In the United Kingdom, the mental health of politicians has traditionally commanded relatively little attention, perhaps due to popular cynicism and distrust towards elected officials. Still, those involved in UK political life have characterised it as a ‘rough old game’, with MPs referencing broken marriages, overwork, loss of friendships and poor sleep. Over time, MPs have spoken more openly about their mental wellbeing, divulging specific episodes of mental ill-health. These anecdotal accounts are supplemented by empirical research showing a sizeable mental health burden amongst members of Parliament. Screening conducted in the 2016 Parliament found elevated rates of probable mental ill-health and harmful alcohol use in MP participants compared to populational control groups.
Distinctive risk factors include onerous working conditions, exposure to violence and abuse, and the pressure of high-profile decision-making. Conceivably, these pressures could be even more pronounced in leadership positions and especially that of the Prime Minister, where decision-making demands, responsibilities for domestic and international affairs, and intense public scrutiny could heighten psychosocial stressors.
Despite the evidence of poor mental health and the high stakes, the consequences of medical incapacity in heads of government are rarely comprehensively regulated through law. Even a constitution as notoriously legalistic as the German Basic Law is conspicuously silent on the matter. The US Constitution goes somewhat further in its 25th Amendment by outlining a more detailed procedure but likewise offers little indication as what constitutes an inability to ‘discharge the powers and duties’ of the presidency. Instead, much of the burden is carried by the political constitution: the conventions and raw arithmetic of political support that create a flexible set of loosely defined norms and arrangements enforced (or at times ignored) primarily through the political process.
The blog post summarises the findings of an inter-disciplinary research project, focusing on the approach of the UK constitution to medical incapacity in the Prime Minister. In doing so, I do not speculate on the mental health of any particular individual, and instead explore hypothetical scenarios as well as the general position on voluntary resignation and involuntary removal.
The office of the Prime Minister
The office of the Prime Minister is a creature of constitutional conventions, which also govern any medical incapacity, including mental health-related limitations. The office is not created nor meaningfully regulated by legislation (and only occasionally referenced). There is no single authoritative source for determining the existence, scope and content of constitutional conventions but the Cabinet Manual is nonetheless a good place to start. While it does not have any formal constitutional status and is therefore not authoritative on the present understanding of conventions, it does at least seek to articulate a common understanding of important constitutional conventions as they were arguably understood when the Manual was written in 2011.
The Cabinet Manual describes the Prime Minister as ‘the Sovereign’s principal adviser’, who ‘chairs Cabinet and has overall responsibility for the organisation of government.’ Accordingly, the Prime Minister leads the current government and has the power to appoint and dismiss all other ministers. The government is collectively accountable to Parliament and must retain the confidence of the House of Commons to remain in power.
Unlike in presidential systems, the UK Prime Minister is not directly elected as head of government but appointed by the Monarch exercising their royal prerogative. Following binding constitutional conventions, the Monarch must select the person best placed to command a parliamentary majority: in virtually all cases, this means the leader of the party with most MPs. Once appointed, the Prime Minister forms a government, remaining in office until they resign, die or are dismissed by the Monarch. Constitutional convention obliges the Prime Minister to resign or call an early general election if defeated on a confidence matter. They also typically resign following major policy defeats, or after losing the support of senior ministers and their party, or indeed at times owing to medical issues, as was last the case with Anthony Eden in 1957.
Short term and foreseeable incapacity
For our purposes, medical incapacity occurs if an individual is unable to perform (temporarily or permanently) some or all prime ministerial duties due to illness or injury. The rule of thumb is that the shorter and more predictable the incapacity, the more straightforward the implications under the UK constitution.
Should the Prime Minister become aware that they will be medically incapacitated in the short-term (with a reasonably foreseeable and specific return to full duties) and be able to make arrangements, then the scenario is comparatively straightforward. The Prime Minister may simply direct a governmental minister to fulfil specific duties in this period of incapacity and impose any advance restrictions they see fit, for instance, limiting the ability to dismiss ministers or instigate major policy changes. This happened most recently in 2020 when Prime Minister Boris Johnson received intensive care for COVID-19 and directed the-then foreign secretary, Dominic Raab, to temporarily assume prime ministerial duties.
At times, the title of Deputy Prime Minister is given to designate an individual who assumes certain responsibilities under such circumstances. If the designated minister in turn falls ill, then the succession would most probably follow the established ministerial ranking order.
Longer term and sudden incapacity
If the medical incapacity arises suddenly, without a reasonably foreseeable return to full duties and is such that it does not allow the Prime Minister to make arrangements in advance, then the constitutional response is more complex. It seems likely that under such circumstances the Prime Minister would resign, and senior government ministers in cabinet would propose an interim successor to the Monarch. This would broadly follow the process after Harold Macmillan’s resignation in 1963, where senior ministers and party officials informally explored the individual best placed to command a parliamentary majority, ultimately recommending Lord Home to the Queen.
As the historic example suggests, Parliament would likely be sidelined in favour of a party political processes. This is hardly surprising because Parliament does not play a role in the appointment of the Prime Minister (which remains a prerogative of the Monarch), nor does it formally endorse a newly formed government (it can merely withdraw confidence). This understanding of the constitution and the UK parliamentary system is supported by the lack of evidence that the medical capacity of a sitting Prime Minister has ever been the subject of formal debate in the House of Commons in modern times (there is no evidence for such debates in Hansard from the 19th century to the present day). The analogous scenario of a Prime Minister dying in office likewise suggests the primacy of the party political process.
The last Prime Minister to die in office was Lord Palmerston in October 1865. While Sir Henry Campbell-Bannerman died at No 10 Downing Street on 22 April 1908, he had resigned 19 days prior due to ill-health. In Lord Palmerston’s case, Parliament had been dissolved prior to his death in anticipation of a General Election and would not meet again until the opening of a new Parliament in February 1866. Similarly, George Canning died between two parliamentary sessions in August 1827. Although in this case Parliament could have conceivably been recalled early, it would not meet again until the next timetabled session in late January 1828. Prime Minister Spencer Perceval was assassinated on 11 May 1812, yet Hansard records no debate on his succession. Finally, William Pitt the Younger passed on 23 January 1806. Although Parliament met the following day, Hansard only records a motion paying tribute to the late Prime Minister.
There is also little indication that matters would unfold differently today. Should Parliament be adjourned, the Speaker of the House of Commons determines whether the House should be recalled early on the basis of representations made by government ministers (House of Commons Standing Order No 13). Combined with the significant control that the government exercises over the parliamentary agenda (House of Commons Standing Order No 14), it seems unlikely that senior ministers would seek a recall of Parliament, much less opt to allocate time to debate the succession of the Prime Minister in the midst of uncertainty over the party leadership. There is ultimately little historical evidence that medical incapacity or even the death of a Prime Minister is considered to call for parliamentary debate. The decision on succession thus effectively falls to the party with a parliamentary majority and depends on its internal political process.
Should a Labour Prime Minister die in office, party rules allow the Labour Party to appoint an acting leader until a leadership contest is concluded. This acting party leader could then plausibly be viewed as capable of commanding the confidence of the House of Commons, giving them a strong claim to be appointed Prime Minister.
Meanwhile, the death of a Conservative Prime Minister would trigger a vacancy, which leads to a leadership election in two stages: Conservative MPs would select two candidates to present to the membership, and then all party members cast their vote. There appears to be no express provision made for appointing an acting leader. Whilst the appointment of a new Prime Minister may be influenced by the order of precedence, the designation of a minister as Deputy Prime Minister, or the appointment of an interim leader, the Cabinet Manual clarifies at para [3.11] that this alone does not restrict prerogative powers for appointing a successor to the Prime Minister. Ultimately, the convention requires the Monarch to appoint the individual best placed to command a majority, and that need not necessarily be the Deputy Prime Minister (especially in a coalition government) nor necessarily the person chosen by party officials as acting leader.
An affirmative cabinet decision (following consultation with backbench MPs) would appear preferrable to confirm who is best positioned to command confidence, especially for coalition governments. Upon appointment, this individual would head the government until the relevant party can hold a leadership election.
Longer term and chronic incapacity
A constitutionally more challenging scenario arises if the Prime Minister is medically incapacitated for a prolonged and uncertain period of time, but remains in principle capable, albeit unwilling, to formally resign.
Section 141 of the Mental Health Act 1983 previously allowed a Member of the House of Commons to lose their seat should they experience a psychiatric disorder that incapacitated them for more than six months. However, this was abolished by s1 of the Mental Health (Discrimination) Act 2013 along with all other rules of the common law that disqualified a person from membership in the House of Commons on the grounds of mental illness.
What remains today is political pressure, which would undoubtedly be exerted on the Prime Minister should they not be in a position to continue in office. This would be akin to the Prime Minister losing support over a controversial policy decision. Recently, Theresa May, Boris Johnson, and Liz Truss were all compelled to resign following the loss of support from backbench MPs and ministers. Should this prove insufficient, further options remain: the governing party can force a leadership election through its own procedures or secure the passing of a parliamentary no confidence vote in coordination with the opposition.
For the Conservative Party, the 1922 Committee of backbench MPs is the primary party-political vehicle for changes to the party leadership. Provided that fifteen percent of Conservative MPs write a letter to the chairperson, an internal confidence vote is held that can trigger a leadership contest. Yet, even if the incumbent leader nominally survives this ballot, substantial political damage is often done: see for instance, the case of Theresa May. Labour MPs do not have similar procedures and instead can only initiate leadership challenges annually, prior to the yearly party conference. Here, an MP must secure the support of twenty percent of Labour MPs to trigger a contest, with the incumbent automatically on the ballot alongside other MPs who wish to stand.
For its part, Parliament can demonstrate a lack of confidence in the government (and by extension the Prime Minister) straightforwardly by passing an explicitly worded no confidence motion, or by defeating the government on a confidence matter. This last occurred in 1979 when the government of Prime Minister James Callaghan was defeated on a confidence matter by a single vote. Subsequently, the Prime Minister requested a dissolution and a general election shortly followed. By constitutional convention, the government is obliged to find parliamentary time for a confidence motion in the government put down by the opposition. The convention is not engaged if the motion is instead directed specifically against the Prime Minister or other Ministers, where it appears from recent precedent that the government is entitled to refuse to table the motion. Thus, both leadership challenges and no confidence votes could entail the involuntary removal of the Prime Minister.
An interesting variant would be a scenario where the Prime Minister lost a confidence vote and was therefore by convention required to resign but unable to do so owing to their incapacity. This would be a rare case where it would be appropriate for the Monarch to dismiss the Prime Minister using their prerogative powers. The Monarch could conceivably act either on their own initiative following informal inquiries with the governing party or in response to a parliamentary prayer motion requesting the dismissal of the Prime Minister. This would be a truly exceptional situation: the last Monarch to dismiss his Prime Minister over a political disagreement was King William IV, who replaced the incumbent Whig Prime Minister Lord Melbourne with Robert Peel in 1834. The Whigs nonetheless continued to maintain a parliamentary majority and Lord Melbourne was eventually reappointed in 1835.
Since then, monarchs have not dismissed Prime Ministers of their own volition pursuant to a robust constitutional convention against involving themselves directly in political matters. An exception is justifiable here because dismissing the Prime Minister if they are unequivocally required to resign but cannot do so for medical reasons would not constitute a politically sensitive intervention. Instead, the Monarch would assume an appropriate, apolitical role to resolve a constitutional impasse according to the wishes of a democratic parliamentary majority – similar in principle to denying a prorogation of Parliament when the government has lost (or is poised to lose) the confidence of the House of Commons.
Overall, the UK constitution thus appears reasonably well equipped to respond to medical incapacity in the Prime Minister. While there is a degree of uncertainty in precisely how the conventions would play out in practice, there appear to be no major constitutional gaps. The shorter in duration and the more foreseeable the incapacity, the easier the process for temporarily replacing the head of government. More challenges admittedly arise where the incapacity is either longer in duration or comes about more suddenly: removing and replacing the Prime Minister in those cases is more complicated.
We should not be surprised that conventions and the party political process dominate, and that legal constitution and Parliament are typically sidelined. In this sense, succession simply mirrors the process which originally led to the appointment of the Prime Minister. Asking which person is ‘best placed’ to command the confidence of the House of Commons is an inherently political question that is difficult to meaningfully regulate through law, save perhaps in specifying a procedure. This task typically falls to internal party rules for leadership contests, which have recently proven themselves as effective, if not entirely uncontroversial, vehicles for replacing Conservative Prime Ministers. Therefore, the absence of regulation through the legal constitution does not, in the end, necessarily indicate an absence of robust procedure.
The author would like to thank Alison Young and the editors for helpful feedback on earlier drafts.
Stefan Theil is Assistant Professor in Public Law at the Faculty of Law, University of Cambridge, and a Fellow of Sidney Sussex College
The blog post is based on a co-authored article entitled Mental health, medical incapacity, and political leadership in the United Kingdom(forthcoming, (2025) 101 International Journal of Law and Psychiatry) together with Alexander Smith and Michael Liebrenz (Bern), Dinesh Bhugra (KCL) and Ashley Weinberg (Salford).
(Suggested citation: S. Theil, ‘Medical Incapacity and the UK Constitution’, U.K. Const. L. Blog (20th May 2025) (available at https://ukconstitutionallaw.org/))
