goodman, The Two- Party System in the United §E§18 (2nd ed.; Princeton: D. Van Nostrand Co. , Inc. , 1960). 18"[T]he two-party system means that there are only two major parties [. . .]." E. E. Schattschneider, Party Government (New York: Rinehart 8: Company, Inc. , 1942), p. 68. "It is not always eaSy 150 make the distinction between two-party and multi-party syten'n 141 with ordinary scholarly usage but, perhaps, is also too narrow and yet too broad. The definition is rather narrow because it excludes party systems that usually have two and only two major parties but that occasionally have three major parties for brief periods. Thus the Canadian two-party system, for example, would not be classed as a two-party system in some years of the twentieth century. 19 This difficulty has been circumvented by a less stringent definition, viz.: "A two-party system is a party system that usually has two and only because there exist alongside the major parties a number of small groups." Duverger, Op. cit., p. 207. Duverger's remark is also suggestive about the inter- locking nature of the three concepts being discussed in the present essay. l9"A1though the Canadian two-party system has repeatedly broken down, it has never been replaced by a permanent multi-party system." Seymour Martin Lipset, "Party Systems and the Represen- tation of Social Groups, " European Journal of Sociofly, Vol. I, NO. 1 (1960), p. 23. Compare with the more usual View that Canada "has basically a two—party system." Carter, Op. cit., p. 58. "Whatever the theoretical justification of the two-party system Canada, ever since the emergence of national parties, has endorsed it consistently [ . . . ] . " Robert MacGregor Dawson, The Government of Canada (Toronto: University of Toronto Press, 1947), p. 497. But compare also with the reference to "[t]he former national multipartism of Canada," by Colin Leys, "Models, Theories, and the Theory of Political Parties," Political Studies, Vol. VII, No. 2 (1959), p. 145. A systematic resolution Of these Conflicting views, of course, pre- supposes an Operational definition Of 'major party'; a suitable definition is proposed below, but at this point, it may be observed that the Cana- dian dilemma also occurs for the United States during the heyday of the Progressive (Bull Moose) Party and for Great Britain during part of the 1920s. Furthermore, if Schattschneider's definition (see above footnote 18) is adopted, then two-party systems cannot be homeostatic but must be static with respect to the number of major political parties. 142 two major parties. "20 This revised definition is liberal in the desired fashion, at the cost Of increased vagueness, but perhaps is too broad. Both versions of the definition are rather broad because they encompass party systems that have two major parties in a perpetual coalition government, or one perpetually governing and one perpetually Opposi- tion party. Thus "the trend[s] toward a two-party system" in the postwar Austrian and We st German party systems, for example, would be prematurely classed as two-party systems. 21 The number (statistical or categorical) of political parties (major or minor) is insufficiert to denotezz the contemporary two-party 20"[A] two—party system[ ,] [t]hat is, there are usually two major parties [. . .]." Carter, Op. cit. , p. 58. See also Dawson, op. cit., PP. 496-497. 21For Austria, see Charles A. Gulick, "Austria's Socialists in the Trend toward a Two-Party System: An Interpretation of Post- war Elections, " Western Political Quartegy, Vol. XI (1958), pp. 539-562; Otto Kirchheimer, "The Waning of Opposition in Parliamen— tary Regimes," Social Research, Vol. XXIV (1957), p. 140. For West Germany, see Duverger, Op. cit. , p. 211; Arnold J. Heiden- heimer, The Governments Of Germany (New York: Thomas Y. Crowell Company, 1961), p. 62; W. Phillips Davison, "Trends in West German Public Opinion, " in Hans Speier and W. Phillips Davison (eds. ), West German Leadership and Foreign Policy (Evanston, Illinois: Row, Peterson and Company, 1957), p. 301. With respect to a stable majority-minority party situation, the postwar Japanese party system is somewhat similar to the postwar West German party system. See above footnote 7 and the sources cited therein. 22'In the language of philosophy, this essay is concerned with the denotation rather than the connotation of terms, that is, with the concepts' empirical referents rather than with the concepts' intrinsic meanings. 143 systems. Since the number of parties per se is inadequate to define two-party systems, the number of parties might be Viewed as an empirical regularity rather than a definitional characteristic, but this possibility has received scant attention in the literature. The number of parties has typically been augrrerted by additional definitional require- ments. A complex definition has been prOposed by Leslie Lipson and adOpted by some subsequent writers, 23 viz.: A state has a two-party system if it satisfies the following con- ditions: (1) Not more than two parties at any given time have a genuine chance to gain power. (2) One of these is able to win the requisite majority and stay in Office without help from a third party. (3) Over a number of decades two parties al- ternate in power. This definition has the virtue of excluding permanent coalitions as well as stable majority-minority systems but is too stringent. A system with a 250-249-1-1 distribution Of legislative seats, among two major and two minor parties, would not be called a 'two-party system', Thus the British party system, "the classic home of two- party government [, ] "25 for example, would often not be classed as 23For example, by Gulick, op. cit., p. 539n. For a criticism, see L. C. Webb, "The Australian Party System," in S. R. Davis, W. McMahon, A. A. Caldwell, and L. C. Webb, The Australian Political Party System (Sydney: Angus and Robertson, 1954), pp. 86-87. 24Lipson,"The Two-Party System in British Politics," p. 338. ZSIbid., p. 337. 144 a two-party system. 26 The first condition, a tacit restriction of a two-party system to two and only two major parties, is also trouble- some, as indicated above. These strict conditions might be relaxed with another definition, viz.: "A two-party system is a party system that usually has two and only two major parties and major parties usually alternate as majority parties in the legislature. "27 This definition is so vague as to be virtually impervious to empirical criticism and, as a result, although perhaps "true" in some sense, should not be accepted. Lipson's definition, although inadequate, is suggestive. A similar definition, prOposed and adopted below, stipulates that, rough- ly speaking, if a party system has satisfied Lipson's three conditions in sorre past period, then the party system has been a two-party system 26Lipson himself equivocated on this point. "Judged by these criteria, Britain may be said to have Operated the two-party system for two and a half centuries. [. . .] A three-party system existed in the latter decades of the nineteenth century [ . . .] and again in the 19208 [. . .] [b] ut the two-party system reasserted itself after each Of these seeming [s_ic_] deviations." liD_1_d. , pp. 338-339. (The problem also exists for classifying the party systems of Canada and the United States.) Lipson's remark about "two and a half centuries" also ob- scures the importance of the democratic prerequisites for a two- party system. These have been tacitly assumed in the present essay but are explicitly discussed below. 27See the sources cited above in footnote 20. In some Of the studies that have been cited, as in this case, the distinction between definition and description is not clearly or explicitly drawn. 145 since that period. Before giving a formal statement of the prOposed definition, it is helpful to distinguish between a two-party system and the environment of a two-party system. In this essay, it is taken as axiomatic that the environment of a two-party system is and must be democratic. The elements of democracy are somewhat difficult to exhaustively enumerate but in- clude bona fide competition for elective legislative Office, a broad and fair suffrage as well as a secret ballot. 2'8 Every two-party system Operates and can only Operate within such an environment. If the environment is disrupted by revolution or civil war, for example, then a two-party system is destroyed, if it exists at the time of the 28Even the conditions that have been mentioned are difficult to specify in detail. 'One Man, One Vote, One Value' and 'Universal Manhood Suffrage' have been rallying cries for the broad and fair suffrage requirement, for example, but these historic slogans are somewhat vague. [See James Mill, An Essay on Government (New York: The Liberal Arts Press, 1955), S ection VIII; Joseph Schumpeter, Capitalism, Socialism, and Democracy (3rd ed.; New York: Harper 8: Brothers Publishers, 1950), Chapters XXII-XXIII.] Furthermore, curiously enough, bona fide competition does not necessarily mean that more than one person is a candidate for a single office since, with some methods of voting, 'no candidate' should be treated as the name of a candidate. [See Duncan Black, The Theory of Committees and Elections (Cambridge: Cambridge University Press, 1963), pp. 3-4, 189.] With respect to the tolerable prOportions of uncontested seats, plural votes and unequal districts, as empirical approximations Of what is intended, the author Should remark that he considers Britain to have acquired a democratic environment in 1885 and to have had a democratic environment continuously since 1885. This is the conventional view. [See Henry Pelling, Modern Britain, 1885-1955 (Edinburgh: Thomas Nelson and Sons Ltd. , 1960), pp. 5-9.] 146 disruption, although it may be succeeded after the disruption by another two-party system. Every democratic environment, Of course, does not have a two-party system. A democracy might have a single- party or multi-party system and, indeed, a nO-party System is a theoretical possibility. 29 Given a democratic environment, a two-party system is defined by its formative cycle. In the formative period, there are two and only two major parties, the same pair for the entire formative period. For convenience, these parties are Party A and Party B and the general elections30 Of the formative period are Election 1, Election 2, etc. The two-party system's formative cycle is: In Election 1, Party A wins a majority (at least fifty percent) of the legislative seats and proceeds to organize the legislature. In Election 2 (since time- lags are permitted, this might be Election 3,etc.), Party B wins a 29"Dictatorial countries have no political parties in the original sense Of the word [. . . ] . " Herman Finer, Theory and Practice Of Modern Government (rev. ed.; New York: Henry Holt and Company, 1949), p. 302. 0Second chambers and executives are so diverse as to be troublesome referents in a comparative study. General elections for the so-called 'first chamber' of the national legislature are used in this study since there is some institutional comparability between, for example, a general election for the British House of Commons and a general election for the United States House of Representatives. 147 majority of the legislative seats and proceeds to organize the legis— lature. In Election 3 (with a permissible time-lag, this might be Election 4, etc. ), Party A wins a majority of the legislative seats and proceeds to organize the legislature. This is the formative cycle Of a two-party system. After the cycle is completed, a two-party system exists and continues to exist as long as the democratic environment is not disrupted. 3’1 Two concrete examples may be used to illustrate the definition Of 'two party system'. (1) The British party system has had a con- tinuously functioning democratic environment since 1885. In the gen- eral election of 1885, the Liberal Party won fifty percent of the seats and organized the House Of Commons. In the general election Of 1886, the Conservative Party (with its Liberal Unionist allies)32 won a large majority of seats and organized the House Of Commons. In the general election of 1892, the Liberal Party (with its Irish Nationalist allies) again won a majority of seats and organized the House Of Commons. Since the Liberal and Conservative 31With this definition, the usual number of major parties-- viz. , two--becomes an empirical regularity rather than a definitional characteristic. A major purpose of the present essay is to suggest the conditions under which three major parties will exist in a two- party system and to indicate which of the three major parties will decline (i.e. , cease to be a major party) as the system returns to its usual state. 32This example anticipates the definition of 'political party' that is given below, Viz. , certain types of coalitions are to be viewed as de facto single parties. 148 parties were the only major parties between 1885 and 1892, the British two-party system began in 1892. 33 (2) The Turkish party system had (let us say) a democratic environment after 1946. In the general election of 1946, the Republican PeOple's Party captured a majority Of the seats and continued tO organize the Grand National Assembly. In the general election Of 1950, the Democrat Party won a majority Of seats and organized the Grand National Assembly. Since the Re— publican People's and Democrat parties were the only major parties between 1946 and 1960, if the Republican PeOple's Party had again won a majority of seats and organized the Grand National Assembly, then a two-party system would have been established in Turkey. There was, however, a considerable fear that the Democrat Party had no intention of following the "democratic rules of the game" and permitting a loss of power in a general election. A military coup ousted the Democrat Party from power in 1960, disrupting the demo- cratic environment, and the postwar Turkish party system did not quite become a two-party system. 34 33The Canadian, Australian, New Zealand, United States and perhaps other. party systems are also two-party systems. 34For an illuminating discussion of the Turkish party system during these years, see Sharabi, op. cit. , Chapter 5. The Austrian, Japanese, West German and perhaps other party systems are also "almost but not quite" two-party systems. 149 This definition seems to extensionally correspond to the pre- dominate usage Of scholars in describing particular party systems as two-party or non-two-party systems. Furthermore, the definition provides (empirically, not definitionally) for important facets of the psychology of politicians and voters. First, the requirement of a full alternation in power, with each party winning and then losing power, conditions the politicians to accept the critical practice of relinquish- ing the perquisites of power.35 Second, the requirement of at least three general elections, with the same two (and only two) major parties, conditions the voters to expect a two-party battle and (prob- ably) inculcates habitual voting for the same party. 36 Finally, elec- tions have demonstrably "made a difference" for both politicians and voters, under the prOposed definition, since a complete alternation in power has occurred during the formative cycle of the two-party system.37 35The fear that the governing party would not yield power in accordance with the "democratic rules of the game" has frequently caused scholars to be reluctant to classify systems, with two and only two major parties, as two-party systems. See Duverger, Op. cit. , p. 214; Burks, op. cit., p. 76. 36For the importance Of habitual partisan voting, for the stability of two-party systems, see Angus Campbell, Philip E. Con- verse, Warren E. Miller, and Donald E. Stokes, The American Voter (New York: John Wiley 8: Sons, Inc. , 1960), p. 553. 37It is not the purpose of this essay to develOp a systematic classification of party systems, but the material in the text can Obviously (and, perhaps, usefully) be extended as follows: (1) It should be noted that the definition Of 'two-party system' denotes those 150 Major Parties and Political Parties. The definition Of 'two- party system' presupposed the concepts 'major party' and 'political party', but these concepts can scarcely be taken for granted since party systems that are usually regarded as stable two-party systems. A class of stable three—party, four-party, . . . , N-party systems can be generated by relaxing the definition of 'two-party system', The revised and general definition may be briefly illustrated by reference to a formation cycle for a three-party system: We assume a demo— cratic environment and a constant three major parties--A, B, and C-- during the formative period. After Election 1, Parties A and B form a parliamentary coalition to organize the legislature. After Election 2, parties A and C form a parliamentary coalition to organize the legislature. After Election 3, parties C and B form a parliamentary coalition to organize the legislature. After Election 4, parties A and B form a parliamentary coalition to organize the legislature. Then the three-party system is established. (This is a sketch; minor- ity governments are permissable; the key point is that each party has and then loses power during the formative cycle.) [The Swedish party system, often mentioned as an unusually stable multi-party system, apparently becomes a stable four-party system with this generalized definition. See Dankwart A. Rustow, The Politics of Compromise (Princeton: Princeton University Press, 1955), Appendix.] (2) From the foregoing, consistent with a common view, a stable one- party or no-party system is not possible, and a stable three-party system is less probable a priori than a stable two-party system because more power Shifts are necessary for the formative cycle. (3) In addition to the stable two-party, . . . , N-party systems, another class Of party systems--viz., nO-party, one-party, . . . , N-party systems--can be generated by defining them in terms of the number of major parties in the system (and, Of course, by excluding those systems that satisfy the generalized definition of a stable N- party system). (4) With this two-fold classification, the stable party systems can be viewed as absorbing states in a Markov Chain. Trans- ition probabilities, conditional upon the non- disruption Of the demo- cratic environment and using general elections as the unit Of time, could be estimated by an historical survey of election results. [For the mathematical aspects of Markov Chains, see John G. Kemeny and J. Laurie Snell, Finite Markov Chains (Princeton: D. Van Nostrand Company, Inc. , 1960.] 151 their variable content is almost proverbial. 38 A sample39 of the literature shows that the 1948 States Rights (Dixiecrat) Party of the United States was a party‘qfO and a non-party;41 the French National Center of Independents after 1954 was a major party‘}2 and a non- party;43 the British Labour Party was a party,44 a pressure group45 and "an appendage to the Liberal party after 1910;"46 and the British 38One man's party is another man's faction. "The study Of party politics has been remarkably confused by the poverty of the English language as far as the vocabulary of politics is concerned. Organizations called 'parties' at various times and in various places have in fact been fundamentally dissimilar, but all alike have been called parties for want of a sufficient variety of words corresponding to the diversity of realities." Schattschneider, Op. cit. , p. 65. "[E]ven the basic referents, such as the party, are at variance [. . .]." Carl J. Friedrich, "Review of Contemporary Political Science, " American Political Science Review, Vol, XLVII, No. 2(June, 1953), p. 540. 39Once again no claim is pretended to an exhaustive or repre- sentative sample of the literature. 40Wilfred E. Binkley, American Political Parties, Their Natural History (3rd ed.; New York: Alfred A. Knopf, 1958), p. 405. 41Anthony Downs, An Economic Theory of Democracy (New York: Harper and Brothers, Publishers, 1957), p. 128. 42Roy C. Macridis and Bernard E. Brown, The De Gaulle Re ublic (Homewood, Illinois: The Dorsey Press, Inc., 1960), pp. 17-18. 43Dorothy Pickles, The Fifth French Republic (rev. ed.; New York: Frederick A. Praeger, Publisher, 1962), p. 65. 44Downs, op. cit., pp. 128—129. 45Henry Pelling, A Short History of the Labour Party (London: Macmillan 8: Co. , Ltd. , 1961), ChapterIT. 46Sir Ivor Jennings, Party Politics (Cambridge: Cambridge 152 Liberal Party was a major party for the last time in 1918,"!7 in 1931,48 and in 1935.49 This confusion can be elucidated by examin- ing a common definition of 'political party' . Perhaps the most widely accepted definition of 'political party' stresses the electoral activity and presumed goals of certain organ- izations, viz.: "A party is an organization that contests elections and seeks to control the government. "50 This definition, in effect, abolishes the distinction between major and minor parties, by relegat- ing most minor parties to the category Of non-parties. 51 Even major University Press, 1961), II, p. 251. See also Hiram Miller Stout, British Government (New York: Oxford University Press, 1953), p. 165. 47Schattschneider, op. cit., pp. 68, 79-80. By inference; logically sound but perhaps not intended by Schattschneider. If such inferences are allowed, then there must be numerous "one-half" parties since so many authors refer to "two-and-a-half” party systems. 48$ee above footnote 26. By inference and consistent with Lipson's apparent intent. 49Duverger, op. cit., p. 208. By inference and consistent with Duverger's apparent intent. 50"A political party is a team of men seeking to control the governing apparatus by gaining Office in a duly constituted election. " Downs, Op. cit. , p. 25. "Political parties are autonomous organized groups that make nominations and contest elections in the hOpe of eventually gaining and exercising control Of the personnel and policies Of government. " Austin Ranney and Willmoore Kendall, Democracl and the American Party System (New York: Harcourt, Brace and Company, 1956), p. 85. See also Schattschneider, op. cit., pp. 35- 36. 51Schattschneider has been one of the few writers to reCOgnize (and accept) the definition's implication that most minor parties are 153 parties occasionally do not aim to control the government but rather seek to minimize a foreseeable electoral misfortune. 52 Some writers have preserved the distinction between major and minor parties by stipulating that a major party not only contests an election but also has a genuine chance to win the election, viz.: "At any given time, a party is an organization that contests an election and a major party is a party that has a genuine chance to gain power through the election. "53 But this definition is also restrictive since a major party, with Obvious prospects of obtaining thirty or forty not really parties: "Only when an organization is in control of the government or is able to create and maintain a widespread expecta- tion that it will take over the government soon does it become a major party or a real party." _Ib_ic_i_. , p. 36. Schattschneider strongly stresses this point. 52When directing the British Labour Party's campaign in 1931, the Leader of the Party, Arthur Henderson, frequently told his younger colleagues that winning a majority in the House of Commons would be a task for the next generation. [See Mary Agnes Hamilton, Arthur Henderson, A BiOgraphy (London: William Heinemann Ltd. , 1938), passim] It might be suggested that this difficulty could be avoided by postulating long-run as well as short-run goals, but this suggestion is virtually useless since very little is known about the goals (short-run or long-run) or, indeed, other characteristics of many minor parties. The definition, in fact, implicitly contains a terribly impoverished theory of human motivation, an almost Hobbesian theory of lusting after power. Even under excellent research condi- tions, the specification and measurement of an organization's goals are extremely difficult and involve contentious philosophical issues about the measurement of values. [See C. West Churchman, _I:r_e_- fltion and Optimal Decision (Englewood Cliffs, N. J.: Prentice- Hall, Inc. , 1961).] 53See above footnote 24 and quotation in the text. 154 percent Of the vote but with no Obvious prospects of obtaining a majority of seats, would not be called a 'major party'. Thus the Republican Party in the United States in 196454 and the British Labour Party in 1931, 55 for example, would not be classed as major part'ns.56 From an operational point of View, the preceding definitions of 'major party' and 'political party', if adOpted, are difficult to use. Chances (genuine or otherwise) are more difficult to ascertain than the results Of a general election. Seeking control of the government 54A somewhat less Obvious case occurred with the Democratic Party's landslide victory in 1936. The Republican Party polled forty percent of the total vote cast in the general election of the United States House Of Representatives in 1936. U. S. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1957 (Washington: U.S. Government Printing Office, 1960), series y 146- 149. 55The Labour Party polled thirty- one percent of the total vote cast in the general election Of the British House Of Commons in 1931. The Constitutional Year Book, 1932 (London: The National Union of Conservative and Unionist Associations, 1932), p. 265. 56Although the phrase "a genuine chance to gain power" is vague, the two examples seem to be plain cases. It might be sug- gested that this difficulty could be avoided by postulating long-run as well as short-run chances, but this suggestion is virtually useless without a predictive theory about long-run chances since, without such a theory, only hindsight could systematically ascertain whether or not a particular party was a major party at a particular time. For example, a party might be said to have (at time 1) a genuine long-run chance of gaining power, and hence might be said to be a major party (at time 1), if it actually gained power in an election within a sub- sequent period of twenty years (by time 1 + 20). This exemplary definition is systematic but is useful only for historians; retrodiction but not prediction would be possible with it. 155 is more Obscure than seeking legislative Office. These comments suggest the approach followed in this essay's definitions of 'major party' and 'political party'. A crucial difference between major and minor parties is that major parties do better than minor parties, in general elections. This difference is measurable either by the parties' percentages Of the total vote or by the parties' percentages of the total seats, but institutional variations complicate the task of selecting a particular percentage as the point Of demarcation between major and minor parties. Since this essay is concerned with two-party systems, the problem may be simplified by focusing upon simple-majority single- ballot electoral systems.57 Given a simple-majority single-ballot system of election, the parties' percentages of the total vote are the more suitable and stable quantitative measure to differentiate between major and minor parks."-B A definitionally stipulated percentage must be somewhat arbitrary, 5"(Duverger's comment (op. cit. , p. 217) that "[a]n almost complete correlation is observable between the simple-majority single-ballot system and the two-party system [ . . . ] " is well known. 58The parties' percentages of the total seats manifest great and rapid changes, without corresponding changes in the parties' per- centages of the total vote. See Enid Lakeman and James D. Lambert, Votingin Democracies (London: Faber and Faber, 1959), Chapter II. 156 but the present definition demarcates the Obviously major and minor parties and also provides a systematic and fairly precise decision rule for marginal cases: "If a party polls at least nineteen or twenty percent of the total vote in a general election, then the party is a major party through the next general election.)59 Although this definition is useful for the subsequent theory Of homeostasis, its general suitability may apprOpriate ly be investigated in other studies; perhaps, the definition will prove to be a convenient indicator for a variety Of characteristic differences between major and minor parties. Two concrete examples may be used to illustrate the definition of 'major party'. (1) Since the British Liberal Party polled only twelve percent of the total vote in the 1918 general election, the Lib- eral Party ceased to be a major party and became a minor party in 1918. 60 (2) Since the Canadian CO-operative Commonwealth Federation 59Several comments are apprOpriate. (1) The definition refers only to general elections for the legislature in countries using the simple-majority single-ballot system of election. (2) The percentage requirement cannot simply be transposed or translated to a system of preferential voting or prOportional representation since a theory of equivalents does not exist to effectuate the translation. The devel— Opment of such a theory--(for example, is 15% of the first preferences in a preferential system comparable to 20% of the votes in the simple- majority single-ballot system? )--is beyond the scope of the present study. (3) The 19 p_r_ 20 latitude in the definition reflects the approx- imate nature Of most election records and the fact that Britain does not but the United States does hold a poll for an uncontested seat. 60See above footnote 47 and related text. This example antici- pates the definition of 'political party' that is given below, viz. , cer- tain types Of coalitions are to be treated as single parties. 157 polled only sixteen percent of the total vote in the 1945 general election, the CO-Operative Commonwealth Federation remained a minor party in 1945.61 A minimal definition of 'political party' is adequate for the present study, viz.: "A party is a number of candidates who cooperate in a general election and who, if elected, (would) vote together in the legislature. "62 This definition should be interpreted in a broad fashion: Parliamentary-electoral coalitions should be viewed as a single party, even if nominally composed of more than one party. 63 Two concrete examples may be used to illustrate the prOposed definition of 'political party'. (1) In 1916-1918 the British Liberal 61See Carter, op. cit., p. 104. 62R seems probable that this definition would be acceptable as a minimal definition to most scholars, with the exception of those persons who equate major parties and political parties. A necessary (but, perhaps, not sufficient) condition for COOperation is a relative absence of rival candidacies in the general election. The notion of voting together in the legislature could also be stipulated in statistical terms as a certain minimal level of cohesion; see Stuart A. Rice, Quantitative Methods in Politics (New York: Alfred A. KnOpf, 1928), pp. 210-211. The temporal sequence implicit in the definition is intended by the author and is important for dating the origin Of any particular party. 63Conversel , more than one art (as defined above) ma be Y P Y Y lumped together under a single name in ordinary discourse. The definition focuses upon behavior, not conventional labels. 158 Party divided between the regular Liberal Party led by Asquith and the Coalition Liberals led by Lloyd George. The Coalition Liberals COOperated with the Unionist Party in the general election of 1918 and voted with the Unionists in the 1919-1922 House of Commons, and concurrently, Opposed the regular Liberal Party in the general election and in the House of Commons. Thus, by definition, the Coalition Liberals and the Unionist Party constituted a single party in 1918-1922, and the regular Liberal Party did not include the Coalition Liberals. 64 (2) During the Second World War, the British Conservative, Labour and Liberal parties formed a Coalition Govern- ment. Partisan electionneering was suspended; most by-elections were filled without a contest by the party that had held the seat. Nevertheless, since this COOperation did not extend to a general election, this Coalition should not be viewed as a single party. 65 III. The Theory The homeostatic theory of two-party systems is a comparative theory of the rise and fall of major parties in two-party systems. 66 64See above Chapter III. For additional examples, see above Chapter II. 65For another example, viz. , Liberal-Labour in 1906-1910, see above Chapter II. 66The existence of the two-party system is taken as a given in the present theory. A theory of the origins Of two-party systems is beyond the sCOpe of the present essay, but if the prOposed definition 159 The concrete parties for the comparative analysis are the Canadian Progressive Party, the Progressive (Bull Moose) Party of the United States, and the British Liberal and Labour Parties. 67 (A suggestive, but less compelling, comparison is also made with the so-called British Liberal Unionist Party. 68) These parties were major parties and their fortunes illuminate (although, perhaps, do not explain) the homeostatic character of two-party systems. The Rise of a New or Third Major Party. The analysis of the rise of new major partie s69 may conveniently proceed under three of 'two-party system' (with its emphasis on the formative cycle) is accepted, then the present study may contribute to the specification Of the phenomenon to be explained by a theory of the origins of two- party systems. 67The only other pertinent case, so far as this writer is aware, is the split in the Australian Labor Party and the rise of the Aus- tralian Country Party. (See above Chapter III.) The theory fits this case also, if the Country Party was a major party in 1922. [Some writers have apparently endorsed the view that the Country Party was a major party. See Geoffrey Sawer, Australian Federal Politics and Law 1901-1929 (Melbourne: Melbourne University Press, 1956), p. 183.] But since Australia adopted a preferential system of voting in 1918, the definition of 'major party' adopted in this essay does not apply and it seems advisable to avoid an ad hoc extension of the definition. Thus this case is cited as a possible but subtle conceptual- empirical test for the present theory. 68Since the Liberal Unionists and the Conservatives formed a parliamentary-electoral coalition immediately after the Liberal Union- ists left the Liberal Party, the Liberal Unionist Party does not meet the definition of 'political party' adOpted in the present study. Except for the early date (1886), this view is consistent with standard works. See A. Lawrence Lowell, The Government of England (New York: The Macmillan Company, 1908), II, pp. 16-17. 69The new major parties may be, and in some cases are, Old parties that are newly major. 160 broad headings, viz. , the conditions for the rise Of a new major party, the sources of those conditions, and the voter's reaction to those conditions. Two and only two major parties constitute the normal state of a two-party system. The existence of three major parties is a de- viation for the system. When the electoral histories of the contem- porary British, Canadian and American two-party systems are sur- veyed, 70 the infrequency and brevity of three-major-party71 deviatiors is striking. Britain had three major parties—-viz. , Liberal, Labour, Unionist--from the general election of 1923 through the general election of 1931. Canada had three major parties--viz. , Progressive, Conservative, Liberal--from the general election of 1921 through the general election of 1925. The United States had three major parties-— viz. , Bull Moose, 72 Republican, Democratic--from the general election of 1912 through the general election of 1914. These were the 70For election returns, see the Constitutional Year Book, 1932, pp. 257-265; Howard A. Scarrow, Canada Votes (New Orleans: The Hauser Press, 1962), passim; Historical Statistics of the United States, Colonial Times to 1957, series y 146-149. 71This locution is awkward, but the term 'third major party' is reserved for the major party that finished third in the total national vote. 7'ZThis nickname is used for convenience, to distinguish it from later Progressive parties in the United States and from the Canadian Progressive Party. 161 only three-major-party deviations in these two-party systems and were occasioned by the rise of the British Liberal, Canadian Pro- gressive and American Bull Moose parties to major party status in 1923, 1921, and 1912, respectively. In addition, the British Labour Party rose to major party status in 1918, but this rise was not accompanied by a subsequent three-major-party deviation since the British Liberal Party simultaneously declined (i.e. , ceased to be a major party). Scholars have frequently noted that a connection existed between the upsurge of each of these four parties and the preceding schisms in the 1922 British Coalition (Unionist-Coalition Liberal) Party, the 1917 Canadian Liberal Party, the 1912 American Repub- lican Party, and the 1918 British Liberal Party, respectively. 73 This correlation suggests the first hypothesis of the theory, viz.: In a two-party system, a severe split in an existing major party is a necessary condition for the rise of a new major party. The additional Observation that these were the most severe schisms to be experienced by major parties, in these two-party systems, suggests the second hypothesis of the theory, viz.: In a two-party system, a severe split in an existing major party is a sufficient condition for the rise Of a new major party. 73For citations and additional detail, see above Chapter III. 162 Although estimates of the magnitudes of these four schisms can only be approximations, the 1918 Split in the British Liberal Party seems to have been the most severe and the 1917 Split in the Canadian Liberal Party apparently was the least severe. 74 The mag— nitude of the schism in the Canadian Liberal Party, in fact, seems to have been near the lower bound for the class of severe Splits. A comparison, with the 1886 schism in the British Liberal Party, is suggestiveh75 (1) The British Liberal Party divided in 1886 over the issue of Home Rule for Ireland, when about twenty-eight percent of the Liberal M. P. s deserted the Leader of the Liberal Party and formed a coalition with the Conservative Party. 76 (2) The Canadian Liberal Party divided in 1917 over the issue of conscription, when about thirty percent of the Liberal M. P. s deserted the Leader of the Liberal Party and formed a coalition with the Conservative Party. 77 The comparative (prOportional) size of these two de serting factions 74See above Chapter 111, Table 3.23. 755ince Britain did not have a two-party system in 1886, as defined above, the comparison is not compelling. 76A detailed account of the development of this split is by J. L. Garvin, The Life of Joseph Chamberlain (London: Macmillan and CO. , Limited, 1933), 11, Chapters XXX-XXXIII. The roll call details may be found in Roy Jenkins, Sir Charles Dilke (London: Collins, 1959), p. 259. 77See above Chapter III. 163 is especially noteworthy since the former schism was not, but the latter schism was, followed by the appearance of a new major party. The 1886 split, in short, although important, was not of sufficient magnitude to be a severe split. These four severe schisms originated in party policy disputes that were inextricably involved with the personalities of the splitting parties' top leadership. The Official leader of the party (Lloyd George, Wilfrid Laurier, William Taft, Herbert Asquith) was Opposed, on a fundamental issue, by the man who was probably the second most important, if not the foremost, man in the party (Bonar Law, G. P. Graham, Theodore Roosevelt, Lloyd George). However, although perhaps prerequisites, these tOp-level divisions do not seem suf- ficient to account for the severity Of the splits. A comparison, with the 1931 split in the British Labour Party, is pertinent: The British Labour Party divided in 1931 Over the issue of a coalition ministry, when the Leader of the Party (J. R. MacDonald) formed a National Government without the support of the Secretary of the Party (Arthur Henderson), but only about five percent of the Labour M. P. s deserted with MacDonald from the Labour Party.‘78 This Split, of course, had 78For voluminous detail, see Reginald Bassett, Nineteen Thirty-one Political Crisis (London: Macmillan, 1958). 164 grave consequences but, although preceded by a tOp-level disagree- ment, was not large enough to be a severe split and was not followed by the appearance of a new major party. Thus a tOp leadership dispute may be a necessary but is not a sufficient condition for the occurrence of a severe split. The voters' reactions to severe splits are difficult to assess since sample surveys postdate the occurrence of these four schisms. However, the psychology of voting may, perhaps, be elucidated by reference to a relatively minor split, in the 1955 Australian Labor Party, that has been subjected to survey analysis.‘79 A strongly anti- Communist, largely Catholic, faction left (or was expelled from) the Labor Party in 1955 and formed the minor Democratic Labor Party. This split per se seems to have cost the Labor Party votes, 80 and more importantly, the split apparently activated the voting expressim Of previously latent economic-class differences among Labor Party supporters, leading some Labor Party voters to switch their votes.81 The broad result was a tendency to break down the traditional alle- gience of Labor Party voters. 82 These findings, even if valid for the 79See Robert R. Alford, Party and Society, The AlpglO-Amer- ican Democracies (Chicago: Rand McNally 8: Company, 1963), pp. 197-218. 801bid., p. 200. 811bid., p. 201. 82Ibid., p. 304. 165 1955 schism in the Australian Labor Party, are hardly decisive for the theory about severe splits. Nevertheless, the example is suggestive since it may plausibly be conjectured that the changes which have been noted would be accentuated by a more severe split. The tentative nature of these remarks about the magnitude and origins of, as well as the voters' reactions to, severe splits should not be allowed to Obscure the fundamental correlation that underlies the two hypotheses about the rise of new major parties in two-party systems: New major parties have arisen after (and only after) an existing major party has suffered a severe split, in the contemporary two-party systems of Britain, Canada and the United States. The Decline of a Major Party. The advent of a new major party does not destroy a two-party system; the system adjusts by the concurrent or subsequent decline of a major party. Judging from the preceding four cases,83 this homeostatic process conforms to the following hypothe sis: Those two (and only those two) parties survive, as major parties, that poll the largest number of votes in the general election which ushers in the new(ly) major party. 84 83Viz. , the advent, as a major party, of the British Liberal Party in 1923, the Canadian Progressive Party in 1921, the American Bull Moose Party in 1912, and the British Labour Party in 1918. 84The hypothesis fits all four cases; it should be noted that the hypothesis itself implicitly defines 'survive'. 166 The 1918 decline of the British Liberal Party was the only one of the four adjustments that was concurrent with the advent of a new major party; the remaining three adjustments --viz. , the 1931 decline of the British Liberal Party, the 1925 decline of the Canadian Progres- sive Party, and the 1914 decline of the American Bull Moose Party-- occurred subsequent to the advent of a new major party. If attention is restricted to the latter three cases, then the following regularity is evident: The new major party soon declined. 85 These three instances of three-major-party deviations merit closer examination since this comparative perspective permits a broader assessment of two leading theses about the decline of the British Liberal Party. From this viewpoint, the center party thesis that the British Liberal Party declined because it was a Center party along the classic Left-Right corfinuum, is not an adequate general explantion. Even prOponents of this explanation confess that the classic Left-Right continuum has little if any applicability tO the major parties of Canada and the United States. 86 Furthermore, if the continuum may reasonably 85This statement, of course, is not valid for the 1918 British case. 86Duverger, op. cit., pp. 418 (for the admission) and 214 (for the original explanation). For another version Of the explanation, see Downs, op. cit. , pp. 128—129. For a trenchant criticism of Downs' viewpoint, with especial reference to the United States, see Donald E. Stokes, "Spatial Models of Party Competition, " American 167 be applied, the Canadian Liberal Party would probably be judged a Center party between the Progressive Left and the Conservative Right?7 and the American Democratic Party might be judged a Center party between the Bull Moose Left and the Republican Right. 88 Thus, given a three-major party deviation, the Center party did not always de- cline as the two-party system returned to its normal state. 89 Political Science Review, Vol. LVII, No. 2 (June, 1963), pp. 368- 377. For the assertion that the "[p] Olitical parties in the Anglo- American countries, more than those in most others, fall along the classic Left-Right continuum" and the immediate qualification that "[i] deOIOgical characterizations are more difficult for the American parties (both United States and Canadian) [. . . ] ," see Alford, _op. _c__i_l_:. , pp. 11, 14n. See also the recent essay by Howard A. Scarrow, "Distinguishing Between Political Parties--the Case Of Canada, " Midwest Journal of Political Science, Vol. IX, NO. 1(February, 1965), pp. 61-76. For a critique, of the explanation's validity as an explana- tion of the 1918 decline of the British Liberal Party, see above Chapter II. 87See Alford, Op. cit., p. 13; Dawson, op. cit., p. 512. 88Professor Joseph A. Schlesinger has remarked, in conver- sations with the present writer, about the fact that the British Center (Liberal) Party has declined while the American Center (Democratic) Party has survived. Schlesinger has a comparative study of "center parties" in process. 891t should be noted that the definite and permanent location Of the British Liberal Party as the center party between the Labour and Conservative parties, is somewhat doubtful since Keynesian political economics were first introduced into the British political arena by the Liberal Party in 1929. [John Maynard Keynes was a Liberal and his policy prOposals antedated the theory presented in his General Theory of Employment, Interest and Money (New York: Harcourt Brace, 1936).] The Liberal Party was actually to the Left of the Labour Party in 1929, according to D. C. Somervell, British Politics Since 1900 (New York: Oxford University Press, 1950), p. 183. 168 From the same viewpoint, the electoral system thesis that the British Liberal Party declined because it was under—represented, as a third party, is also defective as a general explanation. The British Liberal Party and the American Bull Moose parties were under-re- presented"0 but the Canadian Progressive Party was not under-repre- sented, as third parties.91 Indeed, the Progressive Party was over- represented and the Conservative Party was under-represented,92 90In 1923 the British Liberal Party won thirty percent of the vote and twenty-six percent of the seats; in 1912 the American Bull Moose Party captured very roughly twenty-one percent of the vote and four percent of the seats. Constitutional Year Book, 1932, pp. 256, 262; Historical Statistics of the United States, Colonial Times to 1957 series y 139-141, 146-149. (This under-representation per as was hardly "dramatically Obvious" in the 1923 Liberal case. See Leys, op. cit., p. 143.) 91The Canadian Progressive Party in 1921 should not be dis- missed as merely a regional, sectional or inherently local party. The following figures, by province for the 1921 general election, indicate the Progressives were a bona fide national party: Prince Edward Island 4 seats and 3 PrOgressive candidates; Nova Scotia, 16 seats and 8 Progressive candidates; New Brunswick, 11 seats and 5 Pro- gressive candidates; Quebec, 65 seats and 18 Progressive candidates; Ontario, 82 seats and 71 Progressive candidates; Manitoba, 15 seats and 12 Progressive candidates; Saskatchewan, 16 seats and 16 Pro- gressive candidates; Alberta, 12 seats and 11 PrOgressive candidates; British Columbia, 13 seats and 5 Progressive candidates; Yukon, 1 seat and no Progressive candidates. In summary, of the 235 seats in the House of Commons, 149 were sought by Progressive candidates. Scarrow, Canada Votes, p. 36. For the conventional electoral system theory which permits regional parties to be exceptions, see Schatt- schneider, op. cit., p. 75; Duverger, Op. cit., p. 223; Leys, Op. cit., pp. 143-144, but compare with Webb, Op. cit., pp. 89-90. See also above Chapter II. 92The results of the 1921 Canadian general election were: 169 but the Progressive Party declined and the Conservative Party sur- vived. Thus, given a three-major-party deviation, the third major party was not always under—represented and the under-represented party did not always decline as the two-party system returned to its normal state. The center party and electoral system theses are not neces- sarily falsified by these cases since the counter-examples might be dismissed as irrelevant for a number of reasons.93 Nevertheless, the theories do not explain why the new and third major party declined in all three cases. That regularity is, however, consistent with (and, perhaps explained by) the broad findings of survey research into the psychology of voting. The typical voter, in a two-party system in its normal state, is psychologically attached to one Of the two major parties. This Liberal Party, 40. 7 percent Of the vote and 49.4 percent of the seats; Conservative Party, 30. 3 percent Of the vote and 21. 3 percent of the seats; Progressive Party, 23.1 percent of the vote and 27. 7 percent Of the seats; Other, 5.9 percent Of the vote and 1.7 percent Of the seats. Scarrow, Canada Votes, p. 35. 93For example, the claims might be advanced that (1) since the classic Left-Right continuum was weak (or non-existent) in Canada and the United States, the center party explanation simply does not apply to the decline of the Progressive and Bull Moose parties and that (2) since the under-representation hypothesis is statistical, the PrOgressive Party case does not falsify it. 170 allegiance to a major party is not only typical but also durable,94 often passed from generation-to-generation in a family95 and quite resistant to short-term change. The frequency and stability of these psycholOgical identifications are especially striking in a two-party system, by comparison with multi- party systems. 96 Party identification is accompanied, in a two-party system in a normal state, by a high frequency of habitual (or, at least, re- current) voting for the same major party. This is a serious Obstacle for a party that seeks to challenge the existing two major parties. "The conserving influence of party identification makes it extremely difficult for a third party to rise suddenly and with enough popular support to challenge the existing parties. "97 945cc Angus Campbell, Gerald Gurin, and Warren E. Miller, The Voter Decides (Evanston, Illinois: Row, Peterson and Company, 1954), p. 111. 95R. S. Milne and H. C. Mackenzie, Straight Fight (London: The Hansard Society, 1954), pp. 44-45; Paul F. Lazarsfeld, Bernard Berelson, and Hazel Gaudet, The PeOple's Choice (2nd ed.; New York: Columbia University Press, 1948), pp. 140-145. Although comparative survey data are lacking, there is no evidence that traditional loyalties are neither so frequent nor so strong in Canada as in Britain and the United States. See Scarrow, "Distinguishing Between Political Parties --the Case of Canada," pp. 72-76. 96$ee Philip E. Converse and George Dupeux, "Politicization of the Electorate in France and the United States, " Public Opinion Quarterly, Vol. XXVI (Spring, 1962), pp. 1-23; Angus Campbell and Henry Valen, "Party Identification in Norway and the United States, " Public Opinion Quarterly, Vol. XXV (Winter, 1961), p. 524. 97Campbell, Converse, Miller, and Stokes, op. cit., p. 553. 171 A newly major party must make a sustained effort for a con- tinuous series of general elections since its initial success, when it becomes a major party, is not sufficient to capture the traditional (habitual) loyalties of the voter. The 1923 British Liberal Party, the 1921 Canadian Progressive Party and the 1912 American Bull Moose Party failed to make such a sustained effort: After the initial success, in the next general election, the number of candidates fielded by these new major parties declined precipitously. 98 The established major parties, by contrast, continued to run a virtually full slate of can- didates, quite regularly.99 Thus, with these three-major-party deviations, the Opportunity was available for the reactivation of trad- itional allegiances but was not available for the formation Of new voting habits. The British Labour Party did make a sustained effort in the general elections of 1918, 1922 and subsequently. And the British Liberal Party, deserted by the bulk of the Liberal M. P. 8, failed to field a slate of candidates for many Of the seats in the general elections 93The statement is not valid for the British Labour Party, as a new major party, in the general elections Of1918 and 1922. See also the following footnote. 99The statement is not true of the British Liberal Party, when it ceased to be a major party, in the general elections of 1918 and 1922. See the subsequent discussion in the text. 172 of 1918 and 1922.100 Thus, in the 1918 British case, the electorate had an Opportunity to form new habits, without an Opportunity to rapidly re-affirm Old loyalties. The Liberal Party's failure to run full Slates of candidates, a direct result of the extreme severity of its 1918 schism, 101 was critical for the psychOlOgy of voting. These remarks about the psychOIOgy Of voting have been sug- t. gestive rather than definitive since survey research postdates the decline of the 1931 British Liberal Party, the 1925 Canadian Pro- gressive Party, the 1914 American Bull Moose Party and the 1918 British Liberal Party. The interpretation, nevertheless, is plausibly consistent with the recent surveys of voting behavior in two-p arty systems and indicates why the two largest parties102 survived the advent of a new major party. Perhaps a biosocial metaphor is an appropriate informal summary of this comparative analysis of the rise and decline of major 100For the number Of candidates, see above Chapter 111. 101See above Chapter III. 102Viz. , the two parties with the most votes in the general election that ushers in the new major party as a major party. Polit— ical calculations by aspiring politicians--calculating the Odds for a political career-~and tendencies towards social conformity by voters probably also affected the future availability of candidates and votes for the third major party, on the morrow Of that general election. 173 parties in two-party systems: If (and only if) a major party attempts suicide, some heir will appear; if the attempt succeeds, the heir will inherit. IV. Conclusions and a Formal Model of the Theory This homeostatic theory of two-party systems can be trans- lated into a formal model. The model selected for this purpose has been prOposed by Carl G. Hempel in his critique of the lOgic Of functional analysis: [T]he predictive significance of functional analysis is practi- c ally nil--except in those cases where suitable hypotheses Of self-regulation can be established. Such a hypothesis would be to the effect that within a specified range C of circumstanpes, a given system 8 (or: any system of a certain kind S, of which 8 is an instance) is self-regulating relative to a specified range R Of states; i. e. , after a disturbance which moves s into a state outside R, but which does not Shift the internal and external circumstances Of 8 out of the specified range C, the system 8 will return to a state in R. A system satisfying a hypothesis Of this kind might be called self— regulating with reflect to R. 103 The broad hypothesis of the present essay is that two-party systems are self— regulating with respect to the number of major parties. It is also posited that the following connections hold between Hempel's formal model and this comparative theory: (1) "C" is the democratic 103'Carl G- Hempel, "The Logic of Functional Analysis, " in Llewellyn Gross (ed. ), Symposium on Sociological Theory (Evanston, Illinois: Row, Peterson and Company, 1959), pp. 296-297. 174 environment of two-party systems. (2) "S" is the class of two-party systems. (3) "R" is the condition of having two and only two major parties. (4) A severe split in a major party is a "disturbance. " Furthermore, the theory permits the prediction of the particular major party (viz. , the one with the lowest vote in the general election that sees the advent of a newly major party) that will decline as the self-regulating two-party system returns to its normal state. Thus the theory is a modest contribution to "one of the most important tasks Of functional analysis in psychology and the social sciences [viz. ] to ascertain to what extent such phenomena Of self- regulation can be found, and clearly represented by laws of self- regulation, in these fields. "104 From another perspective, the theory shows that the so-called 'decline of the British Liberal Party' was a two-stage process pivoting about 1918 and 1931. This view sharply contrasts with the conventional picture of the decline as a one-stage (albeit lengthy) process. An implication of this contrast, perhaps, deserves a con- eluding emphasis: The theory not only simplifies but also complicates (conventional) reality. 104Ibid., p. 297. CHAPTER V THE LEADERSHIP OF THE BRITISH LIBERAL PARTY, 1868-1917 I. Introduction Previous chapters have argued that the conflict between Asquith and Lloyd George was of decisive importance for the history of the Liberal Party. This chapter attempts to illuminate and evaluate that leadership conflict by elucidating the pattern of leadership selection in the Liberal Party. II. The Selection Of Liberal Leaders, 1868-1908 Gladstone's succession to the Leadership1 of the Liberal Party is a natural starting point for this inquiry Since the formation of the Party is conventionally dated from his ascensim to power in 1868. 2 1This paper focuses upon three positions: The Leader of the Liberal Party. The Liberal leader in the House of Commons. The Liberal leader in the House Of Lords. A capital 'L' designates only the Leader of the Liberal Party. ZR. J. Cruikshank, The Liberal Party(London: Collins, 1948), p. 26; Hamilton Fyfe, The British Liberal Party (London: George Allen 8: Unwin Ltd. , 1928), p. 11; Albert Mabileau, Le Parti Liberal Dans Le gysteme Constitutionnel Britannique (Paris: Libraire Armand Colin, 1953), p. 11; W. E. Williams, The Rise Of Gladstone to the Leadership of the Liberal Party (Cambridge: Cambridge University Press, 1934), pp. 3, 4, 179. 175 176 Gladstone was the heir apparent for several years before he became Leader of the Party. After Lord Palmerston's death, Lord Russell became Prime Minister and Leader of the Party. Since Russell was a member Of the House of Lords, "Lord Granville was relieved of the immediate duties of leadership in the House Of Lords, [although] his real responsibilities were but slightly diminished. "3 Gladstone became the Liberal leader in the House of Commons, and William (later, Sir William) Harcourt: wrote to Lord Houghton that Gladstone's "position in the H[ ouse] of C[ ommons] will be so great and his succession so certain I don't see what more he could desire. "4 In 1867, when a Conservative Government was in office, Russell announced his intention not again to "take office." This 5 announcement, in letters to Gladstone and Granville, was interpreted as Russell's resignation as Leader of the Party. A biographer has stated that "he [Russell] formally reSIgned the leadership of the party 3Lord Edmond Fitzmaurice, The Life of Granville George Leveson Gower, Second Earl Granville (London: Longmans, Green, and Co., 1905), I, p. 497. 4Letter dated, from internal evidence, about October 20, 1865. Quoted in A. G. Gardiner, The Life of Sir William Harcourt (London: Constable 8: Company Ltd. , 1923), I, p. 169. 5Letters of Christmas, 1867 and January 2, 1868, respectively. See John Morley, The Life of William Ewart Gladstone (New York: The Macmillan Company, 1903), II, p. 243; Fitzmaurice, op. cit., I p. 518. 177 to Gladstone and Granville, "6 but in fact, the process was informal and somewhat confused. 7 Russell prOposed to continue to exercise some of the Leader's prerogatives, and while the general election campaign was in prOgress, there was some rather elaborate fencing about who would hold the Party dinner before the opening of the new session of Parliament. Although Lord Russell had abandoned all wish or expectation of again being Prime Minister, he was nevertheless prOposing to invite the Liberal peers to dinner at the commencement of the autumn session of Parliament, thereby apparently intimat- ing that he still considered himself technically to be the leader of the Liberal party, at least in the House of Lords; although Lord Granville had understood that the committal of the Sus- pensory Bill into his hands a few months before had been intended to mark the time and hour of Lord Russell's final retirement . 8 During the 1868 general election campaign, the Liberal can- didates ”adopted with a wonderful unaminity"9 Russell's declaration that Gladstone must undertake the task of forming the next Liberal 6A. Wyatt Tilby, Lord John Russell(London: Cassell 8: Company, Ltd., 1930), p. 252. Emphasis added. 7See the discrepant references to Gladstone's position in Dowager Duchess of Argyll (ed. ), Gage Douglas, Eighth Duke of Aflfll: Autobiography and Memoirs (New York: E. P. Dutton and Company, 1906), II, pp. 242, 346. 8Fitzmaurice, 02. cit., I, p. 530. 9Granville to Russell, November 21, 1868. Quoted in Ibid. , I. p. 533. Apparently, this was not a motion adOpted at a Party meet- ing but rather was Granville's personalassessment. 178 administration. "It is possible in these circumstances," Granville wrote to Russell, ”that he [Gladstone] would prefer having the little meeting in his own house, and that the Commons would like it better.‘10 After the Liberal victory in the general election, the Queen, as a matter of course, commissioned Gladstone to form a Government. The meeting of a ”select few"11 at Gladstone's house, a fortnight later, may be regarded as the Party's formal recognition of Gladstone's succession to the role of Leader of the Party, after becoming Prime Minister. 12 Gladstone's position as Leader of the Party was secure during the life of the "great administration, "13 and even after the electoral defeat of 1874, despite some discontented elements in the Party. 14 His colleagues were taken by surprise, at the last meeting of the Cabinet, IOIbid. 11Ibid. 12"At the end of April 1868, Mr. Gladstone [was] fully recog- nised leader of the party. " Bernard Holland, The Life of Spencer Compton, Eijhth Duke of Devonshire (London: Longmans, Green and Co., 1911), I, p. 70. Emphasis added. The conflict between Holland's view and Fitzmaurice's detailed account is further evidence of the informal neature of the leadership selection process. Fitzmaurice's account, in this writer's Opinion, is more accurate. 13Fyfe's description of Gladstone's first administration, 1868- 1874. Fyfe, op. cit., Chapter III. 14In particular, Russell and Harcourt were disgruntled with Gladstone. See Holland, 0p. cit., I, p. 132; Gardiner, op. cit., I, p. 286. 179 when there came "the startling announcement that Gladstone would no longer retain the leadership of the liberal party, nor resume it, unless the party had settled its differences. "15 Robert Lowe, a member of the Cabinet, immediately protested "against the anarchical experi- ment. "16 A final decision was postponed for some months, during which time there was an "intense and natural and righteous desire to see Gladstone back again."1'7 Nevertheless, Gladstone, who deemed himself "unable to hold it [the Party] together, "18 persisted in his intention. At the last minute, a meeting of the ex-Cabinet was held to bring pressure to bear upon Gladstone. When the effort failed, Granville expressed the fear "that the Liberal Party will fall to pieces in consequence of his [Gladstone's] decision. "19 Gladstone's view was that, upon his (Gladstone's) retirement, Granville would automatically become Leader of the Party as well as l5A Cabinet member's description. Quoted in Morley, Glad- stone, II, pp. 497-498. 161bid. 17Goschen to Granville, in early 1875. Quoted in Fitzmaurice, op. cit., II, pp. 138-139. 18A fragment, dated February 12, 1874. Quoted in Morley, Gladstone, II, pp. 498-499. l9oranviue to the Duke of Argyll, January 14, 1875. Quoted in Fitzmaurice, op. cit., II, p. 141. Gladstone's official resignation was dated January 13, 1875. 180 continuing as Liberal leader in the House of Lords. 20 This view was shared by Harcourt21 and Dr. Lyon Flayfair, 27- members of the "great administration. " Granville himself "took the Opposite view, and took it decidedly, that [ . . .] the question Of the leadership of the party would have [ . . .] to be adjourned to the day when [. . .] the choice Of the Sovereign created a Liberal Prime Minister. "23 Despite this disagreement, 24 a consensus develOped about the need to select a Liberal leader for the House Of Commons. 25 ZOMorley, Gladstone, 11, pp. 621-624; Agatha Ramm(ed.), The Political Correspondence Of Mr. Gladstone and Lord Granville, 1876-1886 (Oxford: Clarendon Press, 1962), I, p. 108. ZIHouand, o2. cit., I, p. 142. zzFitzmaurice, o2. cit., 11, p. 147. 23Ibid., 11. pp. 148-149. 24This disagreement was never Officially resolved, although the evolution Of the Party's customs favored Granville's view that only a Liberal Prime Minister or ex- Prime Minister could be Leader of the Liberal Party. J. A. Spender has stated that this was a "cherished part of Liberal theory" as early as 1896. J. A. Spender, The Life of the Right Hon. Sir Henry Campbell-Bannerman (London: Hodder and Stoughton, n.d. [c. 1923]), I, p. 212. However, Spender has also contended that this only "tided Over the emergency Of 1896 and became part of the unwritten constitution Of the party in subsequent years. " J. A. Spender, Sir Robert Hudson (London: Cassell and Company Ltd. , 1930), p. 53. For convenience of exposition, Granville's view is adopted throughout the present paper, but this does not affect the paper's conclusions. 25This consensus did not exist immediately after Gladstone's retirement. See Gladstone's letter to Granville, March 12, 1874. Quoted in Henry W. Lucy, The RiLht Honorable W. E. Gladstone (Boston: Roberts Brothers, 1895), pp. 80-81. See also Hartington's correspondence. Quoted in Holland, Op. cit., I, pp. 140-144. 181 John Bright, a member of the ex—Cabinet, recommended that "a leader should step into his place by general consent--he should be indicated by his own great qualities. "26 However, _Pu_ncli depicted five candidates: Lord Hartington, Sir William Harcourt, George Goschen, Robert Lowe, and W. E. Forster.27 There were, in fact, only two candidates since Harcourt, 28 Goschen, 29 and Lowe3o were supporting Hartington. Harcourt proposed that the ex-Cabinet should make a recommendation to the Party, 31 but some Of Forster's sup- porters protested, asserting "that they would resist any attempt to nominate a leader without consultation of the whole body Of the Liberal party in the House. "32 ZéBright to Granville, January 15, 1875. Quoted in Fitz- maurice, Op. cit., II, pp. 142-143. 27Gardiner, op. cit., I, p. 288. 281hid., I, p. 290. 29Hon. Arthur D. Elliott, The Life of George Joachim Goschen, First Viscount Goschen (London: Longmans, Green, and CO. , 1911), I, p. 157. 3°1hid., I. pp..156-157. 31(3rdiner, op. cit., I, p. 290; Holland, op. cit., I, p. 142. 32Playfair to Granville, January 15, 1875. Quoted in Fitz- maurice, Op. cit., II, pp. 145-146. 182 With Granville's approval, the Liberal Whip in the House Of Commons called a meeting of the Liberal members of the Commons for February 3, 1875, at the Reform Club. Granville himself played the key role in negotiating "general consent, " before the "whole body of the Liberal party in the House" was consulted at the Party meeting. Neither Hartington nor Forster actively sought the position; each was willing to withdraw rather than have an open contest for the leader— ship. 33 Forster's candidacy suffered from several handicaps. He was disliked by some Whigs because he had "bad manners and inferior education"34 and by some Radicals because he had sponsored the Education Act of 1870. 35 Most members of the ex-Cabinet also Opposed Forster and favored Hartington. 36 After consulting with Granville, Forster withdrew. 37 When Bright took the chair at the 33Ibid., 11, pp. 147—152; Holland, op. cit., I, pp. 144-146. T. Wemyss Reid, Life of the Right Honourable William Edward Forster (London: Chapman and Hall, Limited, 1888), II, p. 96. 34Lowe to Goschen, January 23, 1875. Quoted in Elliott, 2p. _c_i_t. , I, pp. 156-157. See also Henry W. Lucy, A Diary of Two Parliaments: The Disraeli Parliament 1874-1880 (2nd ed.; London: Cassell 8: Company, Limited, 1885), p. 113. 35Roy Jenkins, Sir Charles Dilke (London: Collins, 1958), p. 96; J. L. Garvin, The Life Of Joseph Chamberlain (London: Mac- millan and Co., Limited, 1932), I, p. 222. 36Fitzmaurice, Op. cit., II, p. 148; Morley, Gladstone, II, pp. 503-504. 37Reid, Op. cit., II, p. 96; Fitzmaurice, op. cit. , II, pp. 147—150. 183 Party meeting, Hartington was unOpposed and unanimously elected leader of the Party in the House of Commons. Hartington's diffidence about the leadership, apparently, was genuine. 38 His desultory attendance in the House of Commons, and his dislike of Party organization, did not change after his election as leader. When Joseph Chamberlain extended an invitation for Hartington to address the first annual conference of the National Liberal Feder- ation in 1878, Hartington refused on the grounds that "he did not like any Caucus; and saw that this one must strengthen the Radicals in the party. "39 When Chamberlain attempted in 1879 to have flogging prohibited in the army,a characteristic absence from the House of Commons caused Hartington to appear to criticize the effort. Chamberlain's response was caustic: The noble Lord had not, unfortunately, been in the House during a greater portion of the discussion--a thing which had been very much noticed on previous occasions. It was rather inconvenient that they should have so little of the presence of the noble Lord, lately the Leader of the Opposition, but now the Leader of a section only.40 38Lucy, A Diary, pp. 183-185; Holland, op. cit., I, p. 148. Since Gladstone and Granville contemplated a possible future come- back by Gladstone, perhaps Hartington's diffidence was consequential in their decisions to support Hartington for the vacant post. 39Garvin, op. cit., I, p. 267. See also Holland, op. cit., I, pp. 245-248. 40Hansard, 3rd Series, Vol. CCXLVII, 1806-1807 (July 7, 1879). 184 The quarrel subsided, when Hartington himself moved the abolition of corporal punishment, but "in [Sir Charles] Dilke's view substan- tially damaged Hartington's position and destroyed the possibility of his retaining the leadership. "41 ”42 The "row with Chamberlain and the Radicals was unpleasant, but the great disturbing factor for Hartington's leadership was the uncertainty about Gladstone's intentions. After Gladstone began his furious assault upon the Conservative Government‘s foreign policy, Hartington prOposed to resign, to officially recognize Gladstone's actual leadership. The fact is [Hartington wrote to Granville] that when he chooses to lead, he must be the leader of the party, and that since the autumn, and now, he has chosen to lead, and no amount of disclaimers will alter the fact, though they may in his Opinion relieve him of some of the responsibility which attaches naturally to leadership.43 Gladstone, however, wrote to Hartington saying that "he hoped nothing would be said [ . . .] about his being a leader, or he shall be obliged to be very explicit in an Opposite sense."44 41Jenkins, op. cit., pp. 119-120. 42Hartington's description of the 1879 clash. Quoted in Fitzmaurice, op. cit., II, p. 182. 43May 25, 1877. Quoted in Ibid., II, p. 172. 44Ibid. 185 Doubtless the difficulty is as you put it. [Forster wrote to Granville.] [If Hartington] makes himself a declaration, unless it be irrevocable, it would appear--and might be--use- less. But, if irrevocable, G[ladstone] might meet it with a refusal, and the party might be left leaderless in the Commons, which would be the worst possible result, and in fact fatal.45 During Gladstones' Midlothian campaign, as the prospects for a Liberal victory increased, the difficulty. became more acute. Hartington could not "conceive any arrangement [with Glad- stone holding a subordinate Office] lasting, or indeed any arrangement when Gladstone would be in the House and not leader. "46 It is clear [Hartington wrote to Granville] that a great majorly Of the party will not be satisfied unless Mr. G[ ladstone] is the next Liberal Premier and leader Of the House; and it is natural that it should be so. [. . .] [T] here is not room for argument about the proposition that the man who leads the Liberal party out Of doors ought to lead it in Parliament. The remarkable feeling which has been excited by his late speeches is to a great extent the expression of this conviction in the mind Of the party. If we are convinced, as I think that we must be, that Mr. G[ladstone] is the only possible Prime Minister, it seems to me that it is only fair to the Queen, to the country, to the party, and to myself that this should be acknowledged at once. [ . . .] [I]f he refuses to resume the nominal position which he now occupies virtually, the responsibility Of leaving the party without leadership does not rest with me, but with the man who has created the position. 47 45December 14, 1879. Quoted in Ibid., 11, p. 187. 46Hartington to Granville, October 28, 1879. Quoted in Ibid. , II, p. 182. 47December 7, 1879. Quoted in Ibid. , II, pp. 184-185. 186 By stressing the uncertainty about Gladstone's reaction, Granville dissuaded Hartington from resigning. Although Granville favored Gladstone's return as Prime Minister, Granville was aware that"[ u]p to now, we have done well; H[ artington] '3 official leadership keeping with us the moderates, and G[ ladstone] exciting the enthusiasts [ . . . ] ."48 Thus, during the Midlothian campaign, despite Gladstone's scrupulous avowals that Granville and Hartington were the Party's leaders, Gladstone's return as Prime Minister and Leader Of the Party was expected. After the Liberal Party was victorious in the general election Of 1880, the Queen asked Hartington, as a responsible leader of the victorious party, to form a Government. Hartington replied that a satisfactory Government could not be formed without Gladstone, that Gladstone would accept Office (if at all) only as Prime Minister, and that Her Majesty should call for Gladstone. The Queen did not immed— iately accept Hartington's advice49 but pressed him to ascertain 48Forster to Granville, December 9, 1879. Quoted in £11151 , II, pp. 186-187. "The question after all is this-~18 £9 [Gladstone] or not the real leader? If so, he ought to take the responsibility. " lbi_d. See also M. A. Fitzsimons, "Midlothian: The Triumph and Frustration of the British Liberal Party, " The Review of Politics, Vol. 22, NO. 2 (April, 1960), pp. 187-201. 49This is easily explained by the Queen's antipathy for Gladstone. "She will sooner abdicate than send for or have anything to do with that half-mad fire-brand who would soon ruin everything, 187 directly Gladstone's attitude. Hartington had an interview with Glad- stone, with the result that he (Hartington) had foreseen, and reported this to Her Majesty. The Queen then consulted with both Hartington and Granville, and after both strongly advised that Gladstone be called, Her Majesty reluctantly commissioned Gladstone to form a Government. 50 Gladstone remained Leader of the Liberal Party for fourteen years after the Queen's reluctant call in 1880. During that period he was undoubtedly the dominant figure in the Party, and except for the Home Rule crisis of 1886, his Leadership was not seriously challeng- ed.51 The Irish Question plagued Gladstone's second administration. Under the leadership Of Charles Parnell, M. P. , the different strands and be a Dictator. " Quoted in Philip Magnus, Gladstone, A Biography (New York: E. P. Dutton 8: Co., Inc., 1954), p. 270. 50Gladstone's lengthy memoranda, written at the time, are quoted in Morley, Gladstone, II, pp. 621-628. Sir Algernon West, apparently, is in error when he claims that Hartington attempted to form a Government. See Sir Algernon West, Recollections, 1832- 1886 (London: Smith, Elder, 8: Co., 1899). II, p. 103. 51.1. A. Spender's sweeping claim--"no Liberal would for a moment have disputed that Mr. Gladstone was the leader of the party from the day that he became Prime Minister for the second time in 1880 down to his final retirement in 1894"--is in error, individual exceptions aside, when applied to the Home Rule crisis. See Spender, Campbell-Bannerman, I, p. 212. 188 of Irish nationalist activity came to be joined. 52 The Irish M. P. s began to follow a policy Of systematic Obstruction in the House of Commons, and the conduct of public business was made possible only by a change in the parliamentary rules, permitting closure. 53 With Gladstone's knowledge, Chamberlain negotiated with Parnell, through an intermediary, 54 in an effort to establish an acceptable, conciliatory policy. Chamberlain's efforts resulted in a local government scheme designed to appease (if not satisfy) the Irish demand for Home Rule. The local government plan was rejected, however, by a majority of the Cabinet: "It was supported by all the Commoners except Hartington, and Opposed by all the Peers except Granville. "55 When the Liberal Government resigned, the Irish Question was still pending. During the general election of 1885, Liberal divisions were manifest. Chamberlain and the Radicals advocated a modest measure of Home Rule for Ireland and socio-economic reforms. Hartington 52See F. S. L. Lyons, The Irish Parliamentary Pa rty 1890- 1910 (London: Faber and Faber Ltd. , 1951), Chapter V. 53Morley, Gladstone, III, pp. 52-53; A. Lawrence Lowell, The Government of England (New York: The Macmillan Company, 1908), I, Chapter XV. 54The intermediary was Captain O'Shea, M.P. , the husband of Parnell's mistress. 55Joseph Chamberlain, A Political Memoir, 1880-92, edited by C. H. D. Howard (London: The Batchworth Press, 1953), p. 149. 189 and the Whigs resolutely Opposed both Home Rule and social reform. Gladstone was uncommonly silent, during the campaign, hoping to avoid an irrevocable split in the Liberal Party. The election results imposed a severe strain upon the parlia— mentary system. Without allies, the Conservative Government could not continue, and given the problems of perpetually marshalling a "' majority, even with the Irish as allies, the Conservatives could not carry on in the face of determined Liberal Opposition. The Liberals, conversely, could not form a Government without at least tacit support from either the Conservatives or the Irish. Unless the Conservatives and Liberals formed a temporary alliance, the Irish Nationalist M. P.s would, as the pivotal bloc, exercise great power in the new House of Commons. Gladstone suggested privately that the Conservative Govern- ment bring in a Home Rule measure with tacit Liberal support. This prOposal was rejected by the Conservative Cabinet, 56 chary of a severe split in the Conservative Party. 57 The Conservative 56Winston Spencer Churchill, Lord Randolph Churchill (New York: The Macmillan Company, 1906), II, p. 30. 57"We can have nothing to do with any advances towards the Home Rulers. The latter course would be contrary to our convictions and our pledges, and would be quite fatal to the cohesion of our party." Lord Salisbury to Lord Randolph Churchill, December 9, 1885. Quoted in 113151. , II, p. 14. 190 Government was speedily defeated by a coalition of Irish Nationalists and Liberals. Gladstone returned to office, committed to examine whether it is or is not practicable to comply with the desire widely prevalent in Ireland [ . . .] for the establish- ment by statute of a legislative body to sit in Dublin, and to deal with Irish a s distinguished from imperial affairs [. . .].58 Hartington refused to join a Government constituted on this basis, and with a few exceptions, the Whigs withdrew from the Liberal Party. Perhaps the split with Hartington's Whigs was inevitable, 59 but the split with Chamberlain's Radicals could have been avoided. Chamberlain and George Trevelyan resigned from the Cabinet, when Gladstone refused Chamberlain's prOposal for a continuation of Irish representation in the imperial Parliament. 60 A majority of the Cab- inet supported Gladstone's policy, and when Chamberlain carried the fight to the National Liberal Federation, a majority of the Federation's 61 General Committee also supported Gladstone's views. 58A memorandum which Gladstone read to each person whom he hoped to include in the Cabinet. Quoted in Morley, Gladstone, III, p. 292. 5C)Gladstone, however, was less fatalistic and considered several possibilities, including his own resignation. "I do not see how I could survive a gratuitous declaration of Opposition to me such as Hartington appears to meditate, " Gladstone wrote to Granville in the early months of1886. Quoted in _Ib_1d. , 111, pp. 282-283. 60Chamberlain, Op. cit., p. 212. Gladstone, it would seem, could have yielded the point to Chamberlain. For Gladstone's views, see Morley, Gladstone, III, p. 307. 61Spender, Hudson, pp. 10-12, 38-41. The headquarters of the Federation was soon moved to London, and under Chamberlain's in- fluence, some local units withdrew. 191 In the face of ever-growing embarrassments and importunities, recourse was had to the usual device of a meeting of the party at the foreign Office (May 27). The circular calling the meeting was addressed to those liberals who, while retaining full freedom on all particulars in the bill, were 'in favour of the establishment of a legislative body in Dublin for the manage- ment of affairs specifically and exclusively Irish. ' This was henceforth to be the test Of party membership. 62 Although "in the course of the evening [of the Party meeting] a score Of waverers were found to have been satisfied, ”63 Hartington's Whigs and Chamberlain's Radicals remained recalcitrant. Gladstone continued to refuse the continuation of Irish representation at London. A coalition of Chamberlain's Radicals and Hartington's Whigs, voting with the Conservatives, defeated the Home Rule Bill in the House of Commons. These Liberals were purged from the Liberal Party, but in the general election of 1886, hasty electoral arrange- ments that developed into a permanent coalition with the Conservative Party saved many of them from defeat. 64 After the purge, to use J. A. Spender's phrase, "no Liberal would for a moment have disputed that Mr. Gladstone was the leader Of the party [ . . .] down to his final retirement in 1894. "65 62Morley, Gladstone, III, p. 332-333. 631bid., III, p. 334. 64See Gordon L. Goodman, "The Liberal Unionist Party 1886- 1895, " (University Of Chicago, Ph.D. dissertation, 1956). 658ee above footnote 51. 192 Gladstone became Prime Minister for the last time in 1892 and resigned within two years. The official reasons for his retire- ment were that, at eighty-four years of age, his eyesight and hearing were failing. 66 These very real infirmities, however, were not the immediate stimulus for his retirement. Gladstone's relations with his Cabinets had deteriorated over the years, and in 1894 the Cabinet united against him on the issue of an increase in the naval estimates.67 Gladstone considered the prOposed increase to be "Mad--mad, "68 but when he addressed his colleagues for nearly an hour on January 9, 1894, "he convinced no-one, except the First Commissioner of Works, J. G. Shaw-Lefevre. "69 Both the Prime Minister and the Cabinet remained adamant. The Cabinet's telegraphic reSponse to Gladstone 's prOposal for a dissolution of Parliament was terse: "Your suggestion is impossible. "70 After Gladstone's return from a holiday in France, 66See Gladstone‘s formal letter to the Queen. Quoted in Morley, Gladstone, 111, pp. 514-515. 67Granville's death in 1891 is significant, in this respect, since Granville served as a buffer or mediator between Gladstone and his (Gladstone's) colleagues. Ramm, op. cit., I, pp. xxx-xxxii. Granville was succeeded, as Liberal leader in the House of Lords, by Lord Kimberley. Fitzmaurice, op. cit., II, p. 505. 68Horace G. Hutchinson (ed. ), Private Diaries of the Rt. Hon. Sir Algernon West (New York: E. P. Dutton and Company, 1922), p. 2.3. 6c)Magnus, op. cit., p. 417. 70Hutchinson, op. cit. , p. 271. 193 he resigned as Prime Minister, and this resignation was considered to be a resignation of the Leadership of the Party as well. ”Re signed! I did not resign--I was put out. "71 Gladstone later exclaimed. "I had lost power in my own Cabinet. "72 Lord Rosebery, the Foreign Secetary, and Sir William Harcourt, the Chancellor of the Exchequer, were widely regarded as the two leading possibilities to succeed Gladstone. 73 Both in age and service, Harcourt was senior. Furthermore, Harcourt was probably more acceptable to the National Liberal Federation, and there "was a growing feeling in the H[ ouse] of C[ ommons] against a peer. "74 Harcourt's fatal handicap was the Queen's and the Cabinet's, virtually unanimous, preference for Rosebery. 75 Sir Algernon West, the 71Quoted in J. A. Spender and Cyril Asquith, Life of Lord Oxford and Asquith (London: Hutchinson 8r CO. , 1932), I, p. 89. 72Quoted in John, Viscount Morley, Recollections (New York: The Macmillan Company, 1917), II, pp. 3-4. 73"Mr. Gladstone's Resignation," The Spectator, Vol. 72, (March 3, 1894), pp. 288-289. 74The Marquess Of Crewe, Lord Rosebery (London: John Murray, 1931), II, p. 441. See also Gardiner, Op. cit., II, p. 276. 75Morley, Recollections, H, p. 11; Earl of Oxford and Asquith, FiftLYears of British Parliament (Boston: Little, Brown and Company, 1926), I, Chapter XXXVI. Gladstone favored Lord Spencer; Harcourt favored Harcourt; the other Cabinet members either favored Rosebery or their preferences are unknown. 194 Cabinet's confidant, with the responsibility of informing the Prince Of Wales about Cabinet proceedings, reflected the dominant sentiment in the Cabinet: "[ S]end for Rosebery, " West advised Sir Henry Ponsonby, the Queen's secretary.76 Rosebery became Prime Minister and Leader of the Liberal Party. At Rosebery's request, Harcourt, who, as Chancellor of the Exchequer, had informally served as Gladstone's deputy leader in the House Of Commons, became the Liberal leader in the House of Com- mons. On March 12, 1894, the two leaders addressed a meeting of the Party's M. P. s at the foreign office.7‘7 The meeting went well, but consistent with the unofficial nature of the Leader's position, Rosebery was not formally elected at the meeting. He was simply tacitly accepted as the Leader of the Party.78 Relations between the new leaders were strained. 79 "M [were] no other difficulties," according to Sir Henry 76Hutchinson, Op. cit. , p. 285. For the leadership conflicts in the 1890's, see Peter Stansky, Ambitions and Strategies: The struggle for the leadership of the Liberal Party in the 18908 (Oxford: Clarendon Press, 1964). Stansky's volume, published after the present study was completed, contains a surfeit of detail. 77 Gardiner, Op. cit., II, pp. 275-277; Crewe, op. cit., II, p. 444. Such a meeting seems to have been an innovation. 78Rosebery was never formally elected Leader of the Party. See his letter to Asquith, dated January 29, 1896. Quoted in Spender and Asquith, op. cit., I, pp. 116-117. 79"It would hardly be human nature, and certainly not Sir 195 Campbell-Bannerman, the Minister for War. 80 Campbell-Banner- man's "view was very definitely that Sir William ought either to have retired or to have made up his mind to work amicably with the new Prime Minister.”81 Rosebery found "his personal position less and less tolerable," and in February, 1895, "called a Cabinet to announce that he must resign unless he were better supported in the House of Commons. "82 Harcourt deprecated Rosebery's resignation, and in Rosebery's Opinion, the device was successful. It would Of course not have been possible for me to resign; but it was the only way in which I could restore any discipline, or deal with the Open and in- sulting disloyalty Of one member of the Cabinet.83 After the Liberal defeat in the general election of 1895, Rosebery returned to London to reply to the Speech from the Throne, but Scotland, Paris, and Madrid captured him for the remainder of William Harcourt's nature to defer to Lord Rosebery as all his younger colleagues deferred to Mr. Gladstone. " See above footnote 73. 80Letter of February 12, 1895. Quoted in Spender, Campbel — Bannerman, I, pp. 165-166. Emphasis in the original. See also Earl Of Oxford and Asquith, Op. cit., I, pp. 251-255. 81Spender, Campbell-Bannerman, I, p. 165. 8ZCrewe, op. cit., II, p. 520. 83Rosebery's undated comment. Quoted in Ibid. , II, p. 520. 196 the year. 84 The Spectator soon noted that Rosebery's activities left the Liberal Party "virtually without a leader" and proposed that the leading Liberals in the country, presumably the National Liberal Federation, should get together and insist that a successor be chosen for Rosebery, but nothing Of the sort was done. 85 Despite Rosebery's behavior and Press criticism, his resignation of the Leadership was quite unexpected, although, as Asquith later observed, "the final resignation had been preceded by many premonitory signs, and ought not perhaps to have surprised some of us as much as it did. "86 Rosebery had made clear tO his colleagues, soon after the general election, that the situation in the recent Cabinet could not continue.87 When relations between Rosebery and Harcourt did not improve, and Harcourt continued as Liberal leader in the House of Commons, Rosebery resigned the Leadership Of the Party. The immediate stimulus and public excuse for the resignation was Gladstone's last public speech, disagreeing with Rosebery's policy toward Armenia. 84For the detailed itinerary, see Ibid. , II, pp. 511-516. 85"The Deposition of Lord Rosebery, " The Spectator, Vol. 77, (September 19, 1896), p. 359-360. - 86Earl of Oxford and Asquith, op. cit., II, p. 279. 87spender and Asquith, op. cit., I, pp. 116-117; Lucien Wolf, Efe of the First Marggess of Ripon (London: John Murray, 1921), II, pp. 244-246. 197 "I hope," Rosebery wrote to Gladstone, "that my retirement may at any rate produce some greater amount of unity in the distracted and honeycombed party called 'Liberal. "'88 The Daily News called for the re-election Of Rosebery, but as The Spectator Observed, "the force of circumstances [was] too strong for that. "89 The Spectator predicted that Harcourt's "pre-eminent position in the House of Commons makes it certain that he will succeed to the leadership of the party. "90 That possibility was feared by several members of the ex-Cabinet, but Harcourt rejected a prOposal that he be endorsed by a resolution of the Liberal M. P. s in the House Of Commons.91 "The idiots Of the Press," Harcourt wrote to John Morley, a member of the ex-Cabinet, "seem to think every one is ready to cut one another's throat in order to become 'Leader of the Liberal Party. ' For my part, if I did not think it currish to bolt in the presence of difficulties, I should take up my hat 88Quoted in Crewe, op. cit., II, p. 527. 89"Lord Rosebery's Resignation," The Spectator, Vol. 77 (October 10, 1896), p. 472. 901bid., p. 472. 91For the fears, see Wolf, Op. cit., II, p. 248. For Har- court's action, see Gardiner, Op. cit., II, p. 419. 198 and say good-bye. "92 Harcourt's view seems to have been that the position of Leader of the Party was vacant and that he (Harcourt) simply continued as the Liberal leader in the House Of Commons. In any event, this view was adOpted by Asquith, Sir Henry Fowler, James Bryce, and Lords Ripon, Spencer, and Kimberley, (members Of the ex-Cabinet), as well as by the National Liberal Federation.93 When Rosebery also resigned the Leadership of the Liberal Peers in November, 1896, the Liberal Peers proceeded to re-elect Kimberley as Liberal leader in the House Of Lords. This was done in a "classic" manner: Agreement on a candidate was reached by the Peers who had been members of the ex-Cabinet, other possible can- didates withdrew, and the pre-selected candidate (Kimberley) was unanimously elected at a meeting of the Liberal Peers.94 Thus follow- ing Rosebery's resignations, the Liberals were led by a duumvirate, paralleling the situation in 1875-1880. Personal relations between Harcourt and Kimberley were harmonious, 95 but this did little to quiet the dissensions in the Party. 92Quoted in Ibid., 11, p. 422. 93WOlf, op. cit., II, pp. 248-249; Spender, Hudson, p. 53. 94WOlf, op. cit., II, pp. 248-249; Crewe, op. cit., II, p. 529. 95WOlf, op. cit., II, p. 251. 199 As policy differences increased between the Liberal Imperialists and the Little Englanders, some members of the Imperialist wing fell into the habit of reminding him [Harcourt] that Lord Rosebery's resignation did not settle [the question of a future Liberal Premiership] in his favour or raise him from being leader in the House of Commons to being leader of the Liberal Party.96 Harcourt contemplated resignation for some months before taking the decisive step in December, 1898. The resignation took the form of a public letter to Morley who associated himself with Harcourt's withdrawal. "A party rent by sectional disputes and personal interests is one which no man can consent to lead either with credit to himself or advantage to the country, " Harcourt noted, "I am not, and I shall not consent, to be a candidate for any contested position. "97 The general committee of the National Liberal Federation met a few days after Harcourt's resignation was published. Inevitably, the leadership was discussed at the meeting, and two motions were made on the subject. The first requested Harcourt to reconsider, and the second stated "that, in the Opinion of this meeting, the question 96Spender, Cameell-Bannerman, I, p. 212. Asquith, a Lib- eral Imperialist, denied that Rosebery's friends were intriguing against Harcourt's leadership. See Spender and Asquith, op. cit., I, pp. 119-122. 97The letter is quoted in full in Gardiner, Op. cit. , II, pp. pp. 472-474. 200 Of the leadership Of the Liberal party should be taken into immediate consideration, and calls upon the leaders to close up their ranks. " These motions were withdrawn in deference to a strong feeling that they did not come within the functions of the Federation.98 Asquith, who addressed a mass meeting following the session of the general committee, believed "that the National Federation, which had met the re in the afternoon, had conducted itself with great discretion. "99 Selection Of a new Liberal leader in the House Of Commons was left to the Liberal members of the Commons. The ex-Cabinet members agreed upon a candidate, and their choice was unanimously elected at a meeting Of the M.P. 8. After Harcourt's and Morley's withdrawals, only four members of the late Cabinet were still active in the House" Of Commons: Asquith, Campbell-Bannerman, Fowler, and Bryce. Asquith was asked to undertake the leadership but demurred on per- 100 sonal grounds, namely, his need to continue practice at the Bar. After Asquith's decision, the only difficulty was to persuade 93Lowell, op. cit., I, pp. 546-547. 99Quoted in Spender and Asquith, Op. cit. , I, p. 121. Har- court's view was that the position Of Prime Minister and Leader Of the Party, "when the time came, would only be filled by the Party in Parliament." Gardiner, Op. cit., II, p. 469. 100See Asquith's memorandum Of December, 1898. Quoted in Spender and Asquith, Op. cit., I, pp. 119-122. 201 Campbell-Bannerman to accept. The latter, who remained in Scotland during the negotiations, was soon deluged with letters requesting him to accept the post. "I still hOpe I shall be mate and not captain,"101 he wrote to friends, "but if I receive what in kirk sessions we style a 'call,‘ I am son enough of my country to do my best. "102 After receiving assurances of support from Asquith, Fowler, Bryce, and other leading Liberals, Campbell-Bannerman accepted the "call" and was unanimously elected ata meeting Of the Liberal members Of the House of Commons. '03 The unanimous election of Campbell- Bannerman did not remove the serious divisions Over foreign policy, and reoccurring rumors stated that the leader would resign because of these divisions. All these rumors were baseless. Quite early in the day he had written Off resignation, or even the threat of it, as an available weapon for a man in his position. The fact that his two predecessors had resigned would, he felt, have made it ridiculous for him to seek the same way Of escape. 101Letter to Sir Ralph Knox, dated January 2, 1899. Quoted in Spender, Campbell-Bannerman, I, p. 216. 102Letter to Rosebery, dated January 6, 1899. Quoted in Ibid., 1, pp. 217-218. 103Extensive detail is provided in Ibid., I, pp. 214-220. 104Ibid., 1, p. 287. 202 These divisions become so serious in 1901, however, that Campbell— Bannerman called a Party meeting to confront the question of Party unity. 105 In the course of his speech to the meeting of Liberal members of the House of Commons, Campbell- Bannerman remarked that "this evil [of disunity] can be put down by one force and by one force alone--by the general sense Of the party. "106 A resolution Of confidence in the leader was passed unanimously, supported by Asquith and Sir Edward Grey, the leading Liberal Imperialists in the House of Commons. 107 The leadership of the entire Party remained open after the re-affirmation of Campbell-Bannerman's leadership in the House Of Commons. Kimberley continued to lead the Liberal Peers, and when 105The division was aggravated by a series of dinner speeches in 1901 . "The 'dining history' of the Liberal Party in these weeks was thus summed up by the We stmin ster Gazette: 'There was a dinner on June 14 at which speeches were made which gave great offense to the Imperialist section of the Liberal Party. There was a dinner on June 20 at which Mr. Asquith answered the speeches which gave offense. There is now to be a dinner in recognition of the speech which answered the speeches which gave Offense to the Liberal Imperialists. There will next be a dinner in recognition of the speech which gave the Offense which was answered by the speech which led to the dinner in reCOgnition. The Liberal Party will thus dine and counter-dine itself out of existence or else be dissolved in the laughter of that observant man in the street or balancing elector whose suffrages it so greatly desires to Obtain."_I_l_)_i_<_1., I, pp. 339- 340n. 106A part of the speech is quoted in Ibid. , I, pp. 344-345. lo7Ihid., I, p. 345; Spender and Asquith, op. cit., I, pp. 140. 141. 203 he became ill, Spencer acted for him. Upon Kimberley's death in 1902, "Spencer was elected to succeed him on the prOposal of Ripcn.‘108 Speculation existed that Rosebery might attempt to resume the leader- ship. When Chamberlain launched his tariff campaign in 1903, Ripon wrote to Spencer, "let him [Rosebery] throw himself into this battle, and with his eloquence and his talents he may come out of it our un- questioned leader. "109 Rosebery, however, continued to "plow his own furrow, " with no intention of again taking Office. 110 His final separation from the Party's leadership came in 1905, when Campbell- Bannerman, after consulting with the Liberal Shadow Cabinet, enun- ciated the Liberal policy on Home Rule. Leading Liberal Imperialists, such as Asquith, did not support Rosebery's criticism Of the policy since Campbell-Bannerman had consulted them about the policy.111 Rosebery, in effect, removed himself as a possible future Liberal Minister, by his criticisms. 103WOlf, Op. cit., II, p. 268 n. 109In a letter Of May 30, 1903. Quoted in Ibid., II, pp. 270- 271. 110The phrase is Rosebery's. See his personal memoranda of September 30, 1903, October 14, 1904, and May 2, 1905. Quoted in Crewe, op. cit., II, pp. 585-591. lllspender and Asquith, op. cit., I, p. 169. O 204 Harcourt had died in 1903. Since Spencer was ill, Campbell- Bannerman became Prime Minister and Leader of the Party as a matter of course when the Unionist Cabinet resigned in 1905. Some Liberals had anticipated, and feared, this step because Campbell- Bannerman had not been an impressive Leader of the Opposition in the House of Commons. A concrete manifestation of these apprehen- sions was the "Relugas Compact, " among Asquith, Grey, and Richard Haldane. The substance of the Compact was that, if Campbell- Bannerman became Prime Minister, "he should take a peerage, and Asquith should lead in the Commons as Chancellor of the Exchequer."112 When Carrpbe ll-Bannerman refused to' take a peerage, Asquith immediately accepted Office as Chancellor of the Exchequer, arguing that the cir- cumstances envisaged by the Compact did not exist.113 At Asquith's and Campbell- Bannerman's urging, both Haldane and Grey also entered the Cabinet on Campbell-Bannerman's terms. 114 The fears that had given rise to the Compact proved to be unfounded: "There 112Richard Burdon Haldane, An Autobiography (London: Hodder and Stoughton Limited, 1929). p. 159. 113See Spender and Asquith, op. cit. , I, pp. 174-175. 114Grey was the most reluctant to enter the Cabinet. See George Macaulay Trevelyan, Grey Of Fallodon (Boston: Houghton Mifflin Company, 1937), pp. 108-115. 205 never was in fact a more miraculous change in the 'form' of a public man than from Campbell-Bannerman as leader of Opposition to Campbell-Bannerman as Prime Minister. "115 Campbell- Bannerman was not in good health when he took office, and in February, 1908 became seriously ill. Asquith presided over the Cabinet in the Prime Minister's absence. In March the King left for France but requested Campbell-Bannerman not to resign before his return. 116 In deference to the King's wish, the Cabinet continued to meet without the Prime Minister, but before the King's return, Campbell-Bannerman's health necessitated his retirement. The King had already decided to send for Asquith, when Campbell- Bannerman eventually retired, before the trip to France. 117 After Campbell-Bannerman's retirement, Asquith became Prime Minister as a matter of course. No other possibility was seriously considered. After Asquith became Prime Minister, a unanimous resolution of con- fidence and support was passed at a meeting of the Party's M. P. 8.118 115Spender, Campbell-Bannerman, II, p. 404. 116Ibid.. 11, p. 383. 11"(See Spender and Asquith, op. cit. . I. P- 195- 118The Times, April 30, 1908, p. 8. This seems to have an innovation. 206 Thus, Asquith succeeded Campbell- Bannerman, ”subject to your ratification, " he told a later Party meeting, "in the leadership of the Liberal Party. "119 III. The 1916-1918 Crisis as a Crisis in the Liberal Leadership Selection Process During the tumultous partisan conflicts over LlOyd George's "PeOple's Budget," House of Lords Reform and Home Rule, Asquith's position as leader of the Liberal Party was secure. Except for the resignations of Morley and John Burns, a united Liberal Cabinet entered the First World War. Despite military reversals, Asquith's position was still so paramount in March 1915 that he was able to quell a quarrel between LlOyd George and Reginald McKenna, by threatening to resign. 120 Less than two years later, Asquith was forced to re sign. 119The Liberal Magazine, 1916 (London: The Liberal Pub- lication Department, 1917), p. 621. 120When McKenna succeeded Lloyd George as Chancellor of the Exchequer and Lloyd George became Minister of Munitions, upon the formation of the First Coalition, relations between the two men were further strained because Lloyd George retained the option to return to the Treasury after the munitions crisis was over. See Frank Owen, Tempestuous Journey: LlOyd Gegge His Life and Times (New York: McGraw-Hill Book Company, Inc., 1955), p. 298. Thomas Jones believed Lloyd George's physical separation from Asquith, and McKenna's physical proximity to Asquith, after the change of offices, contributed to the deterioration in the relations between Asquith and Lloyd George. See Thomas Jones, Lloyd Georg_e (London: Oxford University Press, 1951), p. 61. 207 The details of how Lloyd George replaced Asquith as Prime Minister are far from clear. Accounts subsequently written by some Of the chief actors are often in conflict. 121 The fact that the War was going badly for Britain and that Asquith was being virulently criticized in the Press, in 1916, set the stage for the leadership crisis. Despite these difficulties, however, Asquith's position as Prime Minister was secure, as long as the Cabinet supported him, since the alternative would have been to replace the entire leadership team. 127- However, some leading Cabinet members-~notably, Bonar Law, Unionist leader in the House Of Commons, and Lloyd George--were becoming critical of Asquith's conduct as a wartime Prime Minister. After Germany's success in the Balkans in November, 1916, events moved rapidly. On December 3, the Unionist Ministers met and passed a resolution which called for Asquith's resignation. This was verbally, perhaps not clearly, communicated to Asquith by Bonar Law. On December 4, Asquith submitted the resignations of his 121A recent survey of the confused accounts is by Robert Blake, Unrepentant Tory: The Life and Times Of Andrew Bonar Law (New York: St. Martin's Press Inc. , 1956), pp. 312-341. Fortunately, given the perspective of the present study, a brief account of these events is sufficient. After May, 1915, the Cabinet was a coalition of Liberal, Labour, and Unionist members, the so-called 'First Coalition. ' 12‘?‘The weight of this factor, relative to the Press' presumed influence, has been inadequately appreciated by many writers. 208 colleagues and attempted to form a new Cabinet. LlOyd George refused to continue to serve under Asquith. On December 5, Lloyd George asked that he be allowed to state publicly the grounds for his resignation. The Unionist Ministers met again and reiterated their demand for Asquith's resignation. Asquith resigned as Prime Min- ister. The King, who was determined to avoid a general election, 123 asked Bonar Law to form a Government. On December 6, since Asquith refused to accept a subordinate Office, Bonar Law informed the King that he (Law) was unable to form a Government. The King met with the leaders of the Liberal, Labour, and Conservative parties, but no agreement was reached. His Majesty asked Lloyd George to form a Government. On December 7, Lloyd George informed the King that he would form a Government. 124 When Lloyd George accepted the King's commission, his prospects for Labour and Unionist support we re uncertain, '25 but the 123The King had sought legal advice and was prepared to re- fuse a request for a dissolution because of the wartime conditions. Harold Nicolson, K318 George The Fifth (London: Constable 8: Co. Ltd. , 1952), pp. 288-289; Dudley Sommer, Haldane of Cloan (London: George Allen 8: Unwin Ltd. , 1960), pp. 342-343. 124The above schematic description is intended to be totally neutral. The intentions and motivations of the various actors, except the King, are the subjects of great controversy. See Blake, op. cit. pp. 312-341. 125Bonar Law was known to be willing to serve under Lloyd George. 209 leading Labourites and Unionists in the First (Asquithian) Coalition proceeded to accept office in the Second (Lloyd Georgian) Coalition. Within his own Party, Lloyd George could count only upon second- echelon assistance. Although the Liberal members of the First Coali- tion Cabinet agreed that a Government could not carry on without Lloyd George, 136 these leading Liberals associated themselves with Asquith who was not willing to serve under LlOyd George. 127 Never- theless, the evening of December 6, ChristOpher Addison, a Liberal M. P. , informed Lloyd George that he would be supported by a major- ity of the Liberal M. P. s. [O]n Monday [December 4], Kellaway, Glyn Jones and I [Addison] had gone through the list of Members of Parliament which had been made in the summer-time when a crisis was threatened. We divided them into 'doubtfuls' and those whom 126Viscount Samuel, Memoirs (London: Cresset Press, 1945), p. 122. Throughout the period December 3-7, Asquith consulted with his fellow Liberals (with the exception Of Lloyd George) in the Cabinet. Samuel attended these meetings, and his Memoirs contain his notes of the meetings, written virtually immediately after each meeting. 127Samuel was asked to serve but refused. Although associated with Asquith, perhaps some of the other Liberal members of the First Coalition Cabinet would have served, if they had been asked. Walter Runciman stated that "in reforming the Government the present Prime Minister [Lloyd George] gave an invitation to one Liberal Minister. I was not that one. I have been asked by my constituents already why I did not join the new Government. I can only make the simple reply that it was impossible to accept an invitation which I had not received." Liberal Magazine, 1916, p. 598. 210 we thought to be 'for' L.G. , and 1 arranged for a small band of men to canvass round and report through Kellaway. [. . .] Late on Wednesday night I was able to report that L. G. '8 following amounted to 49 out -and-out supporters, whatever happened, and 126 others amongst the Liberal Party who would support him if he could form a Government. 128 During the following months, through the Maurice Debate129 and the general election of 1918,130 Lloyd George retained the support of a majority of the Liberal M. P. s. A Party meeting of the Liberal members of the Houses of Parliament was called for December 8, 1916. The suggestion [ . . .] got about that there was to be a resolution passed [at the meeting] not only swearing loyalty to Asquith, as leader of the Liberal Party, but suggesting that those who supported L.G. were to be cut Off from it. 131 Addison, and a group Of Liberals supporting Lloyd George, protested to the Whips against such a resolution, and somewhat to Addison's lZ8Christ0pher Addison, Politics From Within, 1911-1918 (London: Herbert Jenkins Limited, n.d.), I, pp. 270-271. This seems to have been slightly misinterpreted by Owen, Op. cit. , p. 347. 129Since 98 Liberal M. P. 3 supported Asquith's motion in the Maurice Debate and 71 Liberal M. P. 8 supported Lloyd George's Government, the statement in the text reflects a judgment about how the other Liberals would have voted, if they had voted. The figures are drawn from the Liberal Magazine, 1918, p. 244. 130After the 'coupons', i. e. , Lloyd George's and Bonar Law's endorsements, were distributed for the 1918 general election, Winston Churchill asserted "that the Prime Minister is supported by considerably more than half the Liberal Party in the House of Commons, to say nothing of the general support accorded to him in the constit- uencies [.]" Ibid. , pp. 651-652. 131Addison, op. cit., I, p. 274. 211 surprise, 132 the Liberal supporters of LlOyd George were not ejected from the Liberal Party. The Party meeting endorsed the position outlined in Asquith's address: Why are we here tO-day? We are here to-day because I felt it my duty to resign, not the leadership of our party, though I am quite prepared to do that if I am asked, but I have been compelled to resign the headship of the Government. [. . .] I have been asked and it is a perfectly fair question for you to put to me, why I did not agree to act in a subordinate capacity. [. . .] I really do not think, and my colleagues did not think, that I could as effectually serve the new Government [. . .] as a member of it as I could outside, and outside I am remaining with be mle d3ject[. . .] of lending such help as I can . . .] to assist them in the great task which lies before us.1 3 Eugene Wason, M. P. , moved the following resolution: That this meeting records its thanks to Mr. Asquith for his long and magnificent services to the nation, its unabated confidence in him as leader of the Liberal Party, and its determination to give support to the King's Government engaged in the effective prosecution Of the war. '34 Resolutions of a similar character were soon passed by the National Liberal Federation, the nine English District Liberal Federations, the Scottish Liberal Association, the Ulster Liberal Association, and "also be a large number Of local Liberal Associations. "135 The 132Ibid., 1, p. 275. 133Liberal Magazine, 1916, pp. 621. 625. 134The Times, December 9, 1916. p. 10- 135Liberal Magpzine, 1916, p. 638. 212 Welsh National Liberal Council offered "most cordial congratulations to its fellow-countryman, Mr. Lloyd George," expressed "its deepest gratitude to the Right Hon. H. H. Asquith, " but made no mention of unabated confidence in Asquith as Leader of the Party. 136 During the following months, until the general election of 1918, the resolution adopted at the Party meeting of December 8, 1916 remained the Official position of the Liberal Party.137 The re solution was adOpted unanimously at the Party meeting; Lloyd George, his chief Liberal supporters, '38 and the leading Liberal critics of Asquith did not attend. 139 However, Handel Booth, M. P. , asked "the mover and seconder of the resolution to be a little more candid [. . .] as to what their ideas are in regard to carrying l361bid., p. 638. 137on March 22, 1918 the Council of the National Liberal Federation assured Asquith "of its continued and unabated confidence in him as Leader of the Liberal Party [and pledged] the National Liberal Federation to support the King's Government in the effective prosecution of the War. " Proceedings of the Thirty-Sixth Annual Meeting of the National Liberal Federation, 1919 (London: Liberal Publication Department, 1919), p. 21. 13‘8Addison, for example, did not attend. The Liberal M. P. 5 who attended, including 182 members of the House of Commons, are listed in the Liberal Magazine, 1916, p. 620. 139For example, the younger brothers of Lord Northcliffe, whose newspapers had been vigorously criticizing Asquith, did not attend. Reginald Pound and Geoffrey Harmsworth, Northcliffe (London: Cassell, 1959), p. 514. 213 it out. "140 "With regard to the latter part of the resolution, " Booth continued, "when this meeting terminates it is no use avoiding the fact that men will go from it with totally different ideas of what they voted for. "141 In reply to Booth's query, Asquith observed that "what I myself prOpose to do [. . .] will not affect anybody else [. . .]."1‘1’2 The Liberal Party "suddenly found itself [. . . ] splitting in two."143 Early in 1917, although calling Asquith "the distinguished Leader of the Liberal Party, "144 Lloyd George instituted weekly breakfast meetings with his own (Lloyd George's) followers among the Liberal M. P. 8.145 However, to the end of 1917, it may be said that both in regard to the divisions amongst Liberals and the determination of action in the constituencies, in the event of a general election ensuing, the position was one of drift. 146 140The Times, December 9, 1916, p. 10. 1“Ibid., p. 10. 1421bid., p. 10. 143Auditor Tantum (anon. ), "The New Government, " Fort- Eightly Review, Vol. 107 (January, 1917), p. 44. "The humour of the position is illustrated by the action of two Liberals who on Tuesday reserved seats on both sides of the House. When taken to task by a fellow-member they retorted that they were trying to carry out the Spirit of the Reform Club resolution, which pledged them to support the new Government and to continue under Mr. Asquith's leadership at the same time." The Times, December 14, 1916, p. 9. 144February 3, 1917. Liberal Magazine, 1917, p. 58. 145Addison, op. cit., II, p. 164. 146Ibid., II, p. 166. 214 Early in 1918, Asquith and his supporters sharply criticized Lloyd George's Government 147 and, in the Maurice Debate, moved what Lloyd George considered to be a motion of lack of confidence in the Government. The Government was sustained in the division lobbies by Conservatives, Labourites, and 71 Liberal M. P. s, but 98 Liberal M. P. 3 supported Asquith's motion for an inquiry into the conflicting trOOp estimates of General Maurice and LlOyd George. 148 "This division, The Times pointed out, was 'the debut of an organised Opposition, the first step was taken towards what may become a permanent cleavage".'.149 After the Liberal cleavage in the Maurice Debate and before the general election of 1918, Asquith and Lloyd George each made and refused at least one overture for a reconciliation. During this . period, LlOyd George indirectly Offered Asquith a Cabinet position, 147February, 1918. See the Liberal Migazine, 1918, p. 252; Addison, Op. cit. , II, pp. 234-235. For Asquith's view that a "crisis" was simmering, with what outcome "no one can say exactly, " see Desmond MacCarthy (ed.), H. H. A.: Letters of the Earl of Oxford and Asquith to a Friend (London: Geoffrey Bles, 1933), First Series, 1915-1922, p. 60. 148May, 1918. See the Liberal Magazine, 1918, p. 244. See also above footnote 129. 1‘49Jones, op. cit., p. 150. Jones' volume contains a judicious assessment of the merits of the Maurice Debate. 215 but the invitation was refused. 150 On November 1, 1918 Asquith reaffirmed his Leadership of the Party and extended an indirect invitation to the dissident Liberals: I have been now ten years Leader of the Party. [ . . .] During all that time you gave me your confidence. I have no reason to think that you have withdrawn that confidence now. [. .] so far as I know, there is no disposition in any quarter of the Liberal Party towards the exclusion and ostracism of anybody who is prepared to accept with a whole heart and carry out with conviction the principles of Liberal policy. 151 On November 2, 1918 Lloyd George formally prOposed to Bonar Law that there "should be a Coalition Election [with candidates] pledge[d] to support this Government. "152 The Conservative Party accepted and the Labour Party refused the prOposal. On November 12, 1918 Lloyd George met with his supporters among the Liberal Peers and members of the House of Commons, in addition to a large number of "gentlemen who intend to stand at the forthcoming election as Liberal Coalition candidates. "153 A resolution was unanimously carried at 1505ee the Liberal Magazine, 1918, pp. 651-652, 654; Addison, Op. cit., II, p. 250. 151Speech at Glasgow. Liberal Magazine, 1918, pp. 530- 531. Earlier, on March 22, 1918, Asquith had contended that "I have not resigned that position [i.e. , the Leadership]. I am not aware that I have been deposed from it, and until the day comes when my natural faculties desert me--of which event I am sure I shall receive timely warning from one or other of my candid friends--until that time arrives I have no intention. " Ibid. , p. 154. 157-Ibid.. pp. 585-587. 153Ibid., p. 578. 216 that meeting, stating in part: "Those present at this meeting pledge themselves respectively to stand as or support Liberal Coalition candidates with the Prime Minister as their Leader."154 In the general election campaign, calling Asquith "my late chief, "155 Lloyd George refused to endorse Asquith's supporters. 156 The Liberal headquarters, under Asquith's control, "urged some of the Liberal associations in the constituencies to drOp Coalitionist Liberal candidates [ . . . ] . "157 Lloyd George's Coalition won a sweeping majority in the new House of Commons, and the Liberal Party, under Asquith, was decimated. Asquith, all of his ex-Cabinet 154Ibid., p. 580. Emphasis added. 1551bid., p. 666. 156The point may be made more strongly: Lloyd George's endorsement--the so-called 'coupon'--went to many Opponents Of Asquith's supporters. Those Liberals who had voted for Asquith's motion in the Maurice Debate were, in particular, not given Lloyd George's endorsement. For the most detailed description of the distribution of endorsements, see T. G. Wilson, "The Parliamentary Liberal Party in Britain, 1918-1924, " (Oxford University, D. Phil. dissertation, 1959). 1511251. , pp. 15-16. Neither this policy nor Lloyd George's policy of distributing endorsements, as Wilson's study makes clear, was rigidly and systematically followed. After the election, the National Liberal Federation officially adOpted the policy that "every Liberal Association should jealously guard its independence, and should, without loss of time, cause it to be clearly understood that at the next election its countenance and support will only be given to a Liberal candidate who is independent of other political ties and claims. Proceedings of the Thirty-Sixth Annual Meeting of the National Liberal Federation, 1919, p. 16. 217 colleagues (except, of course, those who joined with Lloyd George), and the entire Liberal Whips' office were defeated. 158 Perhaps Asquith himself has best summarized the 1918 general election and its aftermath for the Liberal Party: The disintegration of the Liberal Party began with the Coupon election of December, 1918. It then received a blow from which it has never since recovered. [ . . .] The Liberal members in the new House were reduced to a handful of little more than thirty. The bulk of the Old Liberal parliamentary party deserted to the Coalition. 159 The conventional judgment, by historians and political scien- tists, is that Lloyd George, rather than Asquith, was primarily re- sponsible for the schism in the Liberal Party, dating from 1916.160 This judgment is commonly based upon the feeling that Lloyd George did not "play the game according to the rules," when he was neither 158Liberal Magazine, 1918, p. 634. 159The quotation is from the 1926 memorandum in which Asquith finally announced his resignation as Leader of the Liberal Party. Quoted in Spender and Asquith, op. cit. , II, pp. 369-371. With these comments, Asquith tacitly admits that Lloyd George was sup- ported by a majority ("the bulk") of the Liberal members of the House of Commons and that the Coalition Liberals were not members of ("deserted") the Liberal Party. 160See D. C. Somervell, "The Twentieth Century," in Sydney D. Bailey (ed. ), The British Party System (2nd ed.; London: The Hansard Society, 1953), p. 39; Samuel H. Beer, "Great Britain: From Governing Elite to Organized Mass Parties," in Sigmund Neumann (ed. ), Modern Political Parties (Chicago: The University of Chicago Press, 1956), p. 16. 218 loyal to his Leader in 1916161 nor loyal to his Party in 1918.162 The thesis of this study is that Asguith, rather than Lloyd Georg_e, did not "play thgpime according_to the rules, " when he remained the Leader of the Party after 1916. 163 The "game" in this analysis is the traditional leadership selection process from Gladstone to Asquith.164 The critical point is that, upon becoming a Liberal Prime Minister, Lloyd George did not automatically become Leader of the Liberal Party. 165 After becoming Liberal Prime Ministers, Glad- stone (1868, 1880), Rosebery (1894) and Campbell-Bannerman (1905) were simply recognized as Leaders of the Liberal Party. Asquith's succession as Leader of the Party (1908) was "subject to [the Liberal M. P. s'] ratification, "166 but Asquith was already (at the time of 161This point is typically made in terms of Lloyd George's allegedly inordinate ambition and/or despicable tactics. 162This point is typically made in terms of Lloyd George's decision to issue 'coupons' as he did in 1918. 163A corollary of the thesis, that need not be argued here, is that the Liberal schism would not have been so severe--indeed, might have been very minor--if Asquith had gracefully acknowledged that Lloyd George became Leader of the Liberal Party by virtue of becoming Prime Minister in 1916. 1640ther "games" might, of course, be defined. This one is not advanced as an absolute standard for judgment. 165This striking departure from convention was frequently and prominently noted at the time. See, for example, Tantum (anon. ), op. cit., p. 35. 166Liberal Magazine, 1916, p. 621. The phrase is Asquith's. 219 "ratification") Leader of the Liberal Party since he was already the Liberal Prime Minister. When Russell (1867) announced his intention not again to take Office as Prime Minister and when Gladstone (1894) announced his final retirement as Prime Minister, these announce- ments were interpreted as also being resignations of the Leadership of the Party. '67 Asquith's resignation as Prime Minister was, by contrast, definitely not a resignation Of the Leadership. 168 The position 'Leader of the Liberal Party' was conventionally an honor accorded each Liberal Prime Minister; Asquith violated this tradi- tion.169 There was ample precedent for LlOyd George's opposition to Asquith in 1916, since intra-party Opposition to the Leader of the Party, although decried, had been rather common. 170 Gladstone, 167The normal resignation of a Prime Minister whose party has suffered an electoral or parliamentary defeat, of course, is a different type. 168Asquith's statement that "I am quite prepared to [resign as Leader] if asked," is quite beside the point. See above footnote 133 and related text but compare with above footnote 151 . 169The motion at the 1916 Party meeting, from the present perspective, should have read: "That this meeting records its thanks to Mr. Asquith for his long and magnificent services to the nation, its [sorrow at his retirement], and its determination to give support to [Mr. Lloyd George as Prime Minister and Leader of the Liberal Party] engaged in the effective prosecution of the war. " For the text of the actual motion that was adOpted, see above in connection with footnote 134. 170Campbell-Bannerman (1895) had criticized Harcourt for 220 Hartington, Rosebery, Harcourt, and Campbell-Bannerman had had virulent (and influential) critics within the Party. 171 When faced with significant Opposition from within the Party, the Leader's usual recourse was to resign, but a determined Leader might suc- cessfully demand that his critics acknowledge him as their Leader, as Campbell-Bannerman did in 1901 , or might purge his critics from the Party, as Gladstone did in 1886. These Options functioned, and were often consciously intended, to unify the Party.172 Asquith did not exercise the traditional Options,173 and the Liberal Party became increasingly estranged. Bright's (1875) judgment of the ideal conditions for leadership --"a leader should step into [and hold] his place by general con- sent--"174 perhaps could not be satisfied in 1916. But Hartington's (1877) judgment of the realities of leadership probably could have neither retiring nor working amicably with Rosebery, but LlOyd George cannot be criticized on similar grounds: he resigned. 171These examples include Liberal leaders in the House of Commons as well as Leaders of the Liberal Party. 172A purge of dissenters is, of course, a drastic tactic for achieving unity. 173In 1918, after Lloyd George ostracized the Asquithians, Asquith finally excommunicated the Lloyd Georgians. l745ee above footnote 26 and related text. 221 been satisfied in 1916. Hartington was the Liberal leader in the House of Commons and Gladstone was a private member, when Hartington wrote to Granville: The fact is that when he [Gladstone] chooses to lead, he must be the leader of the party, [. . .], he has chosen to lead, and no amount of disclaimers will alter the fact, though they may in his Opinion relieve him of some of the re spon- sibility which attaches naturally to leadership. 175 Hartington drew the appropriate inference and tried to resign; but Asquith "[clung] to this barren honour [the Leadership] as a man in a fire might cling to an old coat. "176 IV. Conclusions177 Leaders178 of the Liberal Party were supposed to lead by general consent. This usually meant, in practice, the general consent of the Cabinet or ex-Cabinet members since the Monarch 175See above footnote 43 and related text. 176Blackwood's Magazine, Vol. 201 (1917), p. 137. 177For the leadership selection process in the Conservative Party, which is very similar to the traditional Liberal practices, see R. T. McKenzie, British Political Parties (New York: St. Martin's Press, Inc. , 1955), Chapter II. McKenzie finds Churchill's attitude towards Neville Chamberlain and the Leadership of the Con- servative Party rather puzzling. This is hardly puzzling, if one re- calls that Churchill was a prominent Liberal colleague and friend of Lloyd George during the events described in the present study. See I_b_i§_. , p. 51 n. 178Liberal leaders in the House of Commons and House of Lords, as well as Leaders of the Liberal Party. 222 and Liberal M. P. s normally played only a nominal role, and the National Liberal Federation no role, in the traditional leadership selection process. General consent could be, and often was, with- drawn. Leaders usually resigned, when not supported by a consensus, but sometimes the dissenters were repentant or were excommunicat- ed; thus restoring the unity of the Party. For eight years, Asquith was Leader of the Liberal Party by general consent. When that consent was withdrawn, with Lloyd George's heresy, Asquith kept possession of the Leadership of the Party but neither exacted penitence nor extirpated the dissenters. Thus, from the perspective of the traditional leadership selection process in the Liberal Party, Asquith erred. The informal (albeit patterned) nature of the leadership selection process facilitated Asquith's error of judgment. If pro- cedures for intra-party conflicts had been institutionalized, if a process for removing a leader had been established, perhaps the contest between Asquith and Lloyd George would have been resolved sooner, with less disastrous consequences. 179 The Liberal Party did not recover from Asquith's error. 17C(If these comments are justified, then since the Conserva- tive Party's procedures closely resembled the Liberal Party's practices, the Conservative Party's recent adOption of procedures for electin a leader (when the Party is in Opposition, by a vote of the Conservatives in the House of Commons) probably has reduced the chances that the Conservative Party might suffer a serious schism such as racked the Liberal Party. CHAPTER VI A SUMMARY, CONCLUSION AND PREDICTION The broad thesis of this dissertation has been that the 1918 split was decisive for the decline of the British Liberal Party. This thesis has been supported in a number of ways. First, the leading alternative explanations have been shown to be less persuasive than is sometimes supposed. From an empirical point Of view, 1 a close examination Of the 1900—1914 British elections indicated that the center party explanation was probably not valid and that the electoral system explanation (if valid) should have fore shadow- ed the decline of the Labour Party rather than the decline of the Liberal Party. The last Liberal Government, it was suggested, was more successful than has often been assumed not only in the field of social reform but also in the tri-cornered party competition. , Second, the split thesis has been develOped in some historical and comparative detail. The origins of the schism were traced to lA decision-theoretic model for rational choice behavior by individual voters, presented in Appendix 1, suggests that the center party and electoral system explanations are also not compelling from a purely rational point of view. 223 224 Lloyd George's supersession of Asquith as Prime Minister in 1916, but the most crucial point was the interpretation of the 1918 and 1922 general election results. The interpretation advanced by contemporary Liberals, viz. , that the Liberal Party did not include LlOyd George's faction, was accepted and buttressed by contrasting the Unionist Party's cooperative relations and the Liberal Party's strident rela- tions with Lloyd George's group. The closely related argument that the 1916-1918 Liberal split gave the Labour Party a unique opportunity to become a full-fledged national party, was supported by reference to the 1918 reconstruction of the Labour Party and reinforced by sketches of comparable schisms and the rise of new major parties in other two-party systems. Third, since the absence of a conventionally fixed descriptive terminology was noted at several points, explications were prOposed for key concepts. These definitions were not only consistent with the split thesis but also were consistent with the usual scholarly de scrip- tions of other (non-British) party systems. With this generalized conceptual base, the split thesis was shown to be a special case of a general comparative theory of the rise and decline of major parties in two-party systems. This homeostatic theory was plausibly con- sistent with recent survey research into the psychOIOgy of voting, but since survey research postdates the known instances of the theory, 225 a definite supportive connection could not be demonstrated. The theory was, however, translated into a formal model that should facilitate future testing of its validity since the model is sufficiently rigorous to generate specific predictions. 2 A comparative develOp- ment of the theory also indicated additional shortcomings in the center party and electoral system explanations and suggested that the decline of the Liberal Party was a two-stage rather than single-stage process. Finally, since the Liberal split originated with Lloyd George's supersession of Asquith as Prime Minister, the pattern of previous Liberal leadership conflicts was examined. That pattern was clear: The leader either resigned or forced an early "showdown" with the dissenter(s). Although the Asquith-Lloyd George conflict was com- plicated by the existence of coalition conditions, it was evident that Asquith neither resigned as Leader of the Party nor forced an early "showdown" with Lloyd George. Since the earlier leadership con- flicts had not eventuated in such severe schisms, Asquith's departure from the traditional pattern suggested that he, rather than Lloyd George, should be considered primarily responsible for the Liberal split. 2See below for a prediction about the future prospects of the Liberal Party. 226 Thus the split thesis has been supported both by negative and positive considerations, by British and comparative reflections, and by historical and theoretical arguments. The fundamental conclusion of this analysis is that the split thesis provides the basis for an adequate explanation of the decline of the Liberal Party. Perhaps two aspects of the supporting argu— ments deserve separate mention as additional conclusions. The first is a matter of leadership; the second is a matter of language. The power of party leaders has been stressed by political scientists, since Robert Michels formulated his "Iron Law of Oligarchy. "3 The Iron Law stated that power, within a political party, inevitably gravitated to its leaders. Subsequent scholars--notably Robert McKenzie4 in his excellent study of the distribution of power within the Conservative and Labour parties--have generally con- firmed the Iron Law but have stressed the limitations upon the power of the leaders. Most of these studies--McKenzie's, for example-- have focused upon the policy- making process within parties, although Michels himself was inclined to emphasize the perquisites and status 3Robert Michels, Political Parties, trans. Eden and Cedar Paul (Glencoe, Illinois: The Free Press, 1958). This book was originally published in 1915. 4R. T. McKenzie, British Political Parties (New York: St. Martin's Press, Inc. , 1955). 227 Of the leaders. This study has, in general terms, also confirmed the Iron Law of Oligarchy but, perhaps more importantly, has suggested a relatively unexplored area of the law's application, an area that must raise some doubts about the wisdom of McKenzie's stress upon the limitations of the leaders' power. Within major parties, if the analysis in the preceding chapters may be broadly stated, a few score leaders have the power to endanger the party's survival as a major party. Specifically, within the Liberal Party, perhaps as few as two men--Asquith and Lloyd George--had the power to destroy the Party as a major party. Relative to a political party, this is an awesome power indeed. Informal party practices and internal constraints of conscience, as the Liberal Party's history demonstrates, are not enough to prevent the exercise of that power. The language of political science is (mainly) ordinary languag or, perhaps, the language of politicians. It is a subtle and flexible tool for communication, but as the reference to politicians may have suggested, it is also vague and imprecise. Furthermore, ordinary language is replete with contradictions. The significance of these points for mathematics and science has been stressed by philosOphers of science, since the discovery of Russell's paradox. 5 This study has 5See Willard Van Orman Quine, Mathematical Logi_c_ (rev. ed.; Cambridge: Harvard University Press, 1951), pp. 163-166. 228 illustrated the importance of these considerations for political science in particular. For example, ordinary discourse furnishes at least two, contradictory meanings for the phrase "the Liberal Party in 1918-1922," viz. , Lloyd George's followers either are or are not excluded from the Party. Both of these meanings cannot be retained in a systematic analysis, upon pain of contradiction, and thus a linguistic decision has theoretical significance. Ordinary language is, perhaps, the starting point for a scientific analysis, but as descriptions of the Liberal Party's history illustrate, it is hardly a suitable ending. For systematic, theoretical purposes ordinary dis- course must be refined, if not superseded. This study may appropriately end with a prediction about the Liberal Party's future prospects. If the homeostatic theory Of two- party systems is correct, then assuming the democratic nature of the British political system is not disrupted, the Liberal Party cannot again become a major party unless a severe split racks either the Conservative Party or the Labour Party. Since the Liberal Party has recently been making a serious effort to recoup its lost fortunes, the success or failure Of that effort will probably become apparent in the next few years and will serve as a test for the theory that has been advanced in this study. APPENDIX 1 THE RATIONAL VOTER IN A SINGLE-MEMBER DISTRICT The objective of this note is to present a simplified decision- theoretic model of individual rational voting in single-member dis- tricts under the simple-majority single-ballot method of election, and to derive the following theorem from the model: The rational voter votes for the candidate (party) associated with the outcome he (the voter) most prefers. This theorem may interest political scientists for two reasons. First, within the model, rational voting is a very simple thing. The rational voter needs to know neither the numerical probabilities of the outcomes nor the outcomes' numerical utilities for him. Thus, from the perspective of survey research on voting behavior, the model is realistic in the sense that it does not require the rational voter to make elaborate calculations. Second, the theorem contradicts the common argument that "there is one eventuality in a multiparty system that does not arise in a two-party system: a rational voter may at times vote for a party other than the one he most prefers. "1 1Anthony Downs, An Economic Theory of Democracy (New 229 230 H The model adopts the most widely accepted rule for rational decision-making: The voter (decision- maker) maximizes the expected value of utility. 2 The model also assumes a single-member district Operating under the simple-majority single-ballot system of elections.3 C ,C.The Denote the candidates (parties) by C 2, . . . n 1, alternative courses of action available to the voter are to vote for C1 9 or to vote for C2, ..., or to vote for Cu. (The problem is how the voter should vote rather than whether or not he should vote.) These possible votes are denoted by V , V , .. . , Vn' The possible out- 1 2 comes, ignoring ties,4 are that Cl will win, or C will win, .. . , 2 York: Harper 8: Brothers, Publishers, 1957), p. 47. See also Maurice Duverger, Political Parties, trans. Barbara and Robert North (New York: John Wiley 8: Sons, Inc. , 1955), pp. 225-226. 2"In fact, use of this criterion is often cited as a necessary (if not sufficient) Condition for rational choice. " Russell L. Ackoff, Scientific Method (New York: John Wiley 8: Sons, Inc. , 1962), p. 38. A diffuse intellectual debt is owed to Ackoff's stimulating book. 3This assumption is consistent with the works of Anthony Dawns and Maurice Duverger but, as Joseph Schlesinger has observed, is typically a great simplification of reality. The assumption fits by- elections (special elections) in Britain, Canada and United States, but the general election situation is more complex in general. See Dowm, Op. cit., pp. 23-24; Duverger, Op. cit., p. 223; Joseph A. Schlesing, "The Structure of Competition for Office in the American States, " Behavioral Science, Vol. 5, No. 3 (July, 1960), p. 197. 4Ignoring ties is similar to ignoring the possibility that a coin will fall on its edge. This simplification is convenient since there is no voting choice uniquely associated with a tie outcome. 231 W,...,W.The or C will win. Denote these outcomes by W , n 1 2 _ n voter's relative evaluation of these outcomes (his utilities) are U.5 deSIgnated by U1, U2, .. . , n The (objective) conditional probability that a particular out- come Wi will occur, given a particular vote Vj, can be represented by P(Wi/Vj)' The possible outcomes and alternative courses of action, as well as the corresponding conditional probabilities, are represented in the. following matrix: Pos sible Outcome 3 l W1 W2 . . . Wn Vl P(Wl/Vl) P(WZ/Vl) P(Wn/Vl) V2 P(Wl/VZ) P(WZNZ) P(Wn/VZ) Alternate Courses of Action V P(W/V) P(W/V)... P(W/V) n 1 n 2 n n n It seems clear that the voter's decision has some (although, perhaps, only a very very slight) effect upon the probabilities, that is, that a particular candidate's chances are somewhat enhanced if the voter decides to vote for that candidate. 6 This means, in formal terms, SSee below footnote 8. 6The Shapley-Shubik a priori power index is pertinent. See L. S. Shapley and Martin Shubik, "A Method for Evaluating the 232 that P(Wl/vl) > P(Wl/VZ)’ ..., P(Wl/V1)> P(Wl/Vn)’ and similarly for the probabilities in each of the other columns. If EVi is the expected value of the i'll'1 course of action, the three-candidate problem is given by the following set of equations: (1) [P(Wl/Vl)]Ul + [P(Wz/V1)]U2 + [P(W3/V1)]U3 = EV l (2) [P(Wl/V2)]Ul + [P(WZ/V2)]U2 + [P(W3/V2)]U3 = Ev2 (3) [P(Wl/V3)]U1 + [P(Wz/V3)]U2 + [P(W3/V3)]U3 = Ev3 The voter should select the Vi that maximizes the expected value. These equations can be simplified by observing that, if the voter changes his decision from Vi to V3,, then the probability of Ci winning decreases and the probability of Cj winning increases, but the probability of Cu winning is not affected.7 Thus, in the three-candidate case, P(W1/V2) = P(WI/V3), p(W2/V1) = P(WZ/V3)' P(W3/V1) = P(W3/V2), and the simplified equations are: (4) [P(Wl/Vl)]Ul + [P(Wz/Vl)]UZ + [P(W3/Vl)]U3 EV (5) [P(Wl/V2)]Ul + [P(WZ/VZHUZ + [P(W3/Vl)]U3 EV Distribution of Power in a Committee System," American Political Science Review, V01-XLVIII(1954), pp. 787-792. 7Since this is perhaps counter-intuitive, Objective rather than subjective conditional probabilities are used in the model See the numerical example given below. 233 (6) [P(Wl/V2)]Ul + [P(WZ/V1)]U2 + [P(W3/V3)]U3 = EV3 Subtracting Equation (5) from Equation (4), we obtain: (7) [P(wl /v1)]U1 — [P(Wl/Vz)]U1 + [P(wz/van2 - [P(WZ/VZ)]U2 = EVl-EVZ When the terms in Equation (7) are recombined, we successively Obtain: (8) U1[P(W1/V1) — P(Wl/V2)] + UZ[P(WZ/Vl) - P(Wz/VZ)] = Ev1 - Ev2 (9) U1[P(wl/v1) - P(WI/VZH - U2[P(w2/v2) - P(WZ/Vl)] = EV1 - EVZ Since the alternative courses Of action are assumed to be exclusive and exhaustive, the probabilities in each row of the original matrix have a sum equal to one; and as a result, so do the probabil- ities in each of the Equations (1) - (6). Thus, with respect to Equations (4) and (5), the following result holds: (10) P(Wl/Vl) + P(WZ/Vl) = P(Wl/VZ) + P(WZ/VZ) and transposing we obtain: (11) P(Wl/Vl) - P(Wl/VZ) = P(WZ/VZ) - P(WZ/Vl) If K = P(W1 /V1) - P(Wl/VZ), then K is a positive real number (K > 0) and Equation (9) can be reduced to: (12) KUl - KU2 = EVl - EV2 and recombined as: (13) K(Ul - U2) = EV1 - EVZ 234 Therefore, the ordinal ranking of the expected values depends solely upon the ordinal ranking of the utilities. A corollary is that the rational voter does not need to know the (objective) conditional prob- abilities. By convention the voter most prefers Cl '8 winning--W > 1 8 . > -- Wzand W1 W3 and, hence, U1 > U2 and U1 > U3. Thus, With respect to Equation (13), EV 1 > EVZ' The rational voter selects Vl instead of V2, and by a similar argument9 the rational voter selects V1 instead of V3. Thus, in the three-candidate situation, the rational voter votes for the candidate (party) associated with the outcome he most prefers. A corollary is that the rational voter does not need to know the numerical values of his utilities. This type of argument is Obviously quite general and holds for any number N of candidates in a single-member district using the simple-majority single-ballot system Of election. 8Simple preferences are universally accepted as the basis for utilities. The debate pivots about how to transform the non-numer- ical preferences into numerical utilities, but an isomorphic ordinal ranking is generally accepted as a fundamental characteristic of suitable transformations. See R. Duncan Luce and Howard Raiffa, Games and Decisions (New York: John Wiley 8t Sons, Inc. , 1957), Chapter 2. 9Start by subtracting Equation (6) from Equation (4). 235 III The decision rule (viz. , maximize the expected value of utility) and the analysis of probabilities are, of course, the core of the present model. These can be illustrated, and the relationship between the model and the traditional theory of political parties can be indicated, by a numerical example. Let us suppose that the voter intends to vote for the third candidate C3, viz. , intends to select V3 as a course Of action, and that the (Objective) conditional probabilities are: P(W1/V3) = .40, P(WZ/VB) = .35 and P(W3/V3) = .25. Further, let us suppose the voter changes his mind and decides to vote for the second candidate C2, and that the (objective) conditional probabilities become: P(W1 /V2)= .40, P(WZ/VZ) = .45 and P(W3/V2) = .15. With this hypothetical case, the (let us say) "absolute" probability that Cl will be elected has not changed but the (let us say) "relative" prob- ability that C1 will be elected has strikingly changed; the voter's switch makes C2 rather than C1 the most likely winner. The theory of political parties seems to have focused upon "relative" probabilities, and its implicit decision rule apparently has been: Never vote for the candidate whose winning you least prefer; of the remaining candidates vote for the one who (if you vote for him) has the be st chance of win- ning. This decision rule may be "rational" in some sense, but it is not the decision rule incorporated in the present model. 236 The present model has the advantage, from the perspective of the psychology Of voting, of not postulating calculations by the rational voter. Thus, perhaps, the difference between rational voters and actual voters is smaller than is commonly assumed. This possibility deserves examination by survey research, but the dis- tinction between preferring a candidate's personality and preferring a candidate's winning must be carefully drawn. A highest preference for a particular candidate's personality or stands on the issues or party, is not equivalent to a highest preference for that candidate's winning; in the language of decision theory, the latter is an outcome, the former are not. If this distinction can be successfully drawn in questionnaires, studies of the psychology of voting might be integrated with studies of rational decision- making, with prospects of theoretical advantages for both types of analysis, by construing the so-called sociO-economic determinants of voting choices as determinants of preference structure 5 . who“. am om DOMGOMGD ovvm oNQN neonmd «6mm. womb Edmond.“ squDnBOQ momma a a ~Nw “3:0ch cww hoow neonmd uvow mums Emuoohd coprwounu Nmoo OVNm unweowsb 3mm. mwam usonmd mwow wam Emu cad modoom .onEmmocdd 02.0 N 2:. mmmv when. 2.0m Deficomcb mwwm ooHN mmov vwom comm moons..— oni. memo 02$ oSm mocha. flmnonmd .m . Z . ouMSmxumemiH $23-sz o2; .ooQ oz; .92.. oowmtcoofi com: mooHlOOmL quuumLQ mcofioodm GOSOOEH con—OOH.” mcofloofim oofloomm mcofioofim mo t>m 13230 3.2250 lam Emu 0:00 newm 68.62 *vdodloomfi ..HWHHZOU .HmHZOHZD LmDOmaoiHldémde VJHMDAH MZO ZEH HMOE NEH? mm. UHMHmHQ mmHHHmm N XHQZHnHQJcs 237 238 Hmwm moom umwcomcD gong vwow unendi— omg hwmv Eduonmd .Z 63:93:44 mNHm VNNm amwcowcb mem N ~Nv usonmd ammo oaom Emuonwd co>oO .ouwamxudemd FNNM $3.4” $1.5ch mvvm momm udonmA mohm ahwm #9334 33039me .BOmmSO vowm m How om Hm uwmcomcb oopm ~m¢~ Once. .3934 vooo «.th oowo Emuofld .m 23564 2.5m mmam wawv Homv Damage—D wmov owom NNvm mfiwm usODmA wmwb wmz. mowm Nome Emuonwd anmuoppsm $21.42; 0:; .Oom oz; .cmh oooaleooa boom moaduooom “3.3me mcofioofim cofloofim cofloofim ecofloodm COEOOEH mcofloofim no lemm EmpoGOO EmuOcoO Jam EMHOGOO Jam oEmZ 239 88 SE seasons: 33 82 .333 vamm mot. Eduonwd oBOHO .onwflmoau 22% $3. oumeomeb Nome. mama. ascend 38 $3. Eugen—3 30.2.3. .SMQHDQ N of. o hm~ umchmGD wmfl 5mm: .3083 been momm L333 Loin—own omohcog ahmv mommy ummGOMCD deal 02; Moons...— wmom moon Damaged fisoguoxooO Casing—EGO mums wmmm $3025 mg“ flown usonmd NNVw m H ow. Essen—Had .272 .oflSmxumodfl vHoHlZoH 0:: .OOQ o2; deb moafilooofi mom; moofiloomfi 33”.me mcofioofim GOEOOLM cofioofim moon—03H GOEOBH mcoEooEm mo lam EmuocoO DMHOGOO JAM flamenco Jam 0824 240 Jim: on :3 Scum mm mocodvom Emuoaaou 05 .033 Desmond ofi 5 .HH HODQMSO cw m.~ 3an OD moDOEOOH of com .mooudOm your." wam Nwwm umMGOMGD bayom vaM usonmd omnw woov dduofld >3£wmom om Hm owmw um “c035. ovmm vmbN nsoadd mvam balk fimuonwd uomuumwn— £34 Norm 2.3m ummcoflcD pvwm «comm usonmd mmoo mohm 3.33..— 32 .OumeHmcmA oqmm mvmm umwcowcb momm 05mm pooped Hmmv ammm Emuoohd vcmmxoeafi dosmmm .Eegusfl onmm mwom umwcowcb moan mvcfi usonmg wwhv ammo fiduoohd fuCflSOm .ofifimviow $22.41: oz: .ooQ oz; .53. oomatooafi wood mooalocoa ”.35me mcouooflm cofloofim eoflooflm mcoflooflm cofloofim mcofioofim mo term DNHOGOO fiasco—U Jam Eduoeoo Jam OQLGZ BIBLIOGRAPHY Public Documents and Party Publications Great Britain. Board of Trade (Labour Department). Report on Strikes and Lock-outs. Cd. 5325. 1910. Great Britain. Hansard's Parliamentary Debates (3rd series). Vol. CCXLVII. Great Britain. Ministry of Labour. Eighteenth Abstract of Labour Statistics of the United Kingdom. Cmd. 2740. 1927. Labour Party. Rgport of the Annual Conference Of the Labour Party, January 23-25, 1918. Labour Party. Report of the Annual Conference of the Labour Party, June, 1918. Labour Party. Report of the Special and Annual Conferences of the Labour Party, 1914. Liberal Party. The Liberal Magazine. 1906-1932. Liberal Party. The Liberal Year Book. 1905-1932. Liberal Party. Proceedipgs of the Thirty-Sixth Annual Meeting—of the National Liberal Federation, 1919. Unionist Party. The Constitutional Year Book. 1920-1932. U. S. Bureau of the Census. Historical Statistics of the United States, Colonial Times to 1957. Books Ackoff, Russell L. Scientific Method. New York: John Wiley 8: Sons, Inc., 1962. 241 242 Addison, Christopher. Politics From Within, 1911-1918. 2 vols. London: Herbert Jenkins Limited, n.d. Alford, Robert R. Party and Society. Chicago: Rand McNally 8: Company, 1963. Argyll, Dowager Duchess of, (ed.). George Douglas, Eighth Duke of Argyll: Autobiography and Memoirs. New York: E. P. Dutton and Company, 1906. Bassett, Reginald. Nineteen thirty-one Political Crisis. London: Macmillan, 1958. Bealey, Frank, and Henry Pelling. Labour and Politics 1900 - 1906. London: Macmillan 8: CO. , Ltd. , 1958. Beaverbrook, Lord. Politicians and the War 1914-1916. Garden City, New York: Doubleday, Doran and Company, Inc., 1928. The Decline and Fall of Lloyd Georgp. New York: Duell, Sloan and Pearce, 196 3. Beer, M. A History of British Socialism. 2 vols. London: George Allen 8: Unwin, Ltd. , 1953. Binkley, Wilfred E. American Political Parties, Their Natural History. 3rd. ed. New York: Alfred A. Knopf, 1958. Black, Duncan. The Theory of Committees and Elections. Cambridge: Cambridge University Press, 1963. Blake, Robert. Unrepentant Tory, The Life and Times of Andrew Bonar Law. New York: St. Martin's Press, Inc., 1956. Bowley, A. L. Wages and Income Since 1860. Cambridge: Cambridge University Press, 1937. Burks, Ardath W. The Government of Japan. New York: Thomas Y. Crowell CO. , 1961 . Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes. The American Voter. New York: John Wiley 8: Sons, Inc., 1960. 243 Campbell, Angus, Gerald Gurin, and Warren E. Miller. The Voter Decides. Evanston, Illinois: Row, Peterson and Company, 1954. Chamberlain, Joseph. A Political Memoir, 1880-92. C. H. D. Howard (ed.). London: Batchworth Press, 1953. Churchill, Winston Spencer. Lord Randolph Churchill. 2 vols. New York: The Macmillan Company, 1906. Churchman, C. West. Prediction and Optimal Decision. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1961. Clokie, H. McD. Canadian Government and Politics. Toronto: Longmans, Green 8: Company, 1944. Cole, G. D. H. A History of the Labour Party From 1914. London: Routledge 8t Kegan Paul Ltd. , 1951 . British Working Class Politics 1832-1914. London: George Routledge 8: Sons, Ltd. , 1941. Crewe, Marquess of. Lord Rosebery. 2 vols. London: John Murray, 1931. Cross, Colin. The Liberals In Power (1905-1914). London: Barrie and Rockliff with Pall Mall Press, 1963. Cruikshank, R. J. The Liberal Party. London: Collins, 1948. Dangerfield, George. The Strange Death of Liberal England, 1910- 1914. New York: Capricorn Books, 1961. Dawson, Robert MacGregor. The Government of Canada. Toronto: The University of Toronto Press, 1947. Downs, Anthony. An Economic Theory Of Democracy. New York: Harper 8: Brothers, Publishers, 1957. Duhem, Pierre. The Aim and Structure of Physical Theory. Trans. Philip P. Wiener. New York: Atheneum, 1962. Duverger, Maurice. Political Parties. Trans. Barbara and Robert North. New York: John Wiley 8: Sons, Inc., 1955. 244 Elliott, Hon. Arthur D. The Life of George Joachim Goschen, First Viscount Goschen. 2 vols. London: Longmans, Green and Co. , 1911. Finer, Herman. Theopy and Practice of Modern Government. Rev. ed. New York: Henry Holt and Company, 1949. Fitzmaurice, Lord Edmond. The Life of Granville George Leveson Gower, Second Earl Granville. 2 vols. London: Longmans, Green, and Co., 1913. Friedrich, Carl J. Constitutional Government and Democracy. Rev. ed. Boston: Ginn and Company, 1950. Fyfe, Hamilton. The British Liberal Party. London: George Allen 8: Unwin Ltd. , 1928. Gardiner, A. G. The Life of Sir William Harcourt. 2 vols. London: Constable 8: Company, Ltd. , 1923. Garvin, J. L. The Life of Joseph Chamberlain. 5 vols. London: Macmillan and Co. , Limited, 1932- . Goodman, William. The Two-Parthystem in the United States. 2nd ed. Princeton: D. Van Nostrand Co., Inc., 1960. Haldane, Richard Burdon. An Autobiography. London: Hodder and Stoughton Limited, 1929 . Halevy, Elie. The Rule of Democracy. Trans. E. I. Watkin. New York: Barnes 8: Noble, Inc., 1961. Hamilton, Mary Agnes. Arthur Henderson, A Biography. London: William Heinemann Ltd. , 1938. Heckscher, Gunnar. The Study of Comparative Government and Politics. New York: The Macmillan Company, 1957. Heidenheimer, Arnold J. The Governments of Germany. New York: Thomas Y. Crowell Company, 1961. Hempel, Carl C. Fundamentals of Concept Formation in Emirical Science. Chicago: University Of Chicago Press, 1952. 245 Holland, Bernard. The Life of Spencer Compton, Eighth Duke of Devonshire. 2 vols. London: Longmans, Green and Co. , 1911. Hutchinson, Horace G. (ed. ). Private Diaries of the Rt. Hon. Sir Algernon West. London: John Murray, 1922. Jenkins, Roy. Sir Charles Dilke. London: Collins, 1958. Jennings, Sir Ivor. Party Politics. 3 vols. Cambridge: Cambridge University Press, 1960-1962. The British Constitution. 4th ed. Cambridge: Cambridge University Press, 1961. Jones, Thomas. Lloyd George. London: Oxford University Press, 1961. Karpat, Kemal H. Turkey's Politics: The Transition to a Multi-party System. Princeton: Princeton University Press, 1959. Kemey, John G. and J. Laurie Snell. Finite Markov Chains. Prince- ton: D. Van Nostrand Company, Inc., 1960. Keynes, John Maynard. General Theory of Emplolment, Interest and Money. New York: Harcourt Brace, 1936. Lakeman, Enid, and James D. Lambert. Votingpin Democracies. London: Faber and Faber, 1959. Lazarsfeld, Paul F. , Bernard Berelson, and Hazel Gaudet, The PeOple's Choice. 2nd ed. New York: Columbia University Press, 1948. Lenczowski, George. The Middle East in World Affairs. 2nd ed. Ithaca, New York: Cornell University Press, 1956. Lowell, A. Lawrence. The Government of England. 2 vols. New York: The Macmillan Company, 1908. Luce, Duncan R. , and Howard Raiffa. Games and Decisions. New York: John Wiley 8: Sons, Inc. , 1957. 246 Lucy, Henry W. A Diary of Two Parliaments: The Disraeli Parlia- ment1874-1880. 2nd ed. London: Cassell 8r Company, Limited, 1885. The Right Honorable W. E. Gladstone. Boston: Roberts Brothers, 1895. Lyons, F. S. L. The Irish Parliamentary Party 1890-1910. London: Faber and Faber, Ltd. , 1951. Mabileau, Albert. Le Parti Liberal Dans Le Systeme Constitutionnel Britannigue. Paris: Libraire Armand Colin, 1953. MacCarthy, Desmond, (ed. ). H. H. A.: Letters of the Earl of Oxford and Asquith to a Friend. First Series, 1915-1922. London: Geoffrey Bles, 1933. Macoby, S. English Radicalism 1886-1914. London: George Allen 8r Unwin Ltd. , 1953. Macridis, Roy C. and Bernard E. Brown. The De Gaulle Republic. Homewood, Illinois: The Dorsey Press, Inc., 1960. Magnus, Philip. Gladstone, A BiOQaphy. New York: E. P. Dutton 8t Co., Inc., 1954. Marsh, David C. The Changing Social Structure of England and Wales, 1871-1951. London: Routledge 8r Kegan Paul Ltd. , 1958. McKenzie, R. T. British Political Parties. New York: St. Martin's Press, Inc., 1955. Michels, Robert. Political Parties. Trans. Eden and Cedar Paul. Glencoe, Illinois: The Free Press, 1958. Mill, James. An Essay on Government. New York: The Liberal Arts Press, 1955. Milne, R. S., and H. C. Mackenzie. Straight Fight. London: The Hansard Society, 1954. Mitchell, B. R. Abstract of British Historical Statistics. Cambridge: Cambridge University Press, 1962. 247 Morley, John. The Life of William Ewart Gladstone. 3 vols. London: Macmillan and Co. , Limited, 1903. Recollections. 2 vols. New York: The Macmillan Company, 1917. Morton, W. L. The Progressive Party in Canada. Toronto: Toronto University Press, 1950. Mowat, Charles Loch. Britain Between the Wars, 1918-1940. Chicago: University of Chicago Press, 1955. Mowry, George. Theodore Roosevelt and the Progressive Movement. New York: Hill and Wang, 1960. Muir, Ramsay. How Britain is Governed. New York: Richard R. Smith, Inc., 1930. Nicolson, Harold. King George The Fifth. London: Constable 8: CO. , Ltd. , 1952. Overacker, Louise. The Australian Party System. New Haven: Yale University Press, 1952. Owen, Frank. Tempestuous Journey, Lloyd George His Life and Times. New York: McGraw—Hill Book Company, Inc. , 1955. Oxford and Asquith, Earl Of. Memories and Reflections. 2 vols. Boston: Little, Brown, and Company, 1928. Fifty Years of British Parliament. 2 vols. Boston: Little, Brown and Company, 1926. Parker, John. Labour Marches On. New York: Penguin Books, 1947. Pelling, Henry. Modern Britain, 1885-1955. Edinburgh: Thomas Nelson and Sons Ltd. , 1960. A Short History of the Labour Party.) London: Macmillan 8r Co., Ltd., 1961. Pickles, Dorothy. The Fifth French Republic. Rev. ed. New York: Frederick A. Praeger, Publisher, 1962. i l. l. .l. I 248 Pound, Reginald, and Geoffrey Harmsworth. Northcliffe. London: Cassell, 1959. Quine, Willard Van Orman. Mathematical Logic. Rev. ed. Cambridge: Harvard University Press, 1951. Ramm, Agatha (ed. ). The Political Correppondence of Mr. Gladstone and Lord Granville, 1876-1886. 2 vols. Oxford: Clarendon Press, 1962. Ranney, Austin and Willmoore Kendall. Democracy and the American Party System. New York: Harcourt, Brace and Company, 1956. Reid, T. Wemyss. The Life of the Rt. Hon. W. E. Forster. 2 vols. London: Chapman and Hall Limited, 1888. Rice, Stuart A. Quantitative Methods in Politics. New York: Alfred A. KnOpf, 1928. Rosewater, Victor. Back Stage in 1912. Philadelphia: Dorrance 8: Company, Inc., 1932. Rustow, Dankwart A. The Politics of Compromise. Princeton: Princeton University Press, 1955. Samuel, Viscount. Memoirs. London: Cresset Press, 1945. Sawer, Geoffrey. Australian Federal Politics and Law, 1901-1929. Victoria: Melbourne University Press, 1956. Scalapino, Robert A. and Junnosuke Masumi. Parties and Politics in Contemporary Japan. Berkeley and Los Angeles: University of California Press, 1962. Scarrow, Howard A. Canada Votes. New Orleans: The Hauser Press, 1962. Schattschneider, E. E. Party Government. New York: Rinehart 8: Company, Inc., 1942. Schumpeter, Joseph. Capitalism, Socialism, and Democrafl. 3rd ed. New York: Harper 8: Brothers Publishers, 1950. 249 Sharabi, H. B. Governments and Politics in the Middle East in the Twentieth Century. Princeton: D. Van Nostrand Company, Inc. , 1962. Slesser, Sir Henry. A History of the Liberal Party. London: Hutchin- son 8: Co., n. d. [c. 1944]. Snow, C. P. Science and Government. New York: Mentor Books, 1962. Snowden, Philip, Viscount. An Autobiogragahy. 2 vols. London: Ivor Nicholson and Watson, 1934. Somervell, D. C. British Politics Since 1900. New York: Oxford University Press, 1950. Sommer, Dudley. Haldane of Cloan. London: George Allen 8: Unwin Ltd. , 1960. Spender, J. A. The Life of the Right Hon. SirHenry Campbell- Bannerman. 2 vols. Londm: Hodder and Stoughton, n. d. [c. 1923]. Sir Robert Hudson. London: Cassell and Company, Ltd. , 1930. Spender, J. A. , and Cyril Asquith. Life of Lord Oxford and Asquith. 2 vols. London: Hutchinson 8: Co. , 1932. Stansky, Peter. Ambitions and Strategies: The Strggle for the Leader- ship of the Liberal Party in the 1890's. Oxford: Clarendon Press, 1964. Stout, Hiram Miller. British Government. New York: Oxford Univer- sity Press, 1953. Tilby, A. Wyatt. Lord John Russell. London: Cassell 8: Company, Ltd., 1930. Trevelyan, George Macaulay. Grey of Fallodon. Boston: Houghton Mifflin Company, 1937. West, Sir Algernon. Recollections, 1832-1886. 2 vols. London: Smith, Elder, 8: Co., 1899. 250 Williams, W. E. The Rise of Gladstone to the Leadership of the Liberal Party. Cambridge: Cambridge University Press, 1934. Winkler, Henry R. Great Britain in the Twentieth Century. Publication Number 28 of the Service Center for Teachers of History. Washington: American Historical Association, 1960. Wolf, Lucien. Life of the First Marquess of Ripown. 2 vols. London: John Murray, 1921. Articles, Monographs and Periodicals Anderson, Charles W. "Central American Political Parties: A Functional Approach, " Western Political Quarterly, XV (1962), 125-139. Beer, Samuel H. "Great Britain: From Governing Elite to Organized Mass Parties," Modern Political Parties, Sigmund Neumann, editor. Chicago: University of Chicago Press, 1956. Blackwood's Maggzine, 201 (1917), 13?. Campbell, Angus, and Valen, Henry. "Party Identification in Norway and the United States," Public Opinion Quarterly, XXV (Winter, 1961), 505-525. Carter, Gwendolen M. "The Commonwealth Overseas: Variations on a British Theme," Modern Political Parties, Sigmund Neumann. editor. Chicago: University Of Chicago Press, 1956. Converse, Philip E. , and Depeux, George. "Politicization of the Electorate in France and the United States, " Public Opinion Quarterly, XXVI (Spring, 1962), 1-23. "The Deposition of Lord Rosebery, " The Spectator, 77 (September 19, 1896), 359-360. Eckstein, Harry H. "The British Political System," Patterns of Government, Samuel Beer and Adam Ulam, editors. New York: Random House, 1958. Fitzsimons, M. A. "Midlothian: The Triumph and Frustration of the British Liberal Party," The Review of Politics, 22, No. 2 (April, 1960), 187-201. 251 Friedrich, Carl. J. "Review of Contemporary Political Science, " American Political Science Review, XLVII, No. 2 (June, 1953), 538-541. Grumm, John G. "Theories of Electoral Systems, " Midwest Journal of Political Science, II (1958), 357-376. Gulick, Charles A. "Austria's Socialists in the Trend toward a Two- Party System: An Interpretation of Postwar Elections, " Western Political Quarterly, XI (1958), 539-562. Hempel, Carl G. "Deductive-Nomological vs. Statistical Explanation, " Minnesota Studies in the PhilosophLof Science, III, Herbert Feigl and Grover Maxwell, editors. Minneapolis: University of Minnesota Press, 1962. "The Logic Of Functional Analysis, " Symposium on Sociolog- ical Them, Llewellyn Gross, editor. Evanston, Illinois: Row, Peterson and Company, 1959. Hopkins, J. Castell. The Canadian Annual Review Of Public Affairs, 1917. Toronto: The Annual Review Publishing Company, Limited, 1918. Humphreys, John H. "The Alternative Vote or Proportional Repre- sentation," Liberal Year Book, (1919). 163-168. Karpat, Kemal H. "The Turkish Elections of 1957, " Western Political Quarterll, XIV (1961), 436-459. Kirchheimer, Otto. "The Waning of Opposition in Parliamentary Regimes," Social Research, XXIV (1957), 127-156. Leys, Colin. "Models, Theories and the Theory of Political Parties," Political Studies, VII, NO. 2 (1959). 127-146. Lipset, Seymour Martin. "Party Systems and the Representation of Social Groups, " European Journal of Sociology, I, NO. l (1960), 3-38. Lipson, Leslie. "Party Systems in the United Kingdom and the Older Commonwealth: Causes, Resemblances, and Variations," Political Studies, V11 (1959), 12-31. 252 "The Two-Party System in British Politics, " American Political Science Review, XLVII, No. 2 (June, 1953), 337-358. "Lord Rosebery's Resignation, " The Spectator, 77 (October 10, 1896), 472. "Mr. Gladstone's Resignation," The Spectator, 72 (March 3, 1894), 288-289. Padgett, L. Vincent. "Mexico's One-Party System: A Re-Evaluation," American Political Science Review, LI (1957). 955-975. POpper, Karl R. "Prediction and Prophecy in the Social Sciences, " Theories of History, Patrick Gardiner, editor. Glencoe, Illinois: The Free Press, 1959. Scarrow, Howard A. "Distinguishing Between Political Parties--the Case Of Canada," Midwest Journal of Political Science, IX, No. 1 (February, 1965), 61-76. Schlesinger, Joseph A. "The Structure Of Competition for Office in the American States, " Behavioral Science, 5, NO. 3 (July, 1960), 197-210. Shapley, L. S. , and Shubik, Martin. "A Method for Evaluating the Distribution of Power in a Committee System, " American Political Science Review, XLVIII (1954), 787-792. Slesser, Sir Henry. "The Liberal Party: History," The British Party System, Sydney D. Bailey, editor. 2nd ed. London: The Hansard Society, 1953. Smiley, Donald V. "The Two-Party System and One-Party Dominance in the Liberal Democratic State, " Canadian Journal of Economics and Political Science, XXIV, NO. 3 (August, 1958), 312-322. Somervell, D. C. "The Twentieth Century," The British Party System, Sydney D. Bailey, editor. 2nd ed. London: The Hansard Society, 1953. Stokes, Donald E. "Spatial Models of Party Competition, " American Political Science Review, LVII, NO. 2 (June, 1963), 368-377. Tantum, Auditor. [Anon.] "The New Government," Fortnightly Review, 107 (January, 1917), 32-44. 253 The Time 8 (London). Walsh, W. H. "'Meaning' in History," Theories of History, Patrick Gardiner, editor. Glencoe, Illinois: The Free Press, 1959. Webb, L. C. "The Australian Party System, " The Australian Political Party System, S. R. Davis, W. McMahon, A. A. Calwell, and L. C. Webb. Sydney: Angus and Robertson, 1954. Wildavsky, Aaron B. "A Methodological Critique of Duverger's Political Parties, " Journal of Politics, XXI (1959). 303-318. Unpublished Mate rials Goodman, Gordon L. "The Liberal Unionist Party 1886-1895. " Un- published Ph.D. dissertation, University of Chicago, 1956. King, Anthony. "The Decline of the British Liberal Party. " Paper read before the American Political Science Association Annual Meeting, 1962. Paterson, William. "The Progressive Movement 1919-1930." Un- published M.A. thesis, University of Toronto, 1940. Wilson, T. G. "The Parliamentary Liberal Party in Britain, 1918- 1924.‘' Unpublished D. Phil. dissertation, Oxford University, 1959. llllllllllllllllll 6720 m1 6 o 3 o 3 9 2 1 ” Am