Jump to ContentJump to Main Navigation
The Politics of the PoorThe East End of London 1885-1914$

Marc Brodie

Print publication date: 2004

Print ISBN-13: 9780199270552

Published to Oxford Scholarship Online: January 2010

DOI: 10.1093/acprof:oso/9780199270552.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 26 February 2017

(p.214) APPENDIX 4

(p.214) APPENDIX 4

The Politics of the Poor
Oxford University Press

There are a number of points to be made regarding the evidence used for the analysis made in Chapter 3 to determine any possible links between socio-economic conditions, as represented in Booth’s data, Protestant church-attendance figures, and ‘Conservative’ voting patterns. There are also some issues to be specifically addressed in regard to the validity of the use of Poplar and Bow and Bromley figures for this purpose.

First, it does not seem that the connections that were made in the chapter between church attendance and Conservative voting patterns were a function of, for example, a greater number of churches being placed in particular types of areas, and so by default showing greater apparent attendance. This is shown by the separate Church of England figures—counting only church, not ‘mission’, attendance—with only one church to be counted in nearly all the areas, and with usually a ‘contained’ catchment area of the parish. These figures, with one or two exceptions, show very much the same trend (see Fig. 5. as the total Protestant attendance correlation, as do those for Nonconformist attendance (Fig. 6.. Importantly, as Jeffrey Cox notes, for local studies combined figures are the most accurate measure of attendance,1 but these graphs, when taken together, do suggest that the will to attend church in either the established or dissenting churches was generally higher in the same areas where Conservativism was also apparently stronger.

The church attendances used for these correlations also did not reflect any popularity of ‘generous’ churches to which large numbers of the poor may have flocked. Statements by the ministers concerned show that these were not the types of congregations reflected in the highest levels of attendance. The vicar of the extremely well-attended St Matthias’s, Poplar, did not ‘believe it is the duty of the Church to give help’. At the very popular St John’s, Cubitt Town, the vicar similarly said they had a deliberate policy to ‘confine their relief as far as possible to the old, widows, and the sick’,2 and certainly not to those who may have just come along on a Sunday in hope of assistance.

A further consideration for these figures was that they may have been significantly distorted by differences in the number of Irish Catholics in the population of wards, so affecting the relative numbers of potential Protestant churchgoers. This factor could also, of course, directly affect the political character of wards. But the number of Irish in the Poplar and Bow and Bromley constituences was relatively (p.215) low, in the order of 5–6 per cent of the population,3 and only in one area—the ‘Fenian Barracks’ area around Devons Road—was the concentration significantly higher than this. Otherwise, the number of Irish in each ward was insufficient to directly distort the Protestant church-attendance figures. Similarly, unlike other areas of the East End, the percentage of Jewish immigrants in Poplar was too small at this stage to affect these calculations.4 Finally—and this applies to both social indicators and church attendance correlations—the use of only Poplar and Bow and Bromley in this exercise raises the possibility that circumstances—including a relative social ‘stability’—were different in these areas, and any conclusions cannot be applied to other areas of the East End. Reasons why the same type of analysis was less applicable to other areas were discussed in Chapter 3, and evidence from elsewhere was discussed throughout, which does show that similar conclusions can most likely be reached for the East End as a whole. But the issue of the political representativeness for the East End of Poplar and Bow and Bromley’s electoral results, particularly given that area’s stronger socialist movement, is important, as will be discussed below, and this also touches upon the use of the overall term ‘Conservative’ in analysing these results.


Fig. 5. Church of England attendance/Conservative vote


Fig. 6. Nonconformist attendance/Conservative vote

Electoral Material

In regard to the local electoral material used in the chapter, the ward is clearly the smallest political entity upon which a localized political analysis can be attempted for this period, if the evidence is limited to voting figures. Yet ward-level voting results are only available in the East End for this period for local-government elections. For these to be of assistance in indicating the political ‘complexion’ of an area, there needs first to have been a genuine party-political contest reflected by the results. As Davis and Thompson both note, until the late 1880s such contests were rare in both vestry and board-of-guardians elections, and it was only in the mid-1890s that party divisions became relatively clear in these elections.5 But even after this, voter turnout in these elections remained low, and only increased significantly when the vestries were replaced by the borough councils in 1900. As an example, voter turnout in the wards of Bromley increased from 26.4 per cent in the second-last vestry election, in 1898, to 46.2 per cent in the first borough election two (p.218) years later. By 1912 the turnout in the area was 54.8 per cent.6 While these figures are still below the parliamentary election turnouts of the period, which averaged from 70 to 80 per cent, a turnout of around 50 per cent across the borough elections of 1900–12 seems a relatively good indicator of voter inclinations, and certainly is likely to be more accurate than the results for the previous local bodies. This ignores, for the moment, the question of the representativeness of the franchise and those actually voting, as has been touched upon elsewhere. Board-of-Guardian elections, which continued throughout the period and also provided ward results, had voter turnouts significantly lower than the borough contests—although they were relatively high in Poplar—and tended to be less political in nature and more ‘confused’ in results. Ward boundary changes in 1900 also meant that such results could not be compared between the two periods. All this suggests that the 1900–12 borough-council ward results provide the most useful localized electoral evidence for the East End in this period.

However, there is the problem of whether these results can be used in arguments regarding influences upon parliamentary elections in the same area. The major technical difference between the two types of election was that the borough franchise included women. Women constituted around 15–16 per cent of the municipal electorate, but their turnout seems, in London at least, to have been somewhat lower than men’s.7 It has been suggested that some women did vote in a slightly less ‘politicized’ and more ‘moral’ way, as discussed in Chapter 3, but it also seems likely that this tendency may have been strong, but unrecognized, amongst men. The borough results were probably not ‘distorted’ as an indicator of parliamentary preferences because of the different franchise. It was also argued in the Introduction and Chapter 4 that it seems unlikely that voting patterns were significantly altered by the different issues that applied in local and parliamentary elections, but that it is possible that the personalized influences which may have been acting to a greater extent within localized election campaigns may have acted to maintain the Progressive vote at a higher level than that of the Conservatives.

Duncan Tanner has examined some of the advantages and disadvantages of using municipal ward results as indicators, and has concluded that ‘the differences between the municipal and parliamentary franchises evidently did not have a considerable political impact before 1914’.8

Calculating the level of ‘Conservatism’ in each ward raised a number of difficulties. First, there was the issue of identification of party alignment amongst the myriad of descriptions used by, particularly Conservative, municipal candidates in this period, and also the many inaccuracies regarding this in press reports of results.9 While in parliamentary elections, as seen above, very few party labels were (p.219) necessary (although the ‘Conservative’ candidate could be Conservative, Liberal Unionist, or Conservative and Unionist), in municipal elections the party nomenclature used was much more varied and confusing, particularly amongst the Conservatives. Very broadly, it could be said that the Conservatives adopted the label ‘Moderates’ in local elections, while the Liberals came under a ‘Progressive’ umbrella,10 but all this could vary considerably. The Conservatives, in fact, often ran under the guise of local ratepayers’ or other such organizations, and as ‘Municipal Reform’ or similar groupings (as will be discussed specifically regarding Poplar below).

The major groupings counted as ‘Conservatives’ in borough-council elections for the purposes of analysis in this study included the Union of Stepney Ratepayers, the Municipal Alliance, the Municipal Reform party, the Ratepayers’ Union, and the Moderates, as well as a range of clearly Conservative ‘Independents’ in each round of elections. Interestingly, the Liberal/Progessive side also ran under such groupings as the Stepney Council of Public Welfare.

The problem of identification was most pronounced in areas such as Stepney, but could be overcome by using evidence such as newspaper discussions of alignments, checking past party identifications of ‘Independents’, and cross-checking a wide range of local reports. In Poplar and Bow and Bromley, a more fundamental problem concerned the role of the anti-Socialist Municipal Alliance group, which started to play a major role from the 1906 municipal elections onwards. The votes for this group were counted as forming the majority of ‘Conservative’ support in the area for the correlations in the later part of the period examined. The difficult issue is the extent to which this group was actually gaining votes from both Liberal and Conservative supporters as a united reaction against the strength of socialism in the Poplar borough, as Thompson and others have suggested.11 If this was significant, the local ‘Conservative’ results in this borough cannot be seen as sufficiently representative of alignments in other areas of the East End, and the connections made in Chapter 3 cannot be seen to be more widely applicable. However, an analysis of the Conservative vote in the elections of 1900, before the Alliance was active, and the average Alliance vote across the three elections 1906–12, show that there was very little difference either in the ranking of wards according to the relative Conservative or Alliance vote, or in the actual percentage of votes they gained in wards in each of the periods. This is illustrated in Table 10. This suggests a very close alignment indeed between ‘traditional’ Conservative support and that later gained by the Municipal Alliance. But it could be argued that the Alliance in Poplar and Bow and Bromley maintained the non-Socialist/Progressive vote at a high level in the area at a time when it should have been in decline after peaking in 1900. But the years from 1906, when the Alliance became active, were also very good ones for the Conservatives generally across London in local elections. This can be seen by comparing the voting patterns in borough elections in Poplar with other areas of the East End which were not influenced by a strong Socialist party. The average Conservative/Alliance vote in Poplar and Bow and Bromley increased by 2.1 per cent from 1900 to 1906 (the election in which the Alliance became genuinely active). In the wards of the three surrounding seats of Stepney, Mile End, and Limehouse, the average Conservative municipal vote also increased, by a very similar 1.1 per cent, between the same two elections, suggesting that the Alliance itself had only minimal effect on voter alignments.


Table 10. Comparison of Conservative vote and ranking across wards, 1900 and 19061912


1900 vote (%)

1900 rank

1906–12 rank

Average 1906–12 vote (%)











Bow South





Poplar East










Poplar West





Bow Central















Cubitt Town















Bow North















Bow West















Bromley Central





From this it seems likely that the Alliance vote and its social and geographic basis did not differ greatly from a base Conservative vote, so it can be broadly used for the correlations undertaken in Chapter 3 and as indicative of ‘Conservative’ factors elsewhere in the East End. Throughout the study it is further discussed why the influences on politics in Poplar and Bow and Bromley can be seen as not dissimilar to those in other areas of the East End, despite the stronger and earlier growth of socialism there.

Calculating the actual strength of the Conservative vote in each ward also posed difficulties, given that the multi-member system meant that each voter usually had (p.221) three votes to distribute across a range of candidates. I calculated party strength in two ways. First, by the percentage of ‘Conservative’ councillors elected in each ward respectively across the five borough elections. This could, of course, distort the comparison of wards, as it is possible that 100 per cent of councillors could be elected from one party although the party gained only slightly more than a majority of votes.12 This is less likely because of what was a normal spread of votes with the multi-member system, and ticket-splitting, but I also calculated the actual ‘Conservative’ vote in each ward using the following method:

I assumed that with each party ‘ticket’ in a ward, the solid ‘core’ party vote was represented by the vote gained by the party candidate with the least number of votes, this being the candidate who was least likely to have received any votes on the basis of local popularity or profile. The ‘base’ votes of all parties running tickets were calculated, with ‘Independents’ allocated to parties if it was obvious that they were part of a ticket (either from consistency in voting figures or other evidence). The Conservative percentage of these base votes was then calculated. The suitability of this method was indicated by the fact that it had an over 90 per cent correlation with the percentage of Conservative councillors elected in each ward. It seems sound to combine both these methods as indicators of a ranking of Conservative support across the wards in the five elections, as it is the ranking, rather than the exact level, of Conservative support that is important here.

The method used to calculate vote percentages here had the advantage of minimizing the effect on the results of possible distortion of party support by the local popularity or profile of candidates. This is clearly a more significant issue in local than parliamentary elections, and its partial exclusion from the results used in this chapter means that the factors suggested as influencing the results can be more readily applied also to the parliamentary election results.

Religious Survey

To apply the results of the Daily News church-attendance survey to each electoral ward in Poplar and Bow and Bromley, I first adjusted the figures for each parish to provide the total adult parish population and the actual adult congregation for each church. The method was adopted, in a simplified form, from McLeod’s analysis.13 I multiplied the percentage of persons aged 15 and over (the definition of adult by the Daily News) in the census district (in Poplar’s case, 65 per cent) by each parish population to obtain the adult population. I then took the adult church attendance for each church or parish and multiplied this by 19/22 to eliminate those attending twice in the day.14 These two figures were then used to calculate the percentage of (p.222) adult attendance in the parish, or in only the Anglican or Nonconformist churches in the parish. In the few cases where a ward did not include whole parishes, I calculated as closely as possible the percentage of a parish which was inside the ward and adjusted the figures on this basis. In three-quarters of the cases this was not necessary.

The method of calculating the percentage of working-class people who attended church in the example wards of Bromley North-East and North-West, and Bow North was as follows:

The overall percentage of middle-class and white-collar residents calculated from Booth’s tables for the Poplar borough was just over 21 per cent (5 per cent middle class, 16 per cent other non-working class). For individual areas, only the middle-class figure was available.

In Bow North the middle-class figure was 3.6 per cent, so a middle-class and white-collar figure of around 18 per cent could be assumed, and so an 82 per cent working-class population. On the survey day, 376 adults attended a Protestant church in St Mark’s, the relevant parish. Assuming, as discussed in the chapter, that 25 per cent of this attendance was working class, then the calculation is 25 per cent of 376 = 94. The working-class adult population was then 82 per cent of the adult parish population of 7,753 = 6,357. So the percentage of adult working class who attended church was 94/6,357 = 1.5 per cent, or 1 in every 66. The non-working-class attendance figure comes out as 20 per cent.

The same calculation method for the two Bromley North wards (middle class 4.5 per cent, assumed middle class and white collar 21 per cent), with a notional 50 per cent working-class congregation as also discussed, gives a figure of 7.5 per cent working-class attendance, or 1 in 13. The non-working-class figure comes out as 28.3 per cent.

Charles Booth Material

The following is a list of the relevant Booth ‘blocks’ used to calculate poverty and middle-class figures for each ward:

Cubitt Town—Block 8od

Millwall—80b and 80c

Poplar East—81c and 81d

Poplar West—80a

Bromley South-East—81b

Bromley Central—81a and 1/3 81b

Poplar North West—79c and 1/2 82a and 1/2 79b

Bromley South West—1/2 82a and 1/2 82b

Bromley North East and North West—83 a, b, c, and 1/2 82b

Bow South—84c and 84d

Bow North—85b


Bow West—1/2 85a

Bow Central—1/2 85a

List of Wards in Descending Conservative Vote Order, as shown in Figures

Bow South—69.2


Poplar East—59.3

Poplar West—58.5

Bromley North East and North West (combined)—55.9

Bow Central—55.0

Cubitt Town—50.2

Poplar North West—48.1

Bow North—43.9

Bow West—43.1

Bromley South East—41.9

Bromley South West—37.8

Bromley Central—36.0


(1) Cox, English Churches, 31

(2) Booth MSS B169, 61 and 83.

(3) In 1881 the percentage of Irish residents in Poplar borough was calculated at 6.87% of the population, compared to 26% in St George’s-in-the-East, 22% in Whitechapel, and 13% in Limehouse (Stepney registration district). See Stedman, Jones, Outcast London, 148. In 1897 the Catholic priests in Poplar estimated the Catholic population at just below 5% of the total. See Booth MSS B180, 19, 45, 55, and 83.

(4) The total ‘alien’ population of Poplar in 1901 was only around 1%, and had grown by only 526 people in the previous twenty years. See RC Alien, P.P. 1903 IX, Tables LIX–LX, 1011–1013.

(5) Davis, Reforming London, 25–9, and Thompson, Socialists, Liberals and Labour, 79–83

(6) See East London Advertiser, 28 May 1898, 7 col.d and 10 Nov. 1900, 8 col.a, and London Statistics, vol. 23 (1912–13), 32.

(7) See Hollis, Ladies Elect, 38, and North Eastern Leader, 2 June 1894, 2 col.d.

(8) Tanner, ‘Elections, Statistics’, 905 andpassim.

(9) See ibid.894 and Husbands, ‘East End Racism’, 8, for discussions of this issue.

(10) See Thompson, Socialists, Liberals and Labour, 80–1

(11) See ibid. 83 and 187.

(12) See Tanner’s discussion of this, ‘Elections, Statistics’, 895

(13) McLeod, Class and Religion, 305. I do not believe that the differences between 1903 and the census date of 1901 justify the statistical complexity of McLeod’s method.

(14) See ibid.25.