Politics, parties and voters in parliamentary elections 1835-1862
See also: Nigel Morgans’s Desirable Dwellings
5. Pollbooks and the interpretation of political behaviour
There is probably no source potentially so useful for understanding the nature and behaviour of mid-Victorian communities as the pollbooks (except, of course, the census enumerators’ books), but interpreting the information contained in them presents statistical and historical problems which historians have not yet finally solved. These problems arise from: the difficulty of distinguishing candidates by a uniform standard, both between one constituency and another, and from election to election; the variable numbers of candidates; the splitting of votes in different directions, or plumping for a single candidate; the lack of clear understanding of the meanings of occupational categories (e.g. ‘joiner’ might be a wage employee or a master craftsman who in later periods would be denoted ‘builder’); the printers’ convention of distinguishing household suffrage voters by giving their addresses but omitting their occupations; the relatively small proportion of the adult male population enfranchised between 1832 and 1867, almost everywhere except in Preston (90); and, finally, from the sheer labour of counting and analysing votes manually.
Vincent’s heroic pioneering example (91) is now probably more valuable for the insights of its introduction than for the statistics contained in it, and it certainly raises more questions than it was possible for him to answer. For example, the Toryism of butchers (92) might tempt one to speculate on conservative appetites for blood or butchers’ deference to the meat eating rich; but it is quite as likely that their normal geographical concentration in ‘Shambles’ subjected them to communal pressures within small localities. Nossiter (93) at the same early stage attempted a comprehensive statistical analysis of relative party strength for a very large number of constituencies and elections, by methods which Fraser has challenged (94); and Joyce drew attention to the marked localisation of voting within a single town (95).
In the analysis of the occupational and geographical distribution of votes for the Preston elections of 1832 to 1862 which follows, I have adopted a method similar to Fraser’s, but with a critical modification. Fraser wisely decided to reduce statistics to their simplest manageable form by rendering results as ‘an assumed contest between leading Liberal and leading Conservative’ (96). This idea is well adapted to the clear tendency of electors and non-electors to do the same thing in their public displays; but its main weakness is the assumption that the electors perceived contests in the same standardised way as Dr Fraser. A fairly intimate acquaintance with the details of local parliamentary politics in Preston, especially from 1826 to 1852, shows that this is not a safe assumption. Methods of analysis and calculation are described in detail in Appendix 3(b) (pp 380-5).
Briefly, my modification of Dr Fraser’s method is to calculate the results of contests as perceived by contemporaries. The differences between Dr Fraser’s results and mine are apparent in the table below:-
Table 15. Percentage Conservative share of poll in Preston 1835 to 1859
| 1835 | 1837 | 1841 | 1847 | 1852 | 1857 | 1859 | |
| Fraser: | 50.86 | 40.05 | 43.34 | 49.22 | 51.58 | 48.81 | 56.07 |
| Morgan: | 61.00 | 53.8 | 43.5 | 49.2 | 51.6 | 56.7 | 56.2 |
As perceived by contemporaries the elections of 1835, 1837 and 1857 were all clear ‘conservative’ victories. Fraser’s figures for those years fit only with the results as identified by McCalmont’s Parliamentary Pollbook of All Elections 1832-1918 (97).
‘I hate and detest the name of party,’ said one candidate in 1852, basing his claims to represent the borough mainly on his local connections. Dr Fraser’s intelligible errors arise from the fundamental difficulty of attempting sweeping psephology from a secondary source compiled after the consolidation of two national parties. McCalmont’s entry for 1837 (for example)
Fleetwood, P.H. … C – 2165
Stanley H.T. … L – 2092
Thompson, Col T.P. … L – 1385
Smith, Thos … L – 789
is a source so opaque as to be practically useless.
While the local and specific interpretation of voting behaviour is important for the borough as a whole, it is even more so for the miniature worlds of polling districts’ (1832 and 1835), wards (after 1856) and small localities within wards. Except for 1832, all the raw figures and calculations which follow (shown in tables and graphs) have been derived from the application of Dr Fraser’ general method to each election on its own terms, and from the point of view of the ‘Liberal’ challenge to the local Tory establishment. Explanation of the interpretive approach to each election therefore follows, for without it the meaning of the results is unintelligible.
Because the ‘Liberal”candidate in 1832 was squeezed between the supporters of the prophylactic Reform Act and its inflamed Radical opponents (see Chapter II) the only-useful measure of the challenge to the establishment is therefore between the pairs at the opposite extremes, aggregation of the votes cast for each pair being as useful as distinguishing between a marginal ‘leader’ in each case.
In 1855, with four candidates but only one clear pair, the approximate equality of votes for ‘the blue and orange party’ – Fleetwood and Stanley – in each ward proves that voters polled with (the same solidarity as they fought in the streets against the ‘red’ party. As the latter were presumably the supporters of Mitchell’s radical friend Thomas Smith, and as Col. Thompson was identified with the same group of local leaders who were in communication with tile Philosophic Radical Dr Bowring and the Liberal candidate Crawfurd in 1837 (i.e. they were the nucleus of the Liberal party); and because more people polled for him anyway, I have used Thompson’s votes in relation to those for the leading member of the Tory/Whig pair. For the total vote and for four of the wards this was Fleetwood. I have used Stanley’s figures for the two wards where he led (St. John’s and Christ Church).
That the 1837 election, with three candidates, was a contest between only two of them was obvious to all concerned. With Robert Townley Parker, the squire of Cuerden, now as a Tory candidate Sir Peter Hesketh Fleetwood wisely professed political independence, leaving ‘the greens’ of Crawfurd to fight it out with ‘the blues’ of Parker:
‘a half civilised OURAN-OUTANG or MAN MONKEY, lately caught in the Woods at Cuerden, by a party of Conservative Poachers (98)… A newly invented GAG… has been provided, known by the name of the ‘Operative Conservative Association GAG’… No Roman Catholics, or Dissenters… the MONKEY having taken a most unaccountable dislike to persons so denominated…’ (99)
With the further lesson of Parker’s diminutive loaf beside Crawfurd’s large one (with a large joint of meat as an ‘appendage’), the identification of the Established Church with the Corn Laws was clear. Therefore Liberal strength is shown by the ratio of Crawfurd’s vote to Parker’s in the election of 1837.
The 1841 election was a simple contest between two pairs, so the aggregates of each pair provide the appropriate figures for calculation, as in 1832.
With three candidates in 1847, and a single Liberal committee working for two of them to such effect as to achieve a statistically negligible difference between votes for the advanced opinions of Sir George Strickland and the Fishwickian ‘Liberalism’ of Grenfell, the method of calculation is the same as in 1835: Strickland for four of the wards, Grenfell for Christ Church and St. Peter’s, and Parker the only Conservative.
In 1852, Cardinal Wiseman, Grenfell, and James German between them confused Preston voters. The Conservative leaders made clear that it was a religious contest:
the clergy of the Church of England suffered some degree of reproach because… we put forth a paper recommending the Protestants… to give Mr Parker their support…’ (said the Vicar, John Owen Parr)… ‘I look upon the triumph we celebrate… as a triumph of Protestant truth…. there is a combination of powers against that Protestant truth, which requires that it should be carefully guarded (100)
But a significant number of the voters split between Parker and Strickland, which suggests that they were confused (101). Without (at present) the figures for voting in wards, the calculation is… reduced to ‘Leading Conservative/leading Liberal’ (denoted ‘Rad’ in the pollbook).
By 1857 Grenfell was quite clearly distinguishable from Strickland, and occupied a political position statistically similar to that of Fleetwood in 1837. The Liberals and the Roman Catholics ‘concentrated their voting power upon Sir George (Strickland); the Conservatives primarily… in favour of Mr Cross… in many instances gave their second vote to Mr Grenfell’ (102), and the Conservative inclinations of the Grenfell group was made manifest at the hustings by the fact that his nomination by Thomas Miller was seconded by a merchant who split Conservative both then and in 1847 (103). The paradox of Dr Fraser’s Conservative minority (48.81%) for a clear Conservative victory is explained: Strickland should be taken as the ‘Leading Liberal’ and not Grenfell.
For 1859 the calculation of Liberal strength becomes somewhat unreal, because Grenfell was the only ‘Liberal’ in contest with two Conservatives, and although Goodair now seconded Miller’s support for Grenfell, Clifton, the Conservative nominated in place of Parker, benefitted from ‘a number of Roman Catholic electors’ (104). Therefore the Liberal percentage in 1859 – 43.8% – does not mean what it meant two years before. The old combination of economics and religion was breaking up.
At the bye-election of 1862 in a straight fight between the Liberal candidate presented to the electors by Thomas Miller (Fishwick) and John Goodair (St. Peter’s) and the Conservative presented by C:R. Jacson (St. John’s) and William Birley (Fishwick) – all of them millowners – the Liberals collected less than 40% of the votes. Only the old stronghold of Liberalism, St. Peter’s, remained steadfast (see below). And after Thomas Miller’s death in 1865 his successor as head of Horrocks Miller & Co., Edward Hermon, was elected as one of the two Conservative MPs (105).
The Conservatives were compact as a phalanx. The Liberals, though centrally united, were laterally hampered by numerous self-opinionated, crotchety sections, all having a fairly strong regard for the general cause of Liberalism, but each hugging some particular ‘notion’…’ (106).
Perhaps it may now be taken as final proof of the dangers of standardised psephology throughout my period that it is difficult to date the election which Hewitson described in this way. It was 1874 but might have been any after 1847.
With contests at all nine elections from 1832 to 1862 (inclusive) and an electorate whose minimum (in 1859) was 2,656, and something like 250 different occupations, it is difficult to analyse all the pollbooks comprehensively. In any case the printers omitted the occupations of the household suffrage voters in Preston from 1847 onwards. It has not been possible to complete a comprehensive analysis of the political behaviour of this community for the whole period, although the unusual opportunity offered by Preston’s franchise deserves such attention, especially in the light of Dr Joyce’s findings (107). His work, as Professor Moore has acknowledged, (108) shows ‘that pollbooks can be used to reveal the existence of electoral bloc in certain industrial towns’. As my Appendix 9 shows, this phenomenon possibly ante-dates the period studied by Joyce by at least a generation, and therefore raises the question (for Lancashire factory towns, and possibly others also) whether the pattern discerned by Joyce in the 18608 was a vestigial rural habit compatible with and essentially similar to that discovered by Moore, or a characteristic of a new urban age. It is a question which should also concern Dr J. Foster (109) and Dr Thompson (110).
The method adopted here for simplifying the occupational analysis of the pollbooks is described in Appendix (3). The last pollbook which prints the occupation of every voter is 1841; but as half (51%) of the electorate in 1847 still represents 1,570 identifiable occupations I have included that pollbook in my occupational analysis. In both cases I have counted all the voters in each of the selected representative occupations. For an appreciation of the localisation of political difference I have, first, compared the political behaviour of the wards for every election (except 1852) between 1832 and 1862; and, second, examined the pollbook of 1841 for evidence of ‘areas of employer influence within the town… akin to conventional villages’ or islands of political influence (111), by counting party voting in five limited localities; and I have extended the investigation of two of them to other dates. Given the marked and explicit emphasis on Catholic/Anglican opposition evident in other sources, I have also taken account of the building of churches for each denomination.
The results are not decisive. Although they lean heavily in A Dr Joyce’s direction, the reason may be the long survival of the distinction between spinners and weavers observed at the time of the Reform agitation, rather than employer influence.
The first and most obvious characteristic of the series of pollbooks is the political difference between two of the wards throughout. (112) Christ Church ward, although showing a very marked rise in interest in the Liberal candidates presented in 1837 and 1841, was always the least inclined to support the Liberals (see graph above), while St. Peter’s was the most Liberal in every election – with the significant exception of 1837 when the Irish ghetto on its frontier (west of Friargate, in the notoriously site of unhealthy streets and courts about Canal Street: now the site of Preston Polytechnic). Equally clear is the continuity between Radicalism and Liberalism in St. Peter’s ward (see Chapters 2 and 4). The course of Fishwick’s political behaviour is particularly interesting This was the district of ‘New Preston’ where in 1826 a defiant and insulting population had pelted Stanley’s procession with mud, shouting ‘Cobbett for Ever!’. ‘They were so amazingly rude and dangerous that even tax collectors durst not… go amongst them’ (113). Fishwickians were evidently still defiant enough to resist the combined conservative weight of Thomas Miller sen., of Samuel Horrocks, John Paley sen. and of the Swainsons and Birleys at least until Miller died in 1840. The course of Fishwick’s graph then follows the political inclinations of Thomas Miller jun., (despite the fact that he reputedly abstained from interfering with the freedom of voters (114)) sinking from the top of the Liberal league in 1837 to the bottom in 1862, with the marked discontinuity of 1859 reflecting the peculiarity of Miller’s and Grenfell’s ‘Liberalism’ in that election. ‘The population has thickened and civilisation has penetrated into the region…’ said Hewitson in 1869 (115). The taming social influences of the intervening period may well be reflected in the changing voter pattern (see below p.353). In any case, the correspondence between the parliamentary and the municipal patterns of the two outlying cotton mill wards is obvious.
The graph of voting in St. George’s ward follows a course similar to that of Fishwick but less marked. With its eastern apex at the market place, St George’s ward contained the retail shops, small master workshops and crafts such as tailoring and shoemaking around Friargate; and a significant number of smaller cotton mills near the canal, especially those of the Paley family. It also included the Catholic chapel of St. Mary and ‘Little Ireland’: this may explain the contrasting directions of the voting patterns in Fishwick and St. George’s between 1857 and 1859. As the account of municipal elections in the late 1850s shows (above pp 214-5), the electors of this quarter of the town were becoming less respectful of the assumed right of the Paleys to represent their interests either socially or politically.
Finally, Trinity and St. John’s closely matched one another’s political course through the middle of the field. Trinity included the Market Place and they shared the best end of Church Street. Occupationally very mixed, their invite the most interesting work on the middling sort of Victorian townsmen because they followed the same pattern as ‘the Square’ of Christ Church ward, without going quite as far as St. Peter’s ward, in the great Liberal rise of the late 1830s and early 1840s. This suggests that the strength of early Victorian Liberalism in Preston was associated with the concerns of the middlingly independent shopkeepers and craftsmen. Some possible explanations for this Liberalism in terms of the social competition of its support are investigated below.
First, there is the possibility that the mid-century rise of Liberal influence in Preston’s parliamentary elections was caused by changes in the balance of social classes within the electorate, as the old franchise voters slipped into the minority, assuming the newly-registered ten pound householders to consist of newly-risen men identifying new politics with new opposition to an entrenched Tory elite. Such a theory does not fit the facts. First, the reduction of the old franchise was gradual after the heavy loss in 1833; second, old franchise voters heavily outnumbered ten pounders when the Liberal vote was increasing most rapidly, and were still half the electorate in 1847; third, as the histogram above shows, the absolute total of voters of presumed ‘working class’ occupations was still very great in 1841, and not much reduced in 1847; and, fourthly, the upper social groups were far less Liberal than the lower: three quarters of the weavers voted Liberal in 1841 and 1847, but only 55% of the specialist shopkeepers in 1841. (Note: on the histogram bracketed percentages for 1847 indicate occupations such as grocers and drapers, in which voters were more likely to be registered on the new than the old franchise, and therefore cannot be comprehensively identified in the poll list.) Finally, in 1847 most of the ten pounders voted for Parker, the Tory: 616 (55%) voted for him, and just over 500 (45%) for Strickland and Grenfell. In fact 45% of all Parker’s votes and less than 37% of Grenfell’s or Strickland’s votes were polled by ten pounders (116).
Comparison of the voting of the selected (most numerous) occupational groups in 1837 and 1841 shows who the Liberal voters were (Appendix 8) and the diagrammatic maps below show where they lived (117). For convenience, the following groups have been extracted:
(There are some unlikely discrepancies which may cast doubt on the validity of even a 1/3 sample of 3,738 voters, such as the increased number of labourers; but this may be accounted for by old men being reduced to such occupations or to none.) ‘Manufacturer’ is an uncertain category (see Introduction), but their very marked shift towards the Liberal side compared with tie conservativism of the cotton spinners proves that they viewed politics in a different light, for whatever reason. Apart from the textile employers the decisive Liberal support came from all the more humble occupations {except the mechanics).
The voting of spinners and weavers not only repeats the pattern of 1832 (when 72% of weavers but only 46% of spinners voted Radical: see Chapter 2.4b) but because of the now marked shift of the spinners to the Liberal side (58% in 1841) puts in question the degree of employer influence, and therefore the ‘axiom’ of W.A. Abram quoted and rightly suspected by Dr Joyce (119). Although figures, not for spindles and looms but for hands employed, are available for Preston only in 1847 (see Appendix 1) – after a spate of mill building – the calculation is really immaterial because of the large disparity between the politics of millowners and of spinners alone. (in 1837 18% of the ‘cotton spinners’ and 39% of the ‘spinners’ voted Liberal; in 1841, 21% and 58%.) Very few textile employers changed their opinions between 1837 and 1841. The ‘manufacturers’ George Smith and John Hawkins did not vote in 1837, but with the change to spinning in the more salubrious air of St. Peter’s they voted Liberal. T. and W. McGuffog, inhibited from voting in 1837 while they were drapers in Fishergate, were also (apparently) liberated by the current building of their mill out at Murray Street. Thomas Miller, relieved of his father’s shadow, also voted clearly Liberal in 1841. But only one textile employer actually changed his allegiance from Conservative to Liberal between these elections: Joseph Gillow, the Catholic leader (see Chapter 6.4) (120). Between then Miller: and Gillow could probably explain much of the 19% swing of the spinners to the Liberal side: but how much? The diagram and diagrammatic maps (on pages 347-8) show that there were big changes in St. Peter’s, St. George’s and Trinity wards between 1837 and 1841, but virtually none in Fishwick. The largest Irish born population in 1851 was in St. George’s and St. Peter’s wards (121).
The (virtually) unique value of the Preston pollbooks for the study-of ‘working class’ voting in the early Victorian period makes the problems raised by Catholic influence peculiarly tantalising. But even if Liberal majorities were obtained only by a Catholic margin, the very large numbers of shopkeepers, shoemakers, tailors, mechanics; spinners, weavers and labourers, probably three quarters of whom were not Catholics, still remain to be considered. Were they voting Liberal because they were ‘lower middle’ or ‘working ‘class’? Or because in Preston the ‘lower orders’ sought political expression of a class-conscious opposition to a Conservative upper crust? Or because the years 1826 to 1832 had established a radical tradition to which they adhered as firmly as their superiors adhered to the conservative? Or were they influenced by more commanding but finer and less easily recoverable strands of the social networks?
In an attempt to answer some of these questions I have examined the distribution of votes in 1841 within limited localities in five of the wards, the results of which are shown in the table and the (4) diagrammatic maps above (122). There are very clear local differences between limited localities in Christ Church, St. John’s and Fishwick wards. The differences in St. Peter’s ward are less clear, because the Liberal vote as a whole was larger there, but within the localities (B and E on the map) there were fairly clearly defined conservative enclaves, centred on St. Peter’s Square, Adelphi Street and Harrington Street, and on the streets round Park Lane and Hanover Street mills (marked by dotted lines on the map).
Unfortunately correlation of these differences with occupations has been attempted only in Fishwick ward. Here the clear preference for conservatism shown by neighbours of Swainson and Birley’s ‘Big Factory’ and for Liberalism by the inhabitants of New Hall Lane and the streets east of Leeming Street and King Street, is complicated by the impression that it coincides with the relative absence and presence of weavers; and weavers seem to have gone their own radical way regardless of their millowning neighbours. As the table below shows, local differences in Fishwick were much reduced by 1857.
The result of counting a limited locality within St. George’s ward – a strip between (and omitting) Friargate and the canal, extending from the canal wharf northwards to the boundary of St. Peter’s ward – at Fylde Street – was less clear. Sharp differences within small distances were expected because of the presence of Paley’s three mills in the southern half, and because the northern half was ‘Little Ireland’. With 197 voters found in 1841, and 266 in 1835 (dates chosen before and after the 1837 riots) the sample was large enough; but the pattern was muted. In 1841 a line of difference in voting divided the northern third of the area from the rest, as expected, and comparison with the earlier election showed that the political difference had arisen since 1835. But it was not as clear a difference as appeared in the other wards: perhaps because there was no particular cluster of weavers.
In St. John’s ward the Liberal south eastern quarter was populated in general by the same heavy proportion of weavers and similarly impoverished families as the other side of Leeming Street, and this was the principal centre of cellar dwelling in the town. Compared with the servant-employing gentlefolk at the western side of St. John’s ward, these people were all desperately poor. Between the western and eastern extremes of St. John’s ward lies the middle ground where warpers, coal agents, factory managers, overlookers and shopkeepers lived, The degrees of respectability can be measured from street to street westwards along the line of Avenham Lane, being in Oxford Street and Knowsley Street only relative to the frightening mass east of St. James’s Church, but unquestionable by any Preston standards over the frontier in Latham Street, Bank Parade and Bushell Place (123). It would be easy to attribute the difference in voting to social difference alone; but for the fact that the Catholic chapel of St. Augustine had just been built in. the Liberal locality.
No single and unequivocal interpretation of pollbook evidence has emerged. But religious and occupational influences were clearly important, and employer influence much less so. Social class and social networks were determining influences which probably cannot be disentangled, but the price of a loaf may have mattered more. I suspect but cannot prove yet that the obviously heavy influence of religion had much to do with the importance of ‘secondary network’ affiliations. In 1841 ten of the twelve Catteralls listed voted for Parker and Swainson. The Political Union Radical and opponent of Joseph Mitchell in 1832, John Irvin, split his votes between the Liberal Strickland and Swainson, a Conservative millowner. Few men split their votes in that election.
preston history 