Pairs and Proxies

P is for Pairs and Proxies, both ways of ensuring that individual members of Parliament can be absent without affecting the outcome of any vote, and while P is not for substitutes, since these can be aimed at achieving a similar end, we might as well deal with them here too. The current interest in pairing and proxy voting for the House of Commons stems from the developing discussion about how Parliament can be made truly inclusive, and particularly to overcome the difficulties encountered by pregnant women and co-parenting and caring MPs in fulfilling their roles as Members. It has been further stimulated by the row in July over the breakdown – whether deliberate or accidental – of pairing arrangements for Jo Swinson MP, whose son had been born only a few days before, during a crucial vote over the EU (Withdrawal) Bill.


The system of pairing was described in some detail by Thomas Erskine May in the first edition of his Treatise on Parliamentary Procedure in 1844:

‘It has been shown, that no member can vote unless he be present when the question is put; and no sanction has ever been given, by the house, to any custom partaking of the character of delegation. But a system of negative proxies, known by the name of “pairs,” enables a member to absent himself, and to agree with another member that he also shall be absent at the same time. By this mutual agreement, a vote is neutralized on each side of a question, and the relative numbers on the division are precisely the same as if both members were present. The division of the house into distinct political parties facilitates this arrangement, and members pair with each other, not only upon particular questions, or for one sitting of the house, but for several weeks, or even months, at a time. There can be no parliamentary recognition of this practice; and it is therefore conducted privately, by individual members, or arranged by the gentlemen who are entrusted, by the two great parties, with the office of collecting their respective forces on a division’ [p. 221]

May’s reference to delegation – i.e. the assumption that Members acted simply as delegates for and on the instructions of their constituents – is particularly interesting, as the point underlies much past discussion about the admissibility of the practice of proxy voting. The principle was strongly defended that Members should make up their mind which way to vote on the basis of listening to a debate, rather than vote according to the prejudices of the people who had sent them to Parliament. It seems that pairing was generally regarded as acceptable – since, I suppose, it was refraining from voting, rather than voting – though some clearly still disliked it, and perhaps justifiably, given how frivolously some Members used it.

As with many parliamentary practices it is impossible to tell when pairing began.  The History of Parliament volumes for 1660-90 make assumptions about the pairing of members in particular on the Exclusion Bill: where both members for a single double-member constituency were not recorded as voting in divisions on these intensely divisive bills, a guess has often been made that the two Members agreed not to vote (see e.g. Nicholas Eversfield; George Pelham;  Thomas Cradock;  Anthony LowtherSir George Speke; Thomas Penruddock Samuel Sandys;  William Broxholme; Sir William Bassett; John Knight). This is plausible: double member seats would long provide opportunities for complex exercises in political compromise, and their demise in the nineteenth century helped to contribute to the more politically partisan flavour of both electioneering and Parliament thereafter. But it is impossible to substantiate, and it isn’t quite what we currently mean by pairing, anyway, which is more about convenience and covering necessary absence than (as these probably were) keeping one’s head below the parapet — though, of course, it can still be used on occasion for precisely that purpose.

The first hard evidence of the modern practice comes from the diary of Lord Perceval, who mentioned in February 1730 that ‘the lateness of the night obliged about thirty members to leave the House before the question was put, each taking away with him one of the contrary side’, and a couple of years later referred to skipping a vote in company with his cousin John Finch, who was proposing to vote in the other direction in a division on the salt bill. A motion put to the House on 6 March 1744 suggests that some Members strongly disapproved of the practice, but that more found it too convenient to give up. ‘No Member of the House’, it was proposed, should ‘presume to make any Agreement with another Member to absent themselves from any Service of this House, or any Committee thereof; and that this House will proceed with the utmost Severity against all such Members as shall offend therein’. The motion was defeated in a division by 139 votes to 171. It was a large division, but it isn’t clear from the entry in the Journal what particularly provoked the move, though the identity of the tellers, John Yorke and Humphrey Sydenham, implies that it was the work of independent-minded Members with bees in their bonnets, rather than a conspiracy more deeply laid, and its juxtaposition with the deferral of a call of the House may mean that it was related to a discussion of absenteeism. Peter Thomas, in his House of Commons in the Eighteenth Century, suggests that the practice of pairing had become common by the 1770s for election cases, but not for more important debates: he found the first use of the word itself – ‘to pair off’ for a debate – in 1771. Within the next ten to twenty years it became common parliamentary language. Our biography of Sir Henry Fetherstonhaugh , the companion of the Prince of Wales (later Prince Regent) includes a letter to him from the Prince in Brighton concerning an impending vote in 1785:

If you have not already got a pair off in this business I would go to London tonight and endeavour to get one tomorrow morning, which means you will be back again with us in the evening and ready for the terrible work with the high bred cattle [horse races] that is to ensue the next morning. Pray take [Thomas] Onslow with you upon the same errand.

Many others used the pairing system similarly in order to free them from the tediousness of regular attendance, including this ancestor of Sir Anthony Eden . During the Reform debates in 1831, Henry Hunt, complaining of the influence of members of the peerage in some boroughs told how ‘some years ago he went to the lobby of the House of Commons with the late Duke of Norfolk’s agent, Mr Bryant, who met the member for Steyning coming out of the House. Bryant asked where he was going? The answer was, he was pairing off with general Gascoyne; on which Bryant pushed him back into the House, saying it was the Duke’s orders that he should vote for the question before the House’. Hunt said it made him a reformer.

Informal division lists from the 1780s seem sometimes to have listed ‘paired-off’ members, a practice that was carried over to some of the published lists, including those in the Mirror of Parliament, whose listing of divisions in 1834 provided the names of members who were ‘paired off’, and even, in one instance, ‘shut out’ (i.e. arrived just too late for the division), and Hansard, which for a division on 8 May 1837 (col. 689-90)  listed 38 Members who were ‘paired off’ with each other, indicating whether they were for or against.



As May’s text indicated, pairing was a cousin of the ancient system allowed in the Lords of proxies. The principle of the proxy, though, was very different, based on the idea of legal representation of one person by another, often termed a proctor. On this basis, some liked to think of peers as standing in the same relation to their proxies as their constituents stood to Members of Parliament, and therefore dismissed the idea that there was anything remotely illegitimate about the idea of one person casting a vote on behalf of another when the latter had not heard the debate. Benjamin Disraeli, in his guise as constitutional expert, offered the argument to another member of the government, Lord Malmesbury, when the abandonment of the system was under discussion in 1867. Members of Parliament, he claimed were the ‘proxies or proctors of the estate of the Commons’ and therefore ‘the lords, in using proxies, possess and exercise the same privilege as the Commons, no more, and if it is not convenient for them to attend the meetings of their order, they have the right to choose their representatives’.

The practice of representation in assemblies by proxy was a familiar one in the Church before it became recognised in secular councils sometime in the middle ages; in Parliament it seems to have been used by ecclesiastical magnates before it was used by peers. For both, a proxy was originally dependent on a licence from the crown not to attend Parliament in response to a summons: permission to be absent, and a corresponding proxy, were, in other words, in the gift of the king or queen. Exactly when proxies ceased to mean simply the provision of a substitute in order to represent the interests of the absent man, and started to mean the delegation of the exercise of a vote in the assembly to another member of the assembly remains unclear, but it was certainly the case by the early seventeenth century. In 1626 the system was regulated by Standing Order, in particular to limit the number of proxies any peer could hold to two – a change that followed closely on the heels of the king’s favourite, the Duke of Buckingham, becoming the recipient of no less than thirteen of them.

Whether or not it was the case earlier has been the subject of controversy, but it was very clearly the case in the seventeenth century and afterwards that the crown was using proxies to manage the risk of difficult votes in the House of Lords: the government put considerable effort into soliciting proxies from Members who intended to be absent and distributing them among reliable peers. Very occasionally, the use of proxies saved the administration from losing a vote. But sometimes things went awry, as they did in January 1811, after the government lost a vote in committee on an amendment to the Regency bill concerning the powers to be exercised by the Prince Regent during the illness of the king. Proxies could not be used in committee — hence the lost vote — but the amendment had to be confirmed in the House itself, and the government thus hoped to reverse the outcome by using proxies. The House, however, determined in another knife- edge division that proxies could not be used on this occasion. The government were then defeated again on the amendment. Revelling in the victory, the radical peer Lord Stanhope chortled to Lord Holland, ‘This, my dear citoyen is rare fun. If I could have such nights as these, I would live in the House of Lords: it is such high fun’. Stanhope’s use of the French revolutionary term underlined the odd circumstances of the debate and Parliament’s meeting, with the king incapacitated by illness making some of the old assumptions about the operation of the privileges of the peers more difficult to sustain. [See The Times, 24 January 1811]

But proxies continued for another half-century, despite the fact that, as was frequently pointed out, they were quite at variance with the idea, firmly defended in the Commons, that a Member should only cast a vote in a division if he had listened to the debate that preceded it. Alderman Hughes tried to raise the issue in the Commons in March 1832; Thomas Attwood presented a petition of the Birmingham Political Union against them in February 1836; The most substantial debate in the Commons was contrived by Thomas Duncombe on 9 May 1837, when he was backed by Tennyson D’Eyncourt. In response Lord Stanley mocked D’Eyncourt’s complaint that ‘it was intolerable in reasoning to suppose that a deliberative legislative body in the present state of civilization should have such a power of deciding upon legislative acts’ by referring to how pairing had been used in a division the previous evening on D’Eyncourt motion for repeal of the Septennial Act.

According to Stanley, D’Eyncourt had demanded to know ‘How was it possible for any person to give a proxy, unless after attentively listening to the viva voce reasoning on the subject’. Stanley waspishly asked whether D’Eyncourt had consulted the diplomat Henry Bulwer Lytton, MP for Marylebone who bizarrely combined a radical political career with the diplomatic position of head of the British mission in Belgium, and was presumably paired for a substantial amount of the session as a result, ‘as to the extreme impropriety, the unconstitutional course of proceeding, the custom so absurd that it ought not to be tolerated by a civilised people, of the members of the legislative assembly pairing off for two or three hours, for the purpose of refreshment, or the whole of the night, and then coming down to give a decision on a question on which they had not heard one word of argument’. He pointed out in mock sympathy that while D’Eyncourt had been addressing the House the previous night, there were hardly forty Members present, but when it came to the division, he had had what was then a large minority of eighty-seven.

It appeared… that with a view to refresh themselves, hon. Gentlemen paired off last night for two or three hours, and afterwards came down to vote without listening to any portion of the debate; but this he might be told, should be regarded as nothing, because a person by study might have been convinced of the reasonableness of a proposition without hearing the discussion. He admitted the force of this argument, but how a man could consistently give a vote for the whole of the Session on all questions that might become before that House was perfectly incomprehensible to him. And had they never heard in this House of such a practice as pairing off for a week or month; for half a Session, or for even a whole Session?

It was, though, widely felt outside the confines of such partisan debate that the Lords’ practice of proxies was an abuse in a way that pairing wasn’t, quite. Even in the Upper House they came under regular attack. Robert Grosvenor, Marquess of Westminster, one of the richest men in England and one of the most radical peers, on 6 August 1834 called it an ‘anomalous and absurd’ practice, which ‘tended to make their Lordships unpopular in the country’. A discussion of the subject on 31 July 1849 was initiated by Lord Brougham, apparently provoked by three recent occasions when proxies had overturned the result of a division among those present. After 1830 proxies were used much less frequently, perhaps because peers noticed an increasing public antagonism towards the practice. They were deployed in a division in the House for the last time in 1864 (on British policy towards the dispute between Prussia and Denmark), when, unusually, they altered the result.

It was possibly the imminent arrival in the Lords of the Second Reform Bill that finally brought about reform in 1867. Bagehot had ripped the House of Lords apart in the article he published on it in 1866 which became a chapter of The English Constitution: ‘the cure for admiring the House of Lords’, as he quoted someone saying, ‘was to go and look at it’. A leading article in The Times on 12 June 1867, written as the Commons were about to resume debate in committee on the Reform bill, called it a ‘vapid mimicry of a Parliamentary Assembly’: ‘in its soporific atmosphere the keenest intellect grows dull, the brightest eloquence poor, pointless and wearisome, the degeneracy of one generation propagates and exaggerates itself in the next, and if things go on at this rate, the House of Lords is in a fair way to be the wealthiest, the most highly descended, the most useless, and the most indolent Assembly in the world’. Some of the blame – though only some of it – went to the proxy system.

A debate on 20 June initiated by the former Liberal MP and minister Lord Lyveden (whose claim that the Lords ‘never stood higher in public estimation than at present’ flew heroically in the face of the evidence) produced both opponents and defenders of the system; but it also put proxies onto the agenda of a committee being set up to consider the House’s business, and Lyveden onto the committee. He pushed the issue in the committee with remarkable success. It reported in July, that while the right to vote by proxy was a privilege ‘so ancient as to have become a prescriptive right as inherent in their peerages as that of being summoned to Parliament, and one consequently of which no vote of the House alone can deprive them’, it was open to the House to regulate its use. It expressed the view that ‘the exercise of the right of voting by proxy by peers who have had no opportunity of hearing the debate is objectionable. It diminishes the attendance of peers in the House, and tends to weaken in the public mind the authority of the decisions at which the House arrive’. It recommended that the practice should be discontinued: rather than being abolished (which it thought beyond the power of the House alone), a Standing Order should be passed which would prevent proxies being actually used. The change was finally accepted by the House on 31 March 1868 , proposed by Lord Privy Seal the Earl of Malmesbury, to whom Disraeli had explained the history of proxies the year before. The Standing Order (now No. 60) is still there.


There is one other commonly used way of ensuring that the voices of those constituencies are heard if their Member is in some way unable to attend. It is not unusual in some Parliaments, usually those elected on a proportional, party list, system, to employ substitutes, usually from among those who were offered for election in the same list, but did not secure the required number of votes. Surprisingly, a sort of substitute system was tried once, briefly, in the English Parliament, though the mechanism was more akin to a by-election where the seat was not, technically, vacant. It happened in the course of the Parliament elected in 1572, which met for six weeks that year, and was reconvened for similar periods in 1576 and 1581. It was the longest-lasting Parliament until the Long Parliament itself, elected in 1640, exceeded it in 1649. Not unnaturally over such a period there was considerable turnover: those Members who had died were replaced through by-elections, which were already well-established. But in 1581 there was an apparently systematic attempt to introduce substitutes not only for those who were dead, but for those who were sick, or who were abroad (mostly in royal service as diplomatic envoys). The House discussed these on 19 January,

This Day the House being assembled, and the Speaker elect sitting in the Chair, the Matter began to be debated, touching the said Persons newly returned, in Places of others yet living; of whom Question was made the Day before: And the Case opened to the Effect following ; to wit, that there were some Members of this House absent in her Majesty’s Service, as in Embassade, or in her Highness’ Affairs in Ireland, in whose Places new were returned; some again sick of curable Diseases, as Agues, and such like, and new returned in their Places; One, that being the last Session a Burgess for one Borough, is now newly returned a Citizen for another City, in the Place of a former Citizen for the same City, yet living; and another Burgess now newly returned in the Place of the said former Burgess, so as before now newly returned a Citizen for the said City.

The Journal’s unusually extensive report of the debate indicates that there was considerable weight of opinion against the idea of substitution of this kind, and that even if it were a good idea, ‘such Matter of Service, or Sickness might in any wise be Causes of removing any, and of chusing new in their Places, yet the same ought not to be done upon Suggestion made into the Chancery, but rather by the Judgment of this House’. The House eventually accepted the returns, but made it clear on the last day of the session that it did not consider these substitutes as entitled to sit in any further session, taking considerable pains to attempt to sort out the confusion that resulted where one existing Member had transferred to the seat for which another person, who was still alive, had originally been elected. They also firmly concluded ‘That, during the Time of Sitting of this Court’, no writ for a new election should go out without the House’s warrant to the Clerk of the Crown, ‘according to the ancient Jurisdiction and Authority of this House, in that Behalf accustomed and used’.

If in effect the House had tolerated substitutes in the special circumstances of the 1581 session, it did not do so again. When a related situation arose in 1606 of a series of Members who had been appointed under patents for life to judicial positions in Ireland, the House decided that they had vacated their seats, and ordered that new writs be issued for their constituencies. But that properly belongs to the history of resignation and recall – which conveniently begin with R.

1 thought on “Pairs and Proxies”

Leave a Reply to Resignation – Reformation to Referendum: Writing a New History of Parliament Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s