An introduction to how many people could, and did, vote in the eighteenth century
[5-minute read]
The ‘electorate’ is the group of individuals who were entitled to vote in an election. This is different from the number of people who actually cast their votes at a poll, a group which we might call the ‘voterate’. Surviving poll books can tell us how many people actually cast their votes, but this number would very rarely match the total electorate. Old age, ill health, poor weather, distance, intimidation, or downright indifference could all prevent a voter from appearing at the hustings. Or an election might be abandoned by the trailing candidates before it had ‘polled out’.
It is notoriously difficult to calculate the size of the electorate for eighteenth-century constituencies. The confusing array of parliamentary franchises which determined who could legally vote in each constituency were often opaque or based on unstable qualifications like membership of a corporate body or personal wealth. It was not until the Reform Act of 1832 that the systematic registration of voters was introduced.
Historians have estimated that the total English and Welsh electorate grew from about 200,000 in 1689 to 366,000 on the eve of the Reform Act.[1] But English electorate sizes ranged widely from constituency to constituency. In the mid-eighteenth century, there were two qualified voters in the decayed borough of Gatton, compared to 20,000 of Yorkshire or 12,000 of Westminster. The ‘electorate’ was rarely an unchanging, or even a stable, entity – the number of qualified electors shifted from election to election.
Some constituencies experienced little or no change in their electorate size across the long eighteenth century. Corporation boroughs (where voting was confined to the borough’s governing body) were the most stable. Bath, for example, had an electorate of 30 throughout this whole period. Burgage boroughs were also theoretically static since each vote was tied to a piece of property. In practice, though, these properties could be ‘split’ or conveyed to create extra votes. The number of voters in Pontefract, for instance, grew from about 130 in 1700, to over 250 in 1715, before falling to 130 again by 1774.
The county constituencies and more open boroughs, with householder, scot and lot, or freemen franchises, were more likely to experience significant increases in their electorates. In part, this was a symptom of population growth. The population of England grew rapidly during this period, from around 4.9 million people in 1690, to 8.3 million in 1801, to 11.5 million by 1820. At the same time, the urban population expanded as many people sought out new work in England’s towns and cities.
At other times, however, their electorates were expanded artificially. Freemen boroughs sometimes enrolled huge numbers of new freemen ahead of an election to boost the number of votes for particular candidates. The corporation of Liverpool were accused of admitting more than 380 freemen ahead of the 1734 election, and similar allegations were made in Coventry in 1780 and Bedford in 1774. That said, some of this enrolment was perfectly legitimate, and it was simply good electioneering to encourage potential supporters to qualify for a vote ahead of an election; E. M. Menzies has found no evidence of wrongdoing surrounding the 320 Liverpool freemen admitted before the 1802 election (nor the 480 in 1812 or 450 in 1816).[2]
By contrast, electoral patrons sometimes attempted to engineer a reduction of a constituency’s electorate in order to cement their political control or reduce the cost of an election. At Minehead, the Luttrell family were accused of deliberately allowing the town’s buildings to slide into dilapidation so that they would no longer serve to qualify votes – a decline which was exacerbated by the ‘Great Fire’ of 1791. In Grampound, Christopher Hawkins and Lord Edgcumbe hatched a plot to disqualify 40 or 50 freemen in 1748.
With electorates so uncertain and changeable, it is very difficult to calculate ‘turnout’ (i.e. the proportion of eligible voters who actually cast votes at an election). Sometimes freemen rolls, rate books, or land tax returns can be used to estimate electorates for particular constituencies – or election literature (including poll books) provide figures for the total electorate. For the ECPPEC project’s case study constituencies, the following turnouts can be estimated:
| Constituency | Year | Estimated Electorate | No. of Voters | Turnout |
| Bedford | 1830 | 1,100 | 907 | 82% |
| Bedfordshire | 1831 | 2,800 | 1,869 | 67% |
| Bristol | 1754 | 4,992 | 4,738 | 95% |
| Bristol | 1830 | 7,000 | 6,384 | 91% |
| Cambridge University | 1710 | 418 | 260 | 62% |
| Cambridge University | 1727 | 435 | 345 | 79% |
| Cambridge University | 1734 | 498 | 386 | 78% |
| Cambridge University | 1780 | 753 | 546 | 73% |
| Cambridge University | 1784 | 735 | 588 | 80% |
| Cambridge University | 1790 | 865 | 684 | 79% |
| Cambridge University | 1806 | 1,008 | 609 | 60% |
| Cambridge University | 1807 | 983 | 631 | 64% |
| Coventry | 1761 | 2,500 | 1,596 | 64% |
| Coventry | 1826 | 3,500 | 2,925 | 84% |
| Grampound | 1818 | 69 | 56 | 81% |
| Liverpool | 1802 | 2,750 | 1,429 | 52% |
| Liverpool | 1820 | 3,500 | 2,173 | 62% |
| Liverpool | 1830 | 5,331 | 4,436 | 83% |
| London | 1710 | 8,200 | 6,638 | 81% |
| London | 1727 | Unknown | Unknown | – |
| London | 1796 | 9,600 | Unknown | – |
| London | 1826 | 12,000 | Unknown | – |
| Northumberland | 1826 | 4,000 | 2,984 | 75% |
| Oxfordshire | 1831 | 3,500 | 2,922 | 83% |
| Pontefract | 1826 | 820 | 744 | 91% |
| Westminster | 1820 | 13,000 | Unknown | – |
These figures suggest an average turnout of 76% (lower than the 81% found by Frank O’Gorman[3]). This compares favourably with modern UK elections, where turnout at general elections fell from about 74% in the 1980s to less than 69% in the 2010s. It suggests a high degree of political engagement in eighteenth-century elections amongst the electorate, and reveals an impressive level of mobilisation of voters by candidates and their agents.
However, these numbers should be treated with caution. Many of the estimated electorates represent calculated guesswork and may dramatically over-or-under represent the true figure. In 1818, for example, The Times reported that the number of ‘real voters’ in London stood at about 8,000, rather than the 10,000-plus figure often quoted. It should also be remembered that most eighteenth-century elections went uncontested, and so denied the electorate the opportunity to exercise their franchise. It is striking, when considering the number of voters recorded in poll books, that some of the constituencies with the largest electorates (Yorkshire, Westminster, Bristol) rarely polled close to their full number. The Westminster election of 1784 famously continued for 40 days, with 12,300 voters having polled, and is often said to have ‘polled out’. But the number of voters barely topped 10,000 at any other Westminster election during the century, falling to as low as 3,000 in 1796.
The two university constituencies present a special case. The electorate was composed of members of each university’s governing body (the Senate at Oxford; Convocation at Cambridge), which meant anyone with a Batchelor degree which had been promoted to a Master’s degree (as was normal) and had kept up with payments of a small annual fee to his college. There was no residential qualification, so in theory the electorate was large, perhaps growing from around 1000 to 2500 in each constituency across the long eighteenth century. In practice, however, active participation in elections tended to be the preserve of resident members in both universities. For instance, only 665 voted in the contest for Oxford in 1806, while 796 votes were admitted at Cambridge in the by-election of 1811.
[1] Frank O’Gorman, Voters, Patrons and Parties: The Unreformed Electorate of Hanoverian England 1734-1832 (Oxford: Clarendon Press, 1989), 179.
[2] E. M. Menzies, ‘The Freeman Voter in Liverpool, 1802–1835’, Historical Society of Lancashire and Cheshire Transactions, 4 (1972), 83-107 (89).
[3] O’Gorman, Voters, Patrons, and Parties, 183.
