John Barnes, Historian - Political Change

Political Change

Lecture 2

18 October 2002

The Making of the Conservative Hegemony

The nature of the problem:

Since the passage of the 4^th Reform Act in 1918, there have been twenty-three elections. The Conservatives have won ten in its own right and three with partners - in all fifteen parliaments with a clear Tory majority, even when in coalition. More to the point it has invariably had a sizeable majority with which to govern, while Labour has provided the only minority administrations (three plus one that occurred during the 1974-79 parliament) and the two smallest majorities (parliaments lasting twenty and eighteen months). In addition in three elections that they lost, the Conservatives had the larger share of the vote; the reverse is true only in 1951.
If, for the moment we discount the period after 1997 since it could be argued that a major shift in the pattern looks to have taken place, between December 1918 and May 1997, the Conservative party was in office 58 years, just under three quarters of the period. The record is not much less impressive if we take the postwar period up till 1997, eight victories in fifteen elections but a total of 34 years 7 months in office out of a possible 51 years 7 months, two thirds of the time.
Not only was the 18 year term between 1979 and 1997 comfortably the longest since the 1832 Reform Act, but the longest terms in office before that date were the fourteen years between 1931 and 1945 and the thirteen years between 1951 and 1964. NB: both since 1918 and on both occasions the Conservatives formed the government.
Kenneth Wald has shown conclusively that the pre 1914 pattern of confessional politics gave place to class politics in 1918. David Robertson has shown that at least until 1966, the economic dimension was dominant in British electoral competition and political scientists like Alford and Pulzer have suggested that class is the major determinant of voting behaviour.
Nevertheless in an age of mass suffrage, the Conservatives are in power during the period 1918-1940 with only two short breaks, the first for nine months in 1924 and the second just over two years between 1929 and 1931. The build up of Labour support is checked in 1931 and the evidence suggests that the Conservatives would have won a third successive victory (albeit in conjunction with allies) in 1940 had war not broken out.
On three occasions (1924, 1931, and 1935), once on their own and twice with allies, the Conservatives captured 50% or more of the working class vote, a feat not matched by the Labour party even in 1929. Even if Labour had no middle class support at all in that year (unlikely) their working class vote cannot have exceeded 48%. It is more likely to have been some 4 to 6 points lower.
Kinnear has studied the electoral geography of the Conservative dominance and has shown that of the 200 constituencies with a middle class element of a fifth or more, the Conservatives won between 130 and 180 in the period up to 1931. Much more vital to their success was their ability to win at least 130 seats where the working class element was 80% or more. With the exception of 1923 and 1929 they usually won another hundred seats in this group.
Once the Conservatives had recovered from their defeat in 1945, they were in office from 1951-64, and that result could not have been achieved had not a third of the working class continued to vote Conservative. (see Table 1)

Table 1 Class differences in voting behaviour 1950-64 [Gallup]

Average Plus i.e. Upper middle class (6% of 1964 sample)

Party	1950	1951	1955	1959	1964
Conservative	79	90	89	87	77
Labour	9	6	9	6	9
Liberal	12	4	2	7	14

Average or Middle Class (22% of 1964 sample)

Party	1950	1951	1955	1959	1964
Conservative	69	73	77	76	65
Labour	17	22	21	16	22
Liberal	14	5	2	8	13

Average Minus or working class (61% of 1964 sample)

Party	1950	1951	1955	1959	1964
Conservative	36	44	41	40	33
Labour	53	52	57	54	53
Liberal	11	4	2	6	14

Very Poor (11% of 1964 sample)

Party	1950	1951	1955	1959	1964
Conservative	24	31	44	25	32
Labour	64	67	54	68	59
Liberal	12	2	2	7	9

These figures, drawn from what is generally held to be an era of considerable voting stability, confirm what might already have been inferred from Conservative success both before and after the second world war: class has never been the sole determinant of voting behaviour in Britain and it is foolish to suggest, as Pulzer once did, that "all else is embellishment and detail."ⁱ Butler and Stokes wrote with greater caution on the basis of their survey findings in 1963 - a bad year for the Conservative party - that "there were strong enough cross-currents in each class for partisanship not to have been determined entirely by class. Yet its pre-eminent role can hardly be questioned."ⁱⁱ The point they are making is illustrated by a table linking party support to class self-image.

Table 2 Party Support by Class Self-Image 1963

Class Self-Image

Middle Working

Partisan	Conservative	79%	28%
Self-Image	Labour	21%	72%

9. However, it is worth noting that in their surveys no more than half their sample, unprompted, thought in terms of social class at all and had to be pushed into assigning themselves to a class. (50% in 1964, 40% in 1966, 30% in 1969 and 43% in 1970).

10. Heath, Curtice and Jowell have criticised the earlier election for classifying the electorate dichotomously into manual and non manual categories, based on market research classifications. Instead they suggest a fivefold classification which pays attention to the market situation and employment conditions of the various types of job:

the salariat - managers, administrators, supervisors, professionals and semi-professionals;
routine non manual - clerks, sales workers, secretaries;
the petty bourgeoisie - farmers, small proprietors and own-account manual workers;
foremen and technicians; and
rank and file manual employees.

Table 5 Class and Vote 1964-87

Year	Party	Salariat	Routine N/Manual	Petty Bourgeois	Foremen etc.	Manual
1964	Cons	62	58	75	38	25
	Labour	19	26	14	46	68
	Liberal	18	16	12	15	7
1966	Cons	61	49	67	34	24
	Labour	25	41	19	61	71
	Liberal	15	10	15	5	5
1970	Cons	62	51	70	39	33
	Labour	29	41	19	56	61
	Liberal	9	9	11	5	6
F1974	Cons	54	45	68	39	24
	Labour	22	29	19	40	60
	Liberal	24	26	13	22	16
O1974	Cons	52	44	71	35	21
	Labour	23	32	13	52	64
	Liberal	25	24	16	13	15
1979	Cons	61	52	77	45	32
	Labour	22	32	13	43	55
	Liberal	17	17	10	11	13
1983	Cons	55	53	71	44	30
	Labour	13	20	12	28	49
	Alliance	31	27	17	28	21
1987	Cons	56	52	65	39	31
	Labour	15	26	16	36	48
	Alliance	29	23	20	24	21

11. Although the overall importance of the working class is clearly declining in these years, it is still worth noting that in their election-winning years, the Conservatives take 30% of the core working class vote and a substantially higher percentage of the foremen and technicians.

12. Clearly we have a problem. Had class really been the sole or even the major determinant of electoral behaviour, Labour would have become the natural party of government and Butler and Stokes provided good reason to believe that the erosion of Tory working class support that took place between the pre and post-war periods would continue until Conservative governments became a thing of the past. As the table above shows the reverse was true, with Labour support falling to below half in even the core working class.

In parenthesis, it is worth noting that the phenomenon of voters "crossing over" and voting for the party representing the opposite class is not confined to working class Tories, but historically they are more important because of their numbers and because they ensure the party a near hegemonic role.

Voting studies are a postwar phenomenon, but used carefully they may throw light on possible explanations of working class Toryism. Six may be noted:

McKenzie and Silver - good on the Conservative appeal, but their stress on deference voting has come under heavy fire.
Butler & Stokes, who argued that parental socialisation had taken place before Labour had expanded its hold on the working class and that working class Conservatism was therefore a transient phenomenon.
Embourgeoisement. Very fashionable after Labour's third successive defeat in 1959, the thesis being that growing affluence made the working class more like the middle class in their habits and hence more likely to vote Conservative. The thesis seemed to be effectively refuted by Goldthorpe and Lockwood's study of affluent car workers in Luton. Two cautions are in order: Vauxhalls were a large unionised industrial plant and there is some evidence (Banbury, Briggs writing on Birmingham) that the Conservative working class will be found in smaller industrial firms. Owner occupation, which was part only of the embourgeoisement hypothesis, deserves separate attention.
Self assigned class is clearly correct, but explains nothing.
Nordlinger, although concerned with deference, also traced the connection between trade union membership and Labour voting and offered evidence to suggest that it was in fact the cause of labour voting.
Jessop, who was sharply critical of explanations that derived from deference and Civility (his critique was of Almond & Verba's Civic Culture), and came to the conclusion that "Traditionalism" mattered, but only in conjunction with structural location.

A number of authors have addressed the problem of Conservative success between the wars and arguably in so doing they shed light on some of the reasons for continuing Conservative success. However, before discussing them, it is worth returning to the major precursor of the view put forward by Blair and Paddy Ashdown in the mid 1990s.

Although Martin Pugh made estimates of the relative lack of success the Labour party had in mobilising the working class and drew attention to Kinnear's analysis of the Conservative dominance, he attributes the Conservative success to the break-up of the progressive alliance between the Liberal and Labour parties. Conservatives profited from a divided opposition. To hold this view one has to make certain assumptions:

First, that if some form of PR system had been introduced in 1918, Labour and Liberal voters would have combined against the Conservatives.
Second that, but for 1929-31, Labour would have inexorably extended its hold on the working class.

It is relatively easy to dispose of the second assumption. We know what happened when Liberal candidates withdrew from seats between one election and another and also what happened when they intervened. On the basis of this evidence Ross McKibbin argues that the raw votes actually conceal the Conservative party's popular success. He adopts a two party preferred analysis, that is one which supposes that all voters not voting for the two largest parties are required to choose between them. In calculating the figures for 1923 and 1924, he allocates the Liberal vote as it seems to have split in 1924, that is 3 to 2 in favour of the Conservatives. He suggests that the split in 1929 was even more favourable to them, 3 to 1, although his reasons for thinking so are dependent on his view that by the late 1920s a Liberal vote was a quasi Conservative vote. In the 1930s, with the Liberals split, it is logical to split the Opposition Liberal vote more evenly, although perhaps with a slight edge to the Conservatives and that is what McKibbin does. His conclusion is that the Conservatives would have got a majority of the preferred vote at every election, including the two that they actually lost.

Table 6 Conservative Party Performance 1923-35

Year

1923

1924

1929

1931

1935

Votes actually cast for Cons.

38.0

46.8

38.1

60.7

53.3

Est. Preferred vote

58.0

55.0

67.0

59.0

There are obvious problems with calculating a preferred vote in a country where preferential voting is not required, but the figures for 1924, when the liberal vote was badly squeezed, are revealing and those for the 1930s when a de facto two party system was operating over much of the country, are likely to be more accurate than those for the 1920s. They suggest that we would be right in thinking that throughout the period, there was a large anti-Labour majority in the country against which the Labour party was making little or no headway.

Although one must have some hesitations about a picture coloured by the fact that the Labour party was running an increasing number of candidates in less and less helpful territory, calculations of the vote per opposed candidate go some way to confirm the picture which McKibbin presents.

Table 7

% vote per opposed candidate No of Opposed Candidates

Year

Con

Lab

Lib

Con

Lab

Lib

1923

42.6

41.0

37.8

505

419

442

1924

51.9

38.2

30.9

536

503

334

1929

39.4

39.3

27.7

586

571

513

1931

33.0

509

1935

54.8

40.3

23.9

559

539

161

* National Government

Source: British Political Facts 1900-1985

The failure of the Labour vote per opposed candidate to rise between 1923 and 1929 does not suggest that a dramatic breakthrough was inevitable as Pugh supposed. The obvious qualification, that Labour had more than 150 extra candidates in the field in 1929 than in 1923 is offset by the fact that to be a national candidate for Government, both parties had to be seen to contest most of the seats. What held Labour at bay in 1935 was the consolidation of the anti Labour vote. Even if we accept that the Conservative party were fortunate to be out of office in the great depression, Labour was undoubtedly benefited by World War 2. All the evidence suggests that had there been a peacetime election in 1939 or 1940, the Conservative dominated National Government would have won it.

The evidence proffered by Butler and Stokes to explain the erosion of the Conservative working class vote may well explain some part of the long-term shift to Labour. Table 8 reveals the way in which a degree of class consolidation behind the Labour party took place.

Table 8A Initial preferences among working class children of Conservative and Labour working class parents 1963

Respondent's Initial Parents were

Preference Conservative Labour

Conservative	81%	6%
Labour	19%	94%

Table 8B 1963 preferences among working class children of working class Conservative and Conservative and Labour parents

Respondents current Parents were

Preference Conservative Labour

Conservative	68%	9%
Labour	32%	91%

But on their own figures for 1963 - a bad year for the Tories - more than two thirds of working class respondents with Conservative parents remained obstinately Conservative in their voting habits. Still more to the point, as the lines of transmission back to the electoral preferences of pre 1918 voters lengthen, the persisting strength of the Conservative working class vote suggest that Butler and Stokes put far too much stress on this particular explanation of the decline in their strength. One other factor may come into play. Butler and Stokes noted that social classes were not self-contained and that there was substantial circulation between classes. Overall there was a decline in the number of people doing manual work. On the whole the Conservatives gained more from those moving into the middle class than Labour did from those suffering downward mobility. But since there was a greater tendency for those achieving upward mobility to come from Conservative backgrounds and social aspirations to the middle class, this may in part explain why in the immediate postwar decades Labour appeared to be strengthening its hold on the working class, but without achieving that degree of preponderance that might have allowed the Labour party to replace the Conservatives as the natural party of government.

One important clue to what was happening may be found in Nordlinger's 1967 study of The Working Class Tories. He finds a marked correlation between trade union membership and Labour voting and while this can be explained by the disposition of working class Conservatives not to join unions, there seems to be very little doubt that trade union membership, especially in large industrial plants, does generate Labour voting. The point is reinforced by the electoral history of the West Country. What we have in the Labour party is an expression of organised labour and its interests, and where trade unionism failed to penetrate, as in the south-west, the Labour party made little or no headway. With the partial exception of Falmouth and Camborne, in the interwar years the south west remained highly marginal between the Conservative and Liberal parties, with the former gradually edging ahead in the 1930s. Even after the Second World War, Labour made surprisingly little headway in this area, which latterly has been the scene of a major Liberal revival.

Much of Nordlinger's evidence as with other stories revealed a political culture that was largely elitist in its orientation. Grasping this, Frank Parkin offers the most radical explanation of the phenomenon we are seeking to explain. of these. In his view, what needs to be explained is not Conservative voting amongst the working class, but Labour voting. They were the deviants in a political culture that was essentially Conservative. Briefly, he suggested that within Britain there were a set of dominant or core values, which helped define the characteristics of that society and were antithetical to the values of Socialism. They were to be found in the established Church, the aristocracy, the ancient seats of learning, the public schools, the services, the press, and finally and most important, the institutional complex of private property and capitalist enterprise which dominates the economic sector. They could all be said to in close accord with the ideology of Conservatism. Only the trade unions, the co-operative societies and the non conformist churches offered countervailing forces. Baldwin was alleged to have said that "whichever party may be in office, the Conservatives are always in power". Parkin explained this in what he recognised to be a pluralist society, in which these bodies acted autonomously, by suggesting that these values were so widely diffused that a Labour Government would always find less room for manoeuvre than its Conservative counterpart.

Given that Labour was at odds with both the dominant institutions and the central values of society, Parkin argued that political deviance was "manifested not in working class Conservatism, but rather in electoral support for Socialism on the part of members of any social stratum." Labour voting therefore required certain protective barriers if it was not to be subject to immense pressures to conform, and he variously found these in working class communities made up of a single class with close kinship ties, whether these were the traditional mining villages, communities like Bethnal Green, or the new working class estates like Dagenham. It is not therefore working class status that matters, but access to a normative sub system which can provide the necessary buttress against the dominant value system. A second such buttress, as Marx had predicted, could be found in large scale factory production and in organised labour.

Parkin's was not just an imaginative concept. There was empirical evidence to sustain it, significantly in earlier studies which had not theorised their findings along these lines. Willmott for example found that in Dagenham, no less than 88 per cent of the married men in his sample were Labour voters and that only 13 per cent of the workers identified themselves as middle class, a phenomenon associated with Conservative voting. In less socially homogeneous areas Willmott and Young found that workers were more likely to identify with the middle class, 23 per cent in Greenwich, 31 per cent in Hertford, and 48 per cent in Woodford. "The rule suggested by these four places is that the more the middle class predominates in a district the more working class people identify themselves with it and, incidentally, the more they vote Conservative."ⁱⁱⁱ An earlier study of Greenwich and a later one of Newcastle under Lyme identified the same phenomenon, and, in a famous review of Butler & Stokes for The Times, Iain Macleod found what he memorably dubbed the "chameleon principle" lurking in their pages also.^iv

The most sustained analysis of these phenomena was carried out by Bob Jessop and he is able to establish the importance of "traditionalism" in Britain’s political culture, in effect adherence to the dominant values in society as articulated by the Conservative party. It is not only symbolically identified with the dominant order, but encourages such identification in its literature and platform rhetoric. "The central argument which emerges in the popular party literature is that the Conservatives are uniquely qualified to govern Britain and that the institutions of the country are safe in their hands alone..."^v Not all of the institutions identified by Parkin carry the same weight, but support for the monarchy, religiosity, the independent schools only to a limited extent, the Empire, free enterprise and private property form an important strand in the political culture and one which is particularly associated with Conservative voting. Class consciousness, even of a purely economistic kind, is more strongly associated with Labour voting and is correlated with a structural location which does limit the exposure of working class Labour voters to the central and dominant values in society. In addition they benefit from trade union consciousness which is quite compatible with traditionalism. Jessop concludes that the relationship between traditionalism and electoral choice "can be interpreted in terms of an interaction between structural position, life-style and orientations to the dominant order, such that voting is determined not by any one of these factors alone, but by their dialectical relations. Structural location exercises a constraining influence over life-style and orientations towards the dominant institutional order; in turn these factors influence subsequent social location. MORE"^vi

McKibbin, although clearly not himself enamoured by this explanation, certainly does not dismiss it. Indeed, he could be said to offer an explanation of the success of interwar Conservatism that is in fact complementary to it. Nevertheless he is inclined to play it down, partly by reference to the questionable fact that before 1914 and after 1945 the Conservative vote was "often not the 'normal' vote",^vii and partly because he thinks it exaggerates the extent to which the Labour and Liberal parties disputed the predominant value system. In this he is both right and wrong. The evidence suggests that deference and what has been described as the civility or the civic culture were widely shared values which do little to explain voting behaviour, but they are interrelated with a more significant force, characterised by Jessop as "traditionalism" and this has considerable explanatory force. Finally, he misses the importance of what is also apparent from voting studies, a general respect for private enterprise and private property, which leads to a pragmatic support for the Conservative party as long as it delivers. Although there are dangers in reading back from these findings to an earlier age, it would seem altogether plausible that these feelings might actually be stronger before the state embarked on its role of managing or attempting to manage the economy.

There is considerable evidence that the Conservative party consciously structured its political discourse to reinforce a close identification between its own fortunes and those of the nation, buttressing this with an emphasis on "vertical issues" like Imperialism which distracted attention from class division and on a Disraelian tradition the emphasised the need for social welfare.

Before exploring this theme, it is probably necessary to deal with some contingent factors, none of which wholly explain, although all could be said to contribute to the Conservative's success:

The electoral system - university seats, business vote, county seats. - some favourable effects, but of limited significance.
The Irish breakaway, leaving the Conservatives with Unionist allies in Northern Ireland, but offset by switch of Catholic working class to Labour in defiance of their priests.
Good fortune - the great depression and postwar austerity, but cf. the radicalising effects of WW2, nor is it clear that the Conservatives would have been as ineffective in the face of the onset of the great depression (Tariffs).
The nature of the Labour party - trade union based, patriarchal, economistic - cf. Savage on Preston.
The depoliticisation of industrial relations.

Two other factors are relevant:

Organisation and innovative techniques
The women's vote

But a major factor in the Conservative's was arguably down to the nature of the appeal that Baldwin made:

Protestant and moral. NB Baldwin's success with nonconformity
National with a stress on Englishness
Common sense as opposed to dogma
Social reform
Sound Money
Competence
Constitutionalism and fairness - NB putting Labour in power

Much depended on Baldwin and more particularly his appeal to Liberals, and his mastery of broadcasting/film newsreels enabled him to project his personality. Cf. FDR

NB Philip Williamson's attempt to account for Baldwin's success and significance, also my piece in Solon and Schwarz in Foundations.

Significantly perhaps Churchill (PM 1951-55) had played a major part with Neville Chamberlain in Baldwin's 1924-29 Government in extending social insurance and his principal lieutenants in the postwar Conservative revival (Eden and Butler) could both be seen as Baldwin's "young men".

i. Peter Pulzer: Political Representation and Elections in Britain. Allen & Unwin,1967. p.98

ii. Butler & Stokes 2^nd edition p.77

iii. Notes

iv.

v. McKenzie & Silver Cf. also Harris

vi. Jessop p.193

vii. p.264