Franchise Extension and the British Aristocracy
Samuel Berlinski1
Torun Dewan2
Brenda van Coppenolle3
Abstract. Using evidence from the Second Reform Act, introduced in the United Kingdom
in 1867, we analyze the impact of extending the vote to the unskilled urban population on
the composition of the Cabinet and the background characteristics of Members of Parliament. Exploiting the sharp change in the electorate caused by franchise extension, we separate the effect of reform from that of underlying constituency level traits correlated with the
voting population. Our results are broadly supportive of a claim first made by Laski (1928):
there is no causal effect of the reform on the political role played by the British aristocracy.
1. I NTRODUCTION
Does the expansion of voting rights lead to elected assemblies that are a microcosm of the societies
that they represent? Or are the background characteristics of the men and women elected to office
unaffected by differences in the rules governing the franchise? The question is pertinent if, as
recent evidence suggests, the identity of politicians affects their subsequent performance: studies of
changes in mandated forms of representation in the developing world show that identity is causally
related to different outcomes (Pande, 2003); and recent contributions in political science show that
background characteristics of elected MPS and cabinet ministers affects their performance.
Establishing a relationship between franchise extension and the identity of elected politicians can,
moreover, shed light on an intriguing puzzle in the study of political development. As noted by
Aidt and Jensen (2009) there is a “growing consensus that the extension of the franchise contributed
positively to the growth in government”. But it is unclear which if any political mechanisms links
these factors. In principle there can be several. An obvious candidate is that newly enfranchised
voters are more inclined toward leftist or reform parties. Yet neither Berlinski and Dewan (2011) or
Larcinese (2011) find evidence for such an effect with respect to massive increases in the English
and Italian electorates in the nineteenth and early twentieth centuries.
1
Inter-American Development Bank, IZA and Institute for Fiscal Studies. Email: [email protected]
London School of Economics and Political Science. Email:[email protected]
3
London School of Economics and Political Science. Email:[email protected]
2
1
2
In this paper we return to Victorian Britain and ask whether franchise extension lead to political
institutions that were more open and inclusive. Britain is often seen as the exemplar case of gradual
political development. Unlike in France, where the fate of the aristocracy depended on the form of
the constitution, Britain had an established constitutional monarchy following passage of the Bill
of Rights Act of 1689. Whereas in France the Constituent National Assembly of 1789 swept away
the privileges of the first estate, in Britain successive reform acts of the 19th century established a
broadly representative democracy with the gradual extension of voting rights to ordinary men and
women. But did this affect the characteristics of those who governed?
We explore the relationship between franchise extension and the distribution of political power, and
in particular its impact on the British aristocracy. Cannadine has argued in The Rise and Decline
of the English Aristocracy, that the first three reform acts, passed by Parliament in 1832, 1867, and
1884 respectively, coincided with a decline in the influence of this group in British political and
economic life.4. The claim was disputed by Harold Laski in an article published in the American
Political Science Review in 1928. He argued that
“The three reform acts of the nineteenth century had little effect upon the position of
the aristocracy in politics. Policy may have changed, but the men who made policy
came, in much the same degree, from the same origins as their predecessors.”5
To reach this conclusion Laski looked at the composition of the English Cabinet between 1801 and
1924, focussing in particular on the personnel changes before and after the reforms. His statistics
suggested that little changed between these years. He noted that between 1832 and 1866, the
period that encompasses the First Great Reform Act, the percentage of aristocrats in the Cabinet
decreases from 73 to 64%; whilst between 1867 and 1884, the period that covers the Second and
Third Great Reform Acts, the percentage of aristocrats in the Cabinet decreases further from 64
to 60%. Whilst Laski’s qualitative assessment is that these figures are evidence of no significant
change such aggregate statistics may be misleading for two main reasons.
First this was a period of dramatic change in British society with rapid industrial and urban expansion. It is possible that even the small aggregate changes noted by Laski were incidental to
the three Reform Acts. Moreover, as we show, the effect of the Second Reform Act of 1867, in
particular, was varied at the constituency level: urban boroughs were, by design, more affected
than the rural districts; and, more importantly, whilst some urban boroughs saw a huge increase in
the franchise others were largely unaffected. In short, in order to isolate the impact of the reform
we need to disentangle its effect upon the decline of the English aristocracy, such as it was, from
underlying trends in British society, and look for the effect of the reform not at the aggregate level
but in the constituencies affected by reform.
4
5
Canandine (1999)
Laski (1928)
3
We thus take a fresh look at the data to explore whether the changing role of the aristocracy in
British politics was causally related to extension of the franchise. Central to our empirical strategy
is exploiting the constituency level variation in the impact of these reforms with respect to the
number of registered voters per district. We match data measuring the constituency level impact of
reforms with a unique data-set that records the background characteristics of British MPs. Doing
so, and deploying relevant econometric techniques that allow us to separate the effect of the reform
from underlying constituency level traits correlated with the voting population, as well as broader
tends, we evaluate Laski’s null hypothesis that franchise extension had no effect on the position of
the aristocracy in British politics.
We explore three channels of influence that may have been affected by the reforms. We first
ask whether there is a direct causal relationship between the impact of franchise extension at the
constituency level and the likelihood that the member of a given constituency enters the cabinet. If
Laski’s claim is correct, and in fact membership of the Cabinet remained a preserve of the landed
classes, and so remained unaffected by the large changes in the voting body, then we should find
no significant relationship.
As noted by Cox (1987), however, the accountability of the Cabinet to the majority in Parliament
was already well established by the end of the 19th century. Even if selection to the cabinet did
not change, outcomes could be different if the principal body changed its composition as a direct
consequence of the reforms. So we explore whether changes in the personnel of Parliament were
causally related to franchise extension. Specifically we ask whether constituencies most affected
(treated) by reform were subsequently more or less likely to elect a member of the aristocracy as a
representative.
A third channel that we explore is the ability of politicians to build political dynasties. Whilst
franchise extension may have had no impact on an elected aristocrats immediate political fortunes
it may have weakened his ability to pass on his electoral seat to his offspring. We ask whether
whether seats that were more affected by reform were subsequently less likely to elect the sons and
daughters of those elected to office.
A major focus in this paper is the Second Reform Act of 1867 that, unlike previous and later
reforms, did not compound franchise extension with other constitutional measures that would confound attempts to isolate its causal impact. For much of our analysis we look at a sample of boroughs where franchise reform was not associated with any change in the constituency boundaries
or the number of parliamentary seats between the general elections of 1865 and 1868. Exploiting
the constituency level variation in the impact of franchise reform in this sample we are able to
isolate its subsequent effect on the 1868 parliament.
4
Our results from the years 1865-1868 for this sample show that, once we condition on fixed aspects
of a constituency, there is no direct causal relationship between the growth of the franchise and
the likelihood that an MP becomes a member of the Cabinet. Moreover, the constituencies that
are most affected by the reforms were not more likely to elect an aristocrat to Parliament than
those less affected by franchise extension. Nor does the data suggest that dynastic succession was
related to the impact of the reforms. These results holds when we adopt a very narrow definition
of aristocracy as in Laski’s paper, namely the sons of men possessing hereditary titles, or a broader
definition of such elites that includes landowners and military and naval officers or rentiers.
As a further check we explore whether franchise extension was related to the wealth of politicians
elected to parliament. Lacking good data on the individual wealth of British parliamentarians
we instead exploit a well understood positive relationship between longevity and socio-economic
status that is confirmed by numerous studies. Specifically we ask whether the age at death of a parliamentarian elected in 1868 is causally related to the increase in his constituency level franchise.
Using age at death this as a proxy for wealth we find no evidence suggesting that constituencies
more affected by reform were more likely to elect MPs of lower socioeconomic status.
This analysis thus suggests that Harold Laski was correct, at least with respect to the Second
Reform Act. In terms of the personnel of the British parliament and its executive body the reform
act had no effect. Policies may have changed- this was an era of substantial Liberal reform - but
this was not because parliament was less representative of established elites. In fact the newly
enfranchised electorate were no less inclined to elect such elites to office.
The picture is slightly more complicated, however, when using a less restrictive sample. Then,
when using the broadest definition of aristocracy, we find a nonlinear relationship between franchise extension and election of elites that depends on whether changes to the voting rules were
compounded by differences in the constituency boundaries. Thus constituencies that were hit
by two simultaneous shocks– the inclusion of a lower income tranche of voters and boundary
changes–were less likely to elect aristocrats to office. A conjecture is that the average characteristics of the marginal voter in such districts was different to those of the same voter in a district only
affected by franchise reform.
Next we allow a longer time frame for the reforms to have an effect. We ask whether there was an
impact of the second reform act on those elected to office in 1874. Using the less restrictive sample
we do find a direct effect suggesting that aristocrats were less likely to be elected in constituencies
most affected by reform. This finding is consistent with the conjecture that new voters were less
likely to select such candidates, and/or that aristocrats were less willing to stand for election in
such constituencies. We also find some evidence of a weakening of dynastic effects. We can not
make any strong causal claims, however, due to the introduction of the secret ballot in 1872 that
confounds our analysis.
5
2. E LECTORAL R EFORM
AND
T HE S ECOND R EFORM ACT
Elections in Britain in the Victorian period under investigation took place under the first-past-thepost voting system that is still in place. Whilst some constituencies were single-member districts,
most constituencies elected two candidates and a few elected three and four. The constituency elections were contested by candidates who aligned with one of two major parties, the Conservatives
and the Liberals. The Liberals brought together a loose coalition of Whigs, Radicals, and Peelites
(a faction that had broken from the Conservatives) and by 1860 formed a cohesive parliamentary
block. Following Lord Palmerston’s death in 1865 the Liberals were lead by William Gladstone.
For the immediate period preceding the elections of our investigation, Liberals had held the key
ministries of government. However, between 1865 and 1868 the Conservatives formed a minority
government, first under Lord Derby and then under Benjamin Disraeli.
The Representation of the Peoples Act, otherwise known as the Second Reform Act, was passed
by Parliament on August 15th, 1867. As its name suggests it was the second major voting reform
bill that transformed the political landscape in the Great Britain. The first major extension of the
franchise in the UK took place in 1832. The Great Reform Act of that year introduced several
measures that mitigated malaportionment: increasing representation in the industrialized cities,
and taking away seats from the so-called “rotten boroughs” with small voting populations. The act
also increased the male franchise to around 650,000.
The Second Reform Act, that became law in England and Wales in 1867, extended the franchise in
the boroughs to all males over the age of 21 who were inhabitant occupiers, whether house-owners
or tenants, and to male lodgers whose rent was at least £10 per year. A residence of at least one
year in the borough was required and women were still unable to vote. In counties, the franchise
was extended to holders of life interests, copyholds and leases of sixty years and more worth £5
per annum (from a previous threshold of £10) and to tenants occupying land worth £12 (from a
previous threshold of £50 per annum). The Reform Act for Scotland was delivered in 1868.6 The
reforms were swiftly followed by the election of a new parliament in 1868. This provides us with
a unique window of opportunity for identifying the impact of franchise extension. Later Reform
Acts introduced the secret ballot (1872), placed the counties on an equal footing with the urban
boroughs (1884), reduced the number of multiple member districts (1885), extended the franchise
to all males (1918), and provided women with electoral equality (1928).
6
Ireland had a reform act in 1868 but unlike in England, Scotland and Wales the impact of the reform on registered
voters was marginal.
6
TABLE 1. Number of registered electors by
year county and location
Year
England
Wales
Boroughs Counties
1859
900,084
56,033
439,472
516,645
1865
970,436
61,656
489,146
542,946
1868
1,881,508 127,385 1,216,457 792,436
1874
2,099,402 137,143 1,408,663 827,882
a
Source, Craig (1989), British Parliamentary Election Results 1832-1885
F IGURE 1. Registered Electorate/Population in Boroughs and Counties
Table 1 presents the number of registered electors in 1859, 1865, 1868 and 1874 for England,
Wales, and Scotland in both boroughs and counties.7The Reform Act lead to an increase in the
franchise not witnessed either before or after the 1868 election. From 1859 to 1865 the registered
electorate in England, Scotland and Wales increased by 7% and from 1868 to 1874 by 12%; this
compares with a 97% increase between 1865 to 1868.
The new franchise rules allowed the registration of more than 1,000,000 new electors. The increase was more pronounced in the more densely populated urban boroughs where the number of
7
These figures exclude the electors registered in the university constituencies which were allowed to vote both in the
university constituency and in their town of residence.
7
registered electors increased (on average) by 152% with respect to a 47% increase in the counties. Indeed historians have noted that the most striking feature of the Second Reform Act was the
unexpectedly wide extension of the franchise in the boroughs, when compared to the counties.
A key advantage in studying the 1867 reform is not only that that it was unaccompanied by other
constitutional changes that would confound any attempt to isolate the causal impact of reform, but
that its effect was varied at the constituency level. Specifically, the kernel densities for the logged
difference in the registered voters between 1865 and 1868 show a wide degree of variance in the
effect of franchise extension in these constituencies: in some boroughs the changes in the voting
rules had little discernible impact on the number of eligible voters, whereas in others the size
of the (registered) electorate increased considerably. Adopting the language of the experimental
literature, we can view the extension of the franchise as a ‘treatment’ varying in intensity from
(just below) 0 to (just over) 2 with an average around 1.
3. T HE N EW VOTERS AND THEIR R EPRESENTATIVES
As we have seen, the Reform Act brought into the franchise voters from previously unenfranchised
income brackets. Who were the new voters? Mackenzie (1921) and Bowley (1937) estimate the
income of the head of the household at median, quartile and lowest decile of the income distribution
in 1860. Mackenzie (1921) also provides estimates of household budgets for a typical family (man,
wife, and 3 schoolchildren) which include the amount paid for rent. We present this information
in Table 2 which shows that the annual income of a head of household in the upper quartile of
the income distribution was more than £70; typically he was a semi-skilled worker (e.g, a bricklayer) and paid an annual rent in excess of £10.8 The annual income of the head of household in
the lower quartile was around 60 percent of that in the upper quartile; this would typically be the
income of an unskilled worker (e.g., brick-layer laborer) and paid an annual rent in excess of £6.
At the median of the income distribution the rent paid was close to £8. Although the calculations
are (obviously) rough, given the data limitations, they help us to illustrate the type of households
that gained the vote under the new franchise: it is clear that the extension of the franchise gave the
vote to urban unskilled workers.9
However, there were important qualifications. An aim of the Reform Act was to disqualify the
“feckless”. Paupers and those who did not satisfy the residency requirement were not entitled to
vote. In addition, the rules initially excluded those who did not pay their own rates (local property
taxes) from the franchise. Payment of rates was seen as a signal of personal virtue, the possession
of which was deemed a necessary condition of being granted the vote. Prior to the Second Reform
Act, landlords were able to pay rates on their tenants behalf. This system of “compounding” eased
8
At the time 1 pound = 20 shillings and 1 shilling = 12 cents.
Those in the lowest decile of the income distribution (the agricultural laborers) were only enfranchised in 1884.
9
8
administration on levying rates on houses that consisted of separate units the occupiers of which
paid rents to a landlord. Hodgkinson’s amendment to the original bill, in principle, abolished the
system of compounding - henceforth, those eligible to vote were required to pay their own rates
directly- but it did so only for occupiers who rented a house rather than a room or apartment.
Thus, whereas some potential voters were excluded directly, the partial abolition of compounding
introduced of ambiguity into the system that could only be resolved by local magistrates.The net
effect was what Davis and Tanner have called a “paradoxical and capricious system, still literally
exclusive in that it denied many men the vote, but not selective in the way that its creators, anxious
to limit the vote to those sharing Victorian norms of respectability, had intended.”
Historians have then been concerned with the extent to which the Second Reform Act lead to a
system of one-man-one-vote in the United Kingdom. A form of compounding was reintroduced
by the Liberals in 1869 under the Poor Rate Assessment and Collection Act thereby effectively
enfranchising those who were disadvantaged by Hodgkinson’s amendment. Nevertheless, and
despite these concerns, when looking at the number of voters registered between 1859 and 1874
the large discontinuity occurs between 1865 and 1868. The subsequent increase between 1868
and 1874 is dwarfed by that between 1865 and 1868. Moreover it is of a similar magnitude to the
increase in the registered electorate that occurred between 1859 and 1865.
4. DATA
Our dataset reports individual characteristics for members of parliament between 1832 and 2005.
The personal information about MPs in this dataset relies on biographical sources, most importantly Stenton and Lees (1976, 1978) Who’s Who of British Members of Parliament.10 Information is available about each MP’s parliamentary service, aristocratic connections, educational
background, occupation and dynastic links to other MPs. We consider all members of parliament
elected in general elections.
We investigated how close the sample of MPs covered by the dataset resembles the full universe
of MPs after 1832, by checking these individuals against a full list of elected MPs derived from
Craig (1971, 1977, 1989) up to 1918, and afterwards we used the information from Craig to check
whether the correct number of individuals was returned for each constituency. For each individual
that we found to be missing in our dataset, we collected additional information about parliamentary
service, aristocratic connections and dynastic links from Stenton and Lees (1976, 1978). Our final
sample contains 9,043 individuals or 28,362 electoral records.
Next, we identified the cabinet ministers in this sample using a list of cabinet ministers building
on Cook and Keith (1975), Butler and Butler (2000) and Berlinski, Dewan and Dowding (2012).
10
The dataset was collected and provided by Professor Michael Rush.
9
We exclude cabinet ministers if they were not at least once elected in a general election. Hence,
we exclude the non-elected peers because we have no individual information about them, except if
those peer cabinet ministers were elected at any point before being elevated to the peerage.11 We
identified 506 cabinet ministers or 2,923 cabinet ministerial terms.
The individual information about MPs, including their family connections to other MPs, relies
on biographical information. We exploited all information about dynastic links available in the
dataset and ensured that if person A is reported to have a dynastic link to person B, person B
is also reported to have the inverse dynastic link to person A.12 The dynastic links refer only to
other elected MPs, but the links can be to any MP, elected in general elections, by-elections or
petitions. The dataset does not report the links between MPs and peers explicitly. However, we
have implicit information about such links because the data reports whether an MP was connected
to the aristocracy. An aristocratic connection indicates that the individual was the son, grandson
or nephew of the holder of a hereditary peerage or baronetcy (or similar for women after 1918).13
We also know for the MPs whether they succeeded to a peerage. Tables 2 and 3 below present the
differences in the key characteristics we study before and after the franchise extensions. Finally, to
measure the effect of subsequent treatments of franchise extension, we merged our data to the size
of the electorates and population in each constituency building on Berlinksi and Dewan (2011).14
5. E MPIRICAL S TRATEGY: I DENTIFYING THE E FFECT OF F RANCHISE E XTENSION ON
P OLITICAL O UTCOMES
To understand how political outcomes are causally related to the change in rules governing the
eligibility to vote, we need to isolate the effect of a change in the franchise from other possibly
confounding factors. Although franchise extension was applied nationally and simultaneously in
all constituencies, the magnitude of the change at the constituency level reflects local conditions.
In particular the local impact of a change in the electoral law is related to the constituency level
11
If a peer member of cabinet was elected before, we know the person’s background characteristics, but we do not
if the peer member of cabinet only ever served as a peer in parliament. However, we should not expect to see much
variation in terms of personal characteristics between the peer members of cabinet (who never served as MPs), as the
unelected peers essentially share an aristocratic, landowner background.
12
The biographical entries used to create the dataset for persons A and B would not commonly report these reciprocal
links. Consider an example of person B, who is the son of person A. While the biography of person B would indicate
he is the son of person A, the biographical entry for person A would not necessarily indicate that he is the father of
person B, for example because person B had not yet served in parliament at the time of writing. Therefore, we ensured
that all available one-way dynastic indications between two individuals were transformed into reciprocal links in the
dataset.
13
Aristocratic connections are defined as the son, grandson, or nephew of the holder of a hereditary peerage or baronetcyi.e. hereditary knighthoodwith a concomitant extension to women MPs from 1918. (The Role of the Member of
Parliament Since 1868: From Gentlemen to Players - The Socio-Economic Transformation of the House of Commons,
p31)
14
Sources: Craig (1989, a,b) and Vincent and Stenton (1971)
10
TABLE 2. Distribution of Characteristics, 1865 and 1868
Boroughs
Counties
All
1865 1868 1865 1868 1865 1868
%
Cabinet
Aristocracy
Elite
Rentier
Junior
Senior
Missing
4
28
43
28
31
39
13
2
27
37
23
28
35
10
3
46
78
45
42
51
6
1
51
74
47
49
47
10
3
35
58
35
35
44
10
2
37
53
34
37
40
10
Mean
Age
Age at death
49
72
49
72
48
71
46
71
49
72
47
72
Nr of MPs
394
367
264
291
658
658
TABLE 3. Distribution of Characteristics, 1880 and 1885
Boroughs
Counties
All
1880 1885 1880 1885 1880 1885
%
Cabinet
Aristocracy
Elite
Rentier
Junior
Senior
Missing
3
23
29
16
32
31
11
3
22
23
11
27
30
7
3
41
55
34
42
37
10
2
23
28
15
27
25
9
3
31
41
24
36
33
11
2
22
25
13
27
28
8
Mean
Age
Age at death
49
72
47
74
50
72
50
72
50
72
49
73
Nr of MPs
361
347
291
323
652
670
distribution of income and housing in 1867. In estimating the causal effect of franchise expansion
on political outcomes we face the problem that the change in franchise is systematically related to
a set of constituency level characteristics that are likely to have an independent effect on political
outcomes. Without controlling for these confounding factors, our estimates are likely to be biased
and inconsistent.
11
As long as this differential impact of the reform on franchise levels is driven by community characteristics that are fixed overtime (or that vary slowly), we can measure the impact of franchise
extension by comparing the differences in outcomes between communities where franchise levels vary by different amounts. Thus, we estimate the following benchmark model for MP i in
constituency j at time t:
Yijt = α + βLog(Rjt ) + λt + δj + ǫijt
(1)
where Yijt is one of our outcomes of interest; Log(Rjt ) is the log of constituency level registered
voters; λt is a set of time dummies; δj is a constituency fixed effect and, finally, ǫijt is a random
error term.
If the franchise level is the ratio of registered voters to the relevant constituency population then,
provided that the population remains fixed or its change is uncorrelated with changes in Rjt , equation (1) is similar to regressing Yijt on the log franchise level. Because this may not be the case
we also present estimates where we control for the the local population by including the (logged)
population, Log(Pjt ). Including this term we then estimate
Yijt = α + βLog(Rjt ) + γLog(Pjt ) + λt + δj + ǫijt
(2)
The parameter of interest in equation 1 and 2 is β, the causal effect of changes in the franchise on
electoral outcome
A key advantage in studying the 2nd Reform Act of 1867 is that its effect, varied for different
constituencies, was unaccompanied by other potentially confounding constitutional changes. For
comparison, we provide similar estimates for the changes around the 3rd Reform Act of 1885. The
changes of the 3rd Reform Act affected mainly the counties. However, the franchise extensions of
the 3rd Reform Act were accompanied by other electoral changes. In 1884 many multi-member
constituencies were split, or merged with others, resulting in more single-member constituencies.
If we exclude all constituencies with such changes, we should be able to distinguish the effect
of the franchise extension only. However, doing so decreases the sample size substantially. For
comparison, we also present results for the sample of constituencies that never changed boundaries
for both reforms.15
15
We also checked robustness of these results for a separate sample where constituencies are merged at the highest
possible level across time, see appendix.
12
6. R ESULTS
R EFERENCES
Aidt, T. S. and P. S. Jensen (2009). The taxman tools up: an event history study of the introduction
of the personal income tax. Journal of Public Economics 1-2, 160–175.
Berlinski, S. and T. Dewan (2011). The political consequences of franchise extension: Evidence
from the second reform act. Quarterly Journal of Political Science 6(3–4), 329–376.
Bowley, A. (1937). Wages and Income in the United Kingdom since 1860. Cambridge: Cambridge
University Press.
Canandine, D. (1999). The Decline and Fall of the British Aristocracy. Vintage.
Cox, G. (1987). The Efficient Secret: The Cabinet and the Development of Political Parties in
Victorian England. Cambridge: Cambridge University Press.
Larcinese, V. (2011). Enfranchisement and representation: Italy 1909-1913. unpublished paper.
Laski, H. (1928). The personnel of the english cabinet, 1801-1928. Americal Political Science
Review 22(1).
Mackenzie, W. (1921, October). Changes in the standard of living in the uk, 1860-1914. Economica (3), 211–230.
Pande, R. (2003). Can mandated political representation provide disadvantaged minorities policy
influence? theory and evidence from india. American Economic Review 93(4).
Berlinksi, S. and Dewan, T. (2011) The Political Consequences of Franchise Extension: Evidence
from the Second Reform Act, Quarterly Journal of Political Science, 6 (4): 329-376
Berlinski, S. Dewan, T. and Dowding, K. (2012) Accounting for Ministers: Scandal and Survival
in British Government 1945-2007. Cambridge: Cambridge University Press
Butler, D. and Butler, G. (2000) Twentieth Century British Political Facts, 19002000, 8th edition.
London: Macmillan
Cook, C. and Keith, B. (1975) British historical facts, 1830-1900. London: Macmillan
Craig, F. (1971) British Parliamentary Election Results, 1950-1970. Chichester: Parliamentary
Research Services.
Craig, F. (1983) British Parliamentary Election Results, 1918-1949, 3rd edition. Chichester: Parliamentary Research Services.
Craig, F. (1984) British Parliamentary Election Results, 1974-1983. Chichester: Parliamentary
Research Services.
Craig, F. (1989) British Parliamentary Election Results, 1832-1885, 2nd edition. Aldershot: Parliamentary Research Services. (a)
13
Craig, F. (1989) British Parliamentary Election Results, 1885-1918, 2nd edition. Aldershot: Parliamentary Research Services. (b)
Rush, M. (2001) The Role of the Member of Parliament Since 1868: From Gentlemen to Players.
Chapter 5: The professionalisation of the Member of Parliament, p109-139. Oxford: Oxford
University Press
Stenton, M. and Lees, S. (1976) Who ’s Who of British Members of Parliament: a biographical
dictionary: Volume I, 1832-1885. Hassocks : Harvester Press
Stenton, M. and Lees, S. (1978) Who’s who of British Members of Parliament : a biographical
dictionary: Volume II, 1886-1918. Hassocks : Harvester Press
Stenton, M. and Lees, S. (1979) Who’s who of British Members of Parliament : a biographical
dictionary: Volume III, 1919-1945. Hassocks : Harvester Press
Stenton, M. and Lees, S. (1981) Who’s who of British Members of Parliament : a biographical
dictionary: Volume IV, 1945-1979. Hassocks : Harvester Press
Vincent, J. and M. Stenton (1971). McCalmonts Parliamentray Poll Book, 8th edition. Harvester.
-0.12
[0.083]
-0.617
[1.539]
330
0.627
0.024
[0.089]
0.008
[1.561]
342
0.521
Yes
• log(electorate)
Observations
R-Squared
Control Population Change?
Yes
330
0.59
-0.09
[0.080]
-2.253
[1.888]
(3)
Rentier
Yes
340
0.618
3.323
[3.164]
45.218
[52.950]
(4)
Age at Death
Yes
342
0.274
-0.001
[0.038]
0.918
[1.863]
(5)
Cabinet Min
Source: Craig (1989) and Vincent and Stenton (1971).
Notes: OLS regressions. All regressions include a constant term. Robust standard errors are in parentheses.
Yes
(2)
(1)
Constant
Elite
Aristocrat
Table 4: Changes in Characteristics of Elected Candidates Under the New Franchise (1868-1865)
Yes
342
0.531
-0.018
[0.111]
1.536
[1.325]
(6)
Junior
Yes
342
0.574
-0.195
[0.151]
-2.177
[2.535]
(7)
Senior
183
0.442
190
0.548
Yes
Observations
R-Squared
Control Population Change?
Yes
183
0.469
0.334
[0.295]
617
0.596
-0.062
[0.071]
0.348
[0.505]
-2.111*
[1.272]
-0.053
[0.062]
0.238
[0.146]
-0.087
[1.030]
0.048
[0.122]
488
0.583
-0.067
[0.078]
0.369
[0.509]
-0.056
[0.063]
(3)
Rentier
Yes
190
0.542
1.062
[7.042]
641
0.549
-0.035
[2.256]
-24.281*
[13.894]
-51.946
[59.886]
2.889*
[1.658]
6.207
[6.944]
43.275
[38.085]
-5.265
[4.880]
502
0.566
-0.084
[2.434]
-24.017*
[13.919]
2.854*
[1.664]
(4)
Age at Death
Yes
190
0.256
0.036
[0.045]
643
0.364
-0.044
[0.041]
-0.476
[0.339]
1.448
[1.412]
0.058
[0.041]
-0.167
[0.163]
1.696
[1.538]
-0.215
[0.192]
504
0.337
-0.033
[0.040]
-0.478
[0.338]
0.058
[0.041]
(5)
Cabinet Min
Yes
190
0.49
0.017
[0.440]
643
0.531
-0.055
[0.072]
-0.213
[0.460]
-0.003
[1.915]
0.028
[0.052]
0.026
[0.225]
-0.74
[1.554]
0.063
[0.189]
504
0.555
-0.065
[0.076]
-0.138
[0.458]
0.018
[0.052]
(6)
Junior
Yes
190
0.435
0.082
[0.429]
643
0.567
-0.128
[0.099]
0.001
[0.557]
-1.719
[2.501]
0.012
[0.064]
0.221
[0.288]
-1.71
[1.476]
0.192
[0.181]
504
0.577
-0.138
[0.104]
0.07
[0.564]
0.003
[0.065]
(7)
Senior
Source: Craig (1989) and Vincent and Stenton (1971).
Notes: OLS regressions. All regressions include a constant term. Regressions in Panel B also control for the difference in the number of seats. Robust standard errors are in parentheses.
Yes
0.164
[0.342]
617
0.617
643
0.538
0.475
[0.308]
-0.039
[0.070]
0.739*
[0.379]
-2.333
[1.507]
-0.100**
[0.046]
0.275
[0.174]
-0.904
[0.769]
0.146
[0.092]
488
0.599
504
0.534
-0.028
[0.063]
0.336
[0.481]
0.254
[1.508]
-0.053
[0.055]
-0.019
[0.176]
-1.429
[0.971]
0.163
[0.119]
-0.058
[0.073]
0.750*
[0.390]
-0.102**
[0.047]
-0.027
[0.066]
0.378
[0.483]
-0.059
[0.055]
(2)
(1)
• log(electorate)
Panel C: Counties without boundary or seat changes
Observations
R-Squared
• log(electorate) X Change Boundaries X Change Seats
Change Boundaries X Change Seats
• log(electorate) X Change Seats
• log(electorate) X Change Boundaries
Change Seats
Change Boundaries
• log(electorate)
Panel B: Boroughs with boundary changes and change of seats
Observations
R-Squared
• log(electorate) X Change Boundaries
Change Boundaries
• log(electorate)
Panel A: Boroughs with boundary changes
Elite
Aristocrat
Table 5: Changes in Characteristics of Elected Candidates Under the New Franchise (1868-1865). Different Samples.
-0.084
[0.117]
328
0.53
-0.11
[0.120]
342
0.45
Yes
Observations
R-Squared
Control Population Change?
Yes
328
0.516
-0.027
[0.101]
(3)
Rentier
Yes
340
0.48
-0.198
[3.294]
(4)
Age at Death
Yes
342
0.288
-0.084
[0.063]
(5)
Cabinet Min
Source: Craig (1989) and Vincent and Stenton (1971).
Notes: OLS regressions. All regressions include a constant term. Robust standard errors are in parentheses.
Yes
(2)
(1)
• log(electorate)
Elite
Aristocrat
Table 6: Changes in Characteristics of Elected Candidates Under the New Franchise (1874-1865)
Yes
342
0.419
0.018
[0.119]
(6)
Junior
Yes
342
0.51
-0.25
[0.157]
(7)
Senior
185
0.428
190
0.478
Yes
Observations
R-Squared
Control Population Change?
Yes
185
0.43
0.257
[0.313]
618
0.529
-0.083
[0.082]
-0.045
[0.634]
-1.074
[2.836]
-0.008
[0.075]
0.123
[0.319]
0.184
[2.618]
-0.015
[0.334]
488
0.508
-0.073
[0.086]
-0.01
[0.632]
-0.012
[0.075]
(3)
Rentier
Yes
190
0.438
17.533**
[7.277]
641
0.457
-2.229
[2.132]
-28.110**
[11.439]
-89.894*
[54.130]
3.303**
[1.367]
10.075
[6.168]
-3.08
[41.649]
1.393
[5.205]
502
0.46
-1.065
[2.247]
-26.778**
[11.480]
3.131**
[1.376]
(4)
Age at Death
Yes
190
0.418
0.001
[0.031]
643
0.379
-0.072*
[0.041]
-0.235
[0.283]
0.034
[0.488]
0.026
[0.032]
-0.003
[0.057]
-0.007
[0.616]
0.002
[0.068]
504
0.338
-0.055
[0.041]
-0.231
[0.283]
0.025
[0.032]
(5)
Cabinet Min
Yes
190
0.46
0.411
[0.347]
643
0.442
0.007
[0.087]
-1.117*
[0.648]
2.08
[1.768]
0.111
[0.074]
-0.238
[0.203]
1.845
[1.478]
-0.228
[0.185]
504
0.445
0
[0.088]
-1.005
[0.638]
0.098
[0.073]
(6)
Junior
Yes
190
0.448
0.004
[0.319]
643
0.504
-0.208**
[0.103]
-1.160*
[0.682]
-0.011
[2.652]
0.131*
[0.079]
0.011
[0.297]
-0.319
[1.569]
0.036
[0.196]
504
0.508
-0.221**
[0.104]
-1.065
[0.678]
0.12
[0.078]
(7)
Senior
Source: Craig (1989) and Vincent and Stenton (1971).
Notes: OLS regressions. All regressions include a constant term. Regressions in Panel B also control for the difference in the number of seats. Robust standard errors are in
parentheses.
Yes
-0.076
[0.386]
618
0.527
643
0.469
0.306
[0.410]
-0.144
[0.099]
-0.716
[0.778]
-0.28
[2.922]
0.063
[0.092]
0.032
[0.329]
0.586
[2.743]
-0.056
[0.344]
488
0.499
504
0.472
-0.162**
[0.078]
-0.483
[0.543]
1.649
[1.777]
0.037
[0.063]
-0.184
[0.201]
1.133
[1.429]
-0.132
[0.180]
-0.137
[0.103]
-0.677
[0.775]
0.058
[0.092]
-0.183**
[0.080]
-0.436
[0.542]
0.032
[0.063]
(2)
(1)
• log(electorate)
Panel C: Counties without boundary or seat changes
Observations
R-Squared
• log(electorate) X Change Boundaries X Change Seats
Change Boundaries X Change Seats
• log(electorate) X Change Seats
• log(electorate) X Change Boundaries
Change Seats
Change Boundaries
• log(electorate)
Panel B: Boroughs with boundary changes and change of seats
Observations
R-Squared
• log(electorate) X Change Boundaries
Change Boundaries
• log(electorate)
Panel A: Boroughs with boundary changes
Elite
Aristocrat
Table 7: Changes in Characteristics of Elected Candidates Under the New Franchise (1874-1865). Different Samples.
(2)
-0.341
[0.388]
47
0.88
(1)
-0.365
[0.372]
50
0.863
Yes
• log(electorate)
Observations
R-Squared
Control Population Change?
Yes
47
0.952
-0.208
[0.306]
(3)
Rentier
Yes
49
0.768
-5.141
[7.936]
(4)
Age at Death
Yes
50
1
(5)
Cabinet Min
Source: Craig (1989) and Vincent and Stenton (1971).
Notes: OLS regressions. All regressions include a constant term. Robust standard errors are in parentheses.
Yes
Elite
Aristocrat
Table 8: Changes in Characteristics of Elected Candidates Under the New Franchise (1880-1885)
Yes
50
0.886
-0.229
[0.307]
(6)
Junior
Yes
50
0.961
-0.203
[0.293]
(7)
Senior
264
0.583
276
0.597
Yes
Observations
R-Squared
Control Population Change?
Yes
264
0.524
-0.103
[0.080]
434
0.396
-0.05
[0.146]
-1.199
[2.041]
7.452***
[2.446]
0.08
[0.217]
-0.862***
[0.281]
407
0.397
-0.075
[0.149]
-1.124
[2.015]
0.069
[0.213]
(3)
Rentier
Yes
276
0.517
0.706
[1.989]
443
0.25
0.442
[4.674]
-65.958
[77.862]
2.826
[141.514]
8.139
[8.024]
0.54
[15.630]
415
0.239
-0.12
[4.760]
-77.505
[73.045]
9.398
[7.628]
(4)
Age at Death
Source: Craig (1989) and Vincent and Stenton (1971).
Notes: OLS regressions. All regressions include a constant term. Regressions in Panel B also control for the difference in the
number of seats. Robust standard errors are in parentheses.
Yes
-0.122
[0.081]
434
0.425
445
0.352
0.012
[0.067]
-0.138
[0.129]
-3.296
[3.274]
4.852*
[2.711]
0.308
[0.356]
-0.582*
[0.315]
407
0.418
417
0.324
-0.118
[0.130]
-2.135
[2.164]
4.180**
[1.785]
0.212
[0.231]
-0.495**
[0.214]
-0.161
[0.131]
-3.36
[3.246]
0.242
[0.356]
-0.124
[0.133]
-2.038
[2.126]
0.209
[0.247]
(2)
(1)
• log(electorate)
Panel C: Boroughs without transformations or seat changes
Observations
R-Squared
• log(electorate) X Change Seats
• log(electorate) X Transformed
Change Seats
Transformed
• log(electorate)
Panel B: Counties including transformed and change of seats
Observations
R-Squared
• log(electorate) X Transformed
Transformed
• log(electorate)
Panel A: Counties including transformed
Elite
Aristocrat
Table 9: Changes in Characteristics of Elected Candidates Under the New Franchise (1880-1885). Different Samples.
Yes
276
0.538
-0.057*
[0.032]
445
0.321
0.049
[0.057]
0.674
[0.484]
-0.174
[0.240]
-0.073
[0.051]
0.021
[0.029]
417
0.32
0.051
[0.059]
0.706
[0.492]
-0.068
[0.045]
(5)
Cabinet Min
Yes
276
0.599
-0.045
[0.079]
445
0.316
-0.056
[0.171]
-1.051
[2.283]
0.626
[2.652]
0.071
[0.242]
-0.109
[0.304]
417
0.294
-0.051
[0.173]
-0.542
[2.357]
0.032
[0.268]
(6)
Junior
Yes
276
0.635
-0.032
[0.074]
445
0.335
0.071
[0.153]
1.425
[2.581]
-6.042***
[1.907]
-0.211
[0.278]
0.628***
[0.223]
417
0.318
0.05
[0.152]
1.766
[2.533]
-0.179
[0.283]
(7)
Senior