Ending Life as Cutting Costs: Analysis of ‘The Impact of the Terminally Ill Adults (End of Life) Bill I’ (Professor David Albert Jones)

The Department of Health and Social Care has produced two Impact Assessments and one Memorandum on the Terminally Ill Adults (End of Life) Bill. This paper concerns the first of these, which focuses on the financial costs and cost reductions of implementing the Bill.

The financial IA considers evidence from other jurisdictions but focuses almost exclusively on the United States, and neglects New Zealand, Australia and Canada. However, the United Kingdom is closer to these Anglophone Commonwealth countries in its healthcare systems and societal norms than it is to the United States. On the model of these countries, the potential numbers seeking death under the Bill could be far higher than is estimated. 

The main flaw in the financial IA is not its likely underestimate of numbers of deaths but its decision to measure financial costings without measuring benefit and disbenefit. This effectively sets no value on protecting life but measures the impact of policy only by financial costs and cost reductions. The costs reductions it identifies are not because people could live without the proposed spending or could live with cheaper alternatives, but are because they are dead. Dead people do not utilise healthcare or receive care from local authorities, or benefits from the state. The implementation of the Bill would cost millions, but over time it could pay for itself by ending people’s lives.

A government department has thus provided detailed estimates for how much money the NHS could save if only patients ended their lives earlier. This is truly a sinister document.

The Department of Health and Social Care has produced two impact assessments of the Terminally Ill Adults (End of Life) Bill [the Bill] and a related Memorandum. The first, simply entitled ‘Impact Assessment’ [1] is principally concerned with the financial costs and cost reductions involved in implementing the Bill (‘financial IA’). A second impact assessment [2], co-sponsored by the Ministry of Justice, concerns the potential impact on equality law (‘equality IA’). Alongside these two, the Department Health and Social Care and the Ministry of Justice have also issued a Memorandum [3] on whether the Bill is compatible with the European Convention on Human Rights (‘Memorandum’). This is the first of three papers on the three documents. [4]

The financial IA begins by considering the international context. It identifies 25 jurisdictions ‘where assisted dying is legal and a formal system with regulated processes (such as an established application and approvals process) is in place.’ [5] Note: while this list includes 25 jurisdictions it covers only 11 countries [6], as most jurisdictions listed are within two countries, the USA or  Australia. It is also contentious to describe Switzerland (one of the 25 listed) as having ‘a formal system with regulated processes’. [7] Assisted suicide in Switzerland is provided by private organisations such as EXIT and Dignitas. These each have their own rules but there are no specific statutory regulations for the practice of assisted suicide in Switzerland. There is no equivalent to the Bill in Swiss law.

The financial IA states that ‘in most jurisdictions, the life-ending substance(s) must be self-administered by the Individual’. [8] This is not true, as only 12 out of 25 jurisdictions restrict the practice to self-administration. Furthermore, it is also misleading to speak of the practice in ‘most jurisdictions’ where this represents the practice in only one of two countries. In the case of self-administration, this is a requirement in only 3 out of 11 countries, one of which is Switzerland which, as noted above, has no specific regulatory oversight.

Of the 25 jurisdictions listed, 10 are identified by the financial IA as having ‘comparable eligibility criteria to this Bill’. [9] This choice is important for the methodology of the financial IA, as data from these 10 jurisdictions provide the basis for estimates of numbers of requests and of numbers of deaths. The great majority of those on this list are jurisdictions in the United States (9 out of 10)  [10] all of which are modelled closely on the law in Oregon. The list includes all the jurisdictions in the USA which have passed legislation in this area except New Jersey. [11]

The only non-US jurisdiction held to be ‘comparable’ is New Zealand. However, only two years of data from New Zealand are included. [12] The weight of analysis relies on US jurisdictions and especially on Oregon rather than on New Zealand.

What is unaccountable is the inclusion of New Zealand but exclusion of Australian jurisdictions even though they have very similar laws. [13] This seems to be because the financial IA inclusion criteria specifies ‘a terminal illness from which they are expected to die within 6 months’. [14] However, while it is true that all Australian States permit ‘voluntary assisted dying’ with 12 months expectation of death for neurodegenerative conditions, most (5 out of 6) confine eligibility to 6 months for all other terminal illnesses. [15] In these States, the criterion of 6 months is thus used in 85-91% of cases. [16] Note also that most of those diagnosed with neurodegenerative diseases and an expectation of death within 12 months would have been eligible at the time or later had the time limit been 6 months (as the proportion of assisted death for neurological conditions in Oregon which has a 6 months criterion, is well within the range for States in Australia which have a 12 month criterion). [17] The practices in New Zealand and Australia are clearly ‘comparable’ and if New Zealand is considered as a ‘comparable jurisdiction’ in the financial IA, Victoria, Western Australia, South Australia, Tasmania, and New South Wales should also be so considered.

It should also be noted that the example of Australia was an important model for the sponsor of the Bill and was raised repeatedly in the Committee Stage of the Bill. [18] Australia supplied more international witnesses than any other country [19], and that the language of ‘voluntary assisted dying’ (VAD) which has been incorporated into the Bill since Committee Stage [20], and is used over 100 times in the financial IA [21], is used only in Australian laws on assisted dying, not in any jurisdiction in any other country.

There is also an argument for including consideration of Canada in the financial IA. Canada is also an Anglophone Commonwealth country and has had some influence on the development of the law in Australia. It is true that eligibility criteria are different in Canada, in that they include people who are not terminally ill. However, until 2021 it at least had the requirement that death be ‘reasonably foreseeable’ [22] and, though the law has since expanded to include people whose death is not foreseeable, the great majority still fall within the category of ‘reasonably foreseeable’. [23]

The financial IA acknowledges that ‘there are a range of reasons [why] uptake may still vary across jurisdictions… for example, public health systems, population demographics, societal norms, length of time the service has been operational, design of the process, and mechanisms for reporting usage’. [24] However, the document prioritises relative similarity in eligibility criteria (between the Bill and laws in the USA) over relative similarity in healthcare systems or societal norms (between the UK and New Zealand, Australia and Canada).

If due weight were given to experience in New Zealand, Australia and Canada the higher estimate for numbers of deaths would need to be increased considerably. The financial IA states that ‘it was observed that none [of the comparable jurisdictions] exceeded an assisted death rate of 1% of total deaths in any reported year’. [25] Even without including Canada (which the financial IA acknowledges has reach 4.7% of deaths in year 8 [26]), the examples of New Zealand and Australia suggest that the model adopted by the financial IA underestimates the potential numbers of deaths. The financial IA assumes as its highest estimate, even at year 10, a figure of 0.68% of total deaths [27], whereas New Zealand reached 1.2% of total deaths in year 3 [28], and Western Australia reached 1.6% of total deaths in year 3. These figures are higher than Canada had reached in year 3. [29] The upper estimate of potential numbers in the financial IA could easily be inaccurate by a factor of 5 or more.

Another consequence of focusing on US States is that this gives the impression that the law will remain stable over time: ‘This IA did not identify any significant changes made to VAD services in the 10 comparable jurisdictions’. [30] It is arguable that this is misleading even in the USA where there has recently been a discernible shift. There were no legal changes from 1997 to 2019, but since then there have been 9 changes in 7 States. [31]

It is still more misleading to state that New Zealand has not seen ‘any significant changes’ when, in November 2024, the New Zealand government announced plans for a major revision of the law, including expanding eligibility. [32] Similarly, in February 2025, the government of Victoria, Australia announced plans for a major expansion of its law. [33] In both cases these plans for revision were catalysed by statutory reviews (after 3 years in New Zealand and after 5 years in Victoria). As the Bill also includes a statutory review after 5 years [34], the financial IA ought also to have included, at least as a caveat, the possibility of significant expansion of eligibility and/or of dropping of significant safeguards from the Bill, as a result of the 5-year review.

No quantification of impact on length or quality of life

The financial IA acknowledges that health policies and interventions are ‘usually assessed in reference to Quality-Adjusted Life Years (“QALYs”)’. [35] Nevertheless, a decision has been taken not to do so in this case:

‘QALYs are not designed to quantify the health impacts of choosing to end life. Therefore, they have not been used in this IA.’ [36]

‘… to ensure consistency throughout this assessment, we have expressed such impacts in financial, rather than QALY-value terms.’ [37]

‘Given that the intent of this Bill is to provide lawful assistance to terminally ill adults to end their life, and that this provision may be considered a benefit or disbenefit depending on individual views and circumstances, impacts have not been quantified in QALY terms.’ [38]

It is certainly correct to say that QALYs are not designed to quantify the impact of intentionally ending life. This touches on an important philosophical point: Being dead is not a state of extremely poor health nor a state of extremely low quality of life. Death is not a state of health at all, nor is it a state of life. It is the limit of life. [39]

Some philosophers have also criticised QALY analysis more broadly as a measure of benefit in healthcare. The use of QALYs creates a systematic bias against those who have ongoing health conditions or disabilities and against people who are older and thus have fewer potential life years. It thus discriminates against people who have the greatest need for healthcare and who already face structural forms of discrimination in accessing healthcare. This bias against those in a position of greatest need has been termed the ‘double jeopardy’. [40]

Others criticise attempts to place a value on human life. The problem is not whether a lower value is attributed to the lives of people with disabilities and older people, but is the very idea of estimating or seeking to measure the relative value of anyone’s life. Human life is not ‘valuable’ but invaluable. [41]

Both these criticisms have merit, and when considering public policy it is important to recognise the systematic bias of QALY analysis and to counterbalance this with discussion of wider social value judgements. Nevertheless, an   assessment of the financial or economic impact of a policy requires an assessment of benefits as well as an assessment of costs. One medicine may be more expensive than another but the expense may be justified if it is more effective in relation to extending life or in relation to improving quality of life.

Without any attempt to measure the benefit or disbenefit of implementing the Bill, the financial  IA effectively sets the value of life as zero. If there is no weighing costs against benefits, what remains is weighing costs against cost reductions, including cost reductions that occur only because the person is dead. This frames death as a potential means of cost reduction.

Such a methodology is flawed and can be recognised as flawed whether one supports or opposes the Bill. The financial IA fails to assess whether the Bill would, as intended, increase quality of life at the end of life, whether it would  reduce loss of life by unassisted suicide, or whether it would have a negative impact on quality and length of life through unintended consequences, for example interference with palliative care services [42] or increases in ‘conventional’ (‘unassisted’) suicide. [43] These assessments are subject to great uncertainty but there is some evidence, and without attempting to quantify non-financial benefit what remains is financial cost and cost reduction separated from any connection to non-financial benefit and disbenefit.

Adding up costs and cost reductions

The financial IA overtly frames death as the cause of cost reduction due to ‘unutilised healthcare’. It also makes clear that the reduction in costs will  be proportional to the life years lost.

‘The estimated reduced cost from unutilised healthcare ranges from between £919k to £10.3m in Year 1 (which is half a year), to between £5.84m to £59.6m in Year 10.’ [44]

‘Assuming all assisted deaths occur after 2 months, reducing the length of life by 4 months, then it is estimated that 79% of the associated healthcare costs (for months 4 to 1) are no longer required… this amounts to a potential reduction in spend of between £2.14m to £10.3m in Year 1 (which is half a year) and £13.6m and £59.6m in Year 10 (in 2025/26 prices).’ [45]

‘Assuming all assisted deaths occur after 5 months, reducing the length of life by 1 month, then it is estimated that 34% of the associated healthcare costs (for month 1) are no longer required… this amounts to a potential reduction in spend of between £919k and £4.41m in Year 1 (which is half a year) and £5.84m and £25.6m in Year 10 (in 2025/26 prices).’ [46]

The financial IA also provides estimates to how much local authorities would save from unneeded care home services:

‘In Year 1 (which is half a year), between £49.7k (low cohort estimate, lower bound fee, 1 month not needed) and £1.82m (high cohort estimate, higher bound fee, 4 months not needed);’ [47]

‘In Year 10, between £316k (low cohort estimate, lower bound fee, 1 month not needed) and £10.5m (high cohort estimate, higher bound fee, 4 months not needed).’ [48]

Finally it estimates the reduction in payments for state pensions and state benefits. It states that ‘Social security payments are considered “economic transfers” in HM Treasury’s Green Book, meaning they do not constitute a saving or cost for society’. [49] Nevertheless, the financial IA goes on to provide estimates for these ‘transfers’.

‘for state pensions, between £113k and £2.17m in Year 1 (which is half a year), to between £1.05m and £18.3m in Year 10; [50]

for Attendance Allowance, between £55.7k and £1.07m in Year 1 (which is half a year), to between £354k and £6.20m in Year 10; [51]

for PIP, between £28.5k and £547k in Year 1 (which is half a year), to between £181k and £3.17m in Year 10.’ [52]

It should be emphasised that all these reductions in spending from unutilised services and unclaimed benefits are solely because a person has died sooner than they otherwise would have. 

The financial IA does not provide a table of the cost of implementation against cost savings from implementation, but gives estimates of costs for different elements of implementation separately. Estimates are provided for the cost of the education campaign which would have to accompany the introduction of the Bill [53], the cost of healthcare staff training [54], the cost in healthcare staff time [55], and the cost of  the review panels. [56] The largest single item of expense is ‘the cost of the Voluntary Assisted Dying Commissioner, and their office’, which, on the model of other arms-length bodies, ‘would be approximately £10m per year’. [57]

If these figures are combined (which is not done within the document) the overall costs of implementation, as estimated by the financial IA, are between £7.4m and £20.1m in Year 1 and between  £15.4 and £37.0m in Year 10.

The financial IA thus estimates that there would be considerable upfront costs in implementing the Bill. However, on these estimates, the VAD service has the potential by Year 10 to save more money in ‘unutilised healthcare’ and in other cost reductions than it costs to implement. Furthermore, this cost saving would be greater the greater the number who died in accordance with the Bill’s provisions, and the greater the length of time by which each life was cut short. It has been argued above that if numbers are closer to those in Australia then they will be much higher than the financial IA estimates and the Bill could thus be cost reducing as early as year 3.

It would be unusual in other contexts for ‘reduced cost’ [58] due to death to be considered in this way. For example, discussion of the impact of suicide prevention strategy or of public health interventions for COVID would not usually include, much less seek to quantify, the reduced healthcare costs, reduced local authority costs, and reduced cost to the state in pensions and state benefit due to a person dying prematurely by suicide or from COVID. It is certainly appropriate to look at cost reductions from unused services where patients are using alternative services. However, it does not seem appropriate to quantify cost reductions which are due to patient deaths, especially where these deaths may have been caused (or not been averted) by the intervention.

The presentation of costs and cost savings, without quantifying potential benefit or disbenefit in relation to length and quality of life, frames death by assisted dying / assisted suicide as saving the NHS money. This framing of the issue carries the danger of incentivising government and/or healthcare professionals to promote utilisation of VAD services for financial reasons. In this context, it is of great concern that the law permits healthcare professionals to raise this option pro-actively, without the patient having asked about it. [59]

Time costs of psychiatrists

The most striking omission from the financial IA, and from the other two documents, is any reference to the impact of the Bill on suicide prevention or on mental health services more generally.

There are a number of references to psychiatrists in the financial IA, but nearly all of these refer to the role, availability and cost of a psychiatrist as a member of a VAD review panel. [60] In this context, the financial IA expresses the opinion that there is ‘a low risk to wider capacity’ [61] in mental health services, given that the number of psychiatrists has increased by more than 10%  over the past 14 years:

‘The number of full-time equivalent psychiatry consultants working in all NHS trusts and other core organisations in England has increased by 10.7% since 2010 from 4,121 in September 2010 to 4,561 in September 2024. Over the last year, the number of FTE psychiatry consultants has increased by 82 (1.8%) from 4,479 in September 2023 to 4,561 in September 2024.’ [62]

The fact that the number of psychiatrists has  increased does not show that numbers are adequate to respond to demand for services, as it does not factor in the historic underfunding of mental health services nor the rise in demand for mental health services (even before the COVID pandemic). In contrast to the financial IA statement, the Royal College of Psychiatrists  argued in May 2023 that:

‘Based on NHS England targets to expand the workforce from 2016 onwards, there is currently a shortfall of 690 consultant psychiatrists across England – 15% of the current consultant psychiatrist workforce.’ [63]

In June 2024 the BMA published research conducted through ‘in-depth interviews with doctors across the mental health system, including those working in psychiatry, general practice, emergency medicine, and public health’. [64] The name of the report is self-explanatory: “It’s broken”: Doctors’ experiences on the frontline of a failing mental healthcare system. [65]

Examples could easily be multiplied. There is no evidence of any spare capacity in the mental health system. In this context, the ‘opportunity cost to staff time working on this’ [66], is one that would have a direct adverse impact on the delivery of mental health services, taking some thousands of hours a year in staff time. This should have been acknowledged more overtly in the document.

Reinforcing a pernicious message

Another consequence of framing death as cost reduction is that it reinforces the pernicious belief that those receiving care are a burden to society. This idea is one that can result in indirect pressure on a person to end their life. This danger is acknowledged in the financial IA to be one of the ‘main reported disbenefits’ of legalising such practices. [67]

‘There are a range of reasons put forward to explain why individuals might feel actual or perceived pressure to end their life. In Oregon, research conducted with terminally ill adults prior to their assisted death and published in the 2024 Annual Report showed ‘burden on family, friends/caregivers’ (42%) and ‘financial implications of treatment’ (9%) as end-of-life concerns.’ [68]

To this it may be added that, while autonomy remains the main motivation for seeking death under the Oregon law, concerns about being a burden and about the financial implications of treatment have both increased significantly since the law came into force. [69]

A clear example of how such framing death as cost cutting can endanger patients is given by the late Nicholas Tonti-Filippini, an Australia bioethicist who suffered from a severe, incurable, chronic illness which required regular burdensome treatment and frequent hospitalisation. He describes the experience of receiving a letter telling him how much his care  was costing:

‘For several years, until I objected, I received from my health insurer a letter that tells me how much it costs the fund to maintain my health care. I dreaded receiving that letter and the psychological reasoning that would seem to have motivated it. Each year I was reminded how much of a burden I am to my community. The fear of being a burden is a major risk to the survival of those who are chronically ill.’ [70]

The Equality and Human Rights Commission has also highlighted the need, before passing a law that would allow assisted dying / assisted suicide, to address social conditions that would otherwise lead to people dying through a form of indirect coercion.

‘It is also necessary to consider coercion beyond the risks posed by individuals. One of the most important protections against people feeling coerced into seeking an end to their life is to ensure social conditions, support, care and services are in place so that people with disabilities or serious or terminal illnesses do not feel that they are a burden to their loved ones or to society. This goes beyond adequate funding and access to health and social services, and must include active efforts to create a society where people are able to live life on equal terms, free from discrimination.’ [71]

The Department of Health and Social Care has produced separate documents on equality law (equality IA) and on human rights (Memorandum) which are discussed in separate papers. [72] These will return to the issue of the relation between social inequality, feelings of being a burden and the pressure to end one’s life.

The key relevance of this issue in the context of the financial IA is that the methodology employed, and especially the framing of premature death as cost saving, could exacerbate the problem of people wrongly feeling that their lives constitute a burden to others. [73]

The approach taken by the financial IA sends out a pernicious message. Nevertheless, the financial IA is perhaps helpful in making this attitude overt. It demonstrates that critics have been right to express the concern that this Bill could be regarded as a means of cutting costs. The financial IA shows that the provisions of the Bill can be construed in this way by a government department. The Department of Health and Social Care has provided Parliament with detailed estimates for how much money the NHS could save if only patients ended their lives earlier. This is truly a sinister document. [74]

If the issues discussed here affect you or someone close to you, you can call Samaritans on 116 123 (UK and ROI), visit their website https://www.samaritans.org/ or contact them on jo@samaritans.org.

If you are reporting or writing about a case of death by suicide, whether assisted or non-assisted, please consult media guidelines https://www.samaritans.org/about-samaritans/media-guidelines/ on how to do so responsibly.

[1] Terminally Ill Adults (End of Life) Bill (as amended in the House of Commons Public Bill Committee), Department of Health and Social Care (DHSC) / Ministry of Justice (MoJ) (14 May 2025).

[2] Terminally Ill Adults (End of Life) Bill: equality impact assessment, Department of Health and Social Care / Ministry of Justice (14 May 2025).

[3] Terminally Ill Adults (End of Life) Bill: ECHR memorandum, Department of Health and Social Care / Ministry of Justice (2 May 2025).

[4] The other two being: An Equal Opportunity to Live: The Impact of the Terminally Ill Adults (End of Life) Bill II; and A Human Right to Suicide Prevention: The Impact of the Terminally Ill Adults (End of Life) Bill III (both upcoming).

[5] Financial IA, para 6. The 25 Countries are listed in the financial IA Table 57, page 115: Austria, Belgium, California, Canada, Colorado, Hawaii, Luxembourg, Maine, New Jersey, New Mexico, New South Wales, New Zealand, Oregon, Portugal, Queensland, South Australia, Spain, Switzerland, Tasmania, The District of Columbia, The Netherlands, Vermont, Victoria, Washington and Western Australia.

[6] Austria, Australia, Belgium, Canada, Luxembourg, New Zealand, Portugal, Spain, Switzerland, the Netherlands, and the United States.

[7] Op. cit., Financial IA, para 6.

[8] Ibid., para. 7.5.

[9] Ibid., para. 8.

[10] The ‘comparable jurisdictions’ are California (USA), Colorado (USA), Hawaii (USA), Maine (USA), New Mexico (USA), New Zealand, Oregon (USA), the District of Columbia (USA), Vermont (USA), and Washington (USA).

[11] The omission of New Jersey is due to a mistake in Table 57 where it states that New Jersey extends eligibility to 12 months for neurodegenerative diseases. A possible source for this error is confusion between New Jersey and the Island of Jersey which is proposing a 12 month timeframe for neurodegenerative diseases.

[12] In fact 3 years data was available from New Zealand through the Annual Service Reports whereas the financial IA has taken data from the Registrar (assisted dying) Annual Report to the Minister of Health which uses the financial year, hence only two years of data are available.

[13] See D.A. Jones, Wrong Side of the World: The Misplaced Reliance on Australia in the UK Debate on ‘Assisted Dying’, Oxford: Anscombe Bioethics Centre, 2024.

[14] Op. cit., Financial IA, para. 8.

[15] Victoria, Western Australia, South Australia, Tasmania, and New South Wales.

[16] Total with neurological conditions for Victoria, 9% (Victoria Voluntary Assisted Dying Review Board Annual Report (July 2022 to June 2023), p. 20); for Western Australia, 12.4% (Voluntary Assisted Dying in Western Australia – Annual Report 2023–24, p. 25); for South Australia, 10% (Voluntary Assisted Dying Review Board 2023-24 Annual Report, p. 19); for Tasmania, 15% (Voluntary Assisted Dying Annual Report 2023–24, p. 10); and in New South Wales, 9.9% (NSW Voluntary Assisted Dying Board – Annual Report 2023–24, p. 10).

[17] Proportion with neurological conditions in Oregon over all years is 11.4% (Oregon Death with Dignity Act Data Summary 2024, p. 14) which is mid-range between Victoria (9%) and Tasmania (15%).

[18] In total there are 245 references to Australia in the consolidated transcript of the Public Bill Committee Debates on the Bill.

[19] Op. cit., D.A. Jones, Wrong Side of the World (2024), p. 3.

[20] See Terminally Ill Adults (End of Life) Bill (as amended in Public Bill Committee), clauses 4, 38, 39 and Schedule 1, compare with the Bill as introduced.

[21] See op. cit., Financial IA, paras. 2, 9, 12, 22, 23, 35, 40, 49, 52 (Table 1), 53, 75, etc.

[22] Government of Canada: Canada’s medical assistance in dying (MAiD) law.

[23] In 2023, 95.9% of MAID cases (n=14,721) were individuals whose death was reasonably foreseeable (Track 1) and 4.1% (n=622) were individuals whose death was not reasonably foreseeable (Track 2)’. See Government of Canada: Fifth Annual Report on Medical Assistance in Dying in Canada, 2023, Section 2.2.

[24] Op. cit., Financial IA, para. 60.

[25] Ibid.

[26] Ibid., para. 9.

[27] Ibid., para. 70.

[28] There were 453 assisted deaths in New Zealand in year 3 (Assisted Dying Service Annual Service Report 2024, p. 12), out of 37,722 deaths.

[29] Voluntary Assisted Dying in Western Australia – Annual Report 2023–24, p. 6.

[30] Op. cit., Financial IA, para. 12.

[31] Oregon 2019, California 2021, Vermont 2022, Hawaii 2023, New Mexico 2023, Oregon 2023, Washington 2023, Vermont 2023, and Colorado 2024 see Jones DA, ‘A Different Slippery Slope’, The American Journal of Bioethics, (2024) 24.2: W1-W1, doi: 10.1080/15265161.2023.2296406.

[32] Ministry of Health, 2024: Review of the End of Life Choice Act 2019, Wellington: Ministry of Health, pp. 10–22.

[33] Voluntary Assisted Dying Laws Still The Compassionate Choice, Minister for Health Mary-Anne Thomas, Media Release (Thursday, 20 February 2025).

[34] Op. cit., Terminally Ill Adults (End of Life) Bill (as amended in Public Bill Committee), clause 46.

[35] Op. cit., Financial IA, para. 92.

[36] Ibid., para. 50.

[37] Ibid., para. 51.

[38] Ibid., para. 92.

[39] S.E. Barrie, ’QALYs, euthanasia and the puzzle of death.’ Journal of Medical Ethics 41.8 (2015): 635-638.

[40] J. Harris, ‘QALYfying the value of life.’ Journal of Medical Ethics 13.3 (1987): 117-123.

[41] R. Stith, ‘Toward freedom from value’. Jurist, 38 (1978), 48; R. Stith, ‘The priority of respect: How our common humanity can ground our individual dignity.’ International philosophical quarterly 44 (2004): 165.

[42] D.A. Jones, Evidence of Harm: Assessing the Impact of Assisted Dying / Assisted Suicide on Palliative Care, Oxford: Anscombe Bioethics Centre (2024).

[43] D.A. Jones, Suicide Prevention: Does Legalising Assisted Suicide Make Things Better Or Worse?, Oxford: Anscombe Bioethics Centre (2024).

[44] Op. cit., Financial IA, para. 254, bold in the original.

[45] Ibid., para. 269, bold in the original.

[46] Ibid., para. 270, bold in the original.

[47] Ibid., para. 297.1.

[48] Ibid., para. 297.2.

[49] Ibid., para. 322, bold in the original.

[50] Ibid., para. 327.1, bold in the original.

[51] Ibid., para. 327.2, bold in the original.

[52] Ibid., para. 327.3, bold in the original.

[53] Ibid., para. 123.

[54] Ibid., para. 152.

[55] Ibid., paras. 214, 218.

[56] Ibid., para. 239.

[57] Ibid., para. 229, bold in the original.

[58] E.g. Ibid., para. 254, bold in the original.

[59] Op. cit., Terminally Ill Adults (End of Life) Bill (as amended in Public Bill Committee), clause 5(2).

[60] Op. cit., Financial IA, paras. 27, 101, 172, 178, 203, 206, 225, 230, 241, 423.1. The only reference outside this context is the possibility of referral to a psychiatrist for an expert assessment of decision-making capacity (para 28).

[61] Ibid., para. 249.

[62] Ibid., para. 248.

[63] Press Release: Royal College of Psychiatrists calls for urgent publication of NHS Workforce Plan as psychiatrist numbers stagnate, Royal College of Psychiatrists (25 May 2023).

[64] Summary Page: “It's broken”: Doctors’ experiences on the frontline of a failing mental healthcare system, British Medical Association (as updated, Friday 28 June 2024).

[65] “It’s broken”: Doctors’ experiences on the frontline of a failing mental healthcare system. British Medical Association (2024).

[66] Op. cit., Financial IA, para. 249.

[67] Ibid., para. 90.

[68] Ibid.

[69] D.A. Jones, D. Paton and P. Rutaquio, ‘Trends in End-Of-Life Concerns after the Legalization of Assisted Suicide’ (January 21, 2025). Available at SSRN: https://ssrn.com/abstract=5105556

[70] N. Tonti-Filippini About Bioethics, Volume 2. Ballan, VIC: Conor Court (2012), p. 112.

[71] Written evidence submitted by The Equality and Human Rights Commission (ADY0317) to the House of Commons Health and Social Care Committee on Assisted Dying / Assisted Suicide (January 2023).

[72] Op. cit., An Equal Opportunity to Live: The Impact of the Terminally Ill Adults (End of Life) Bill II; and A Human Right to Suicide Prevention: The Impact of the Terminally Ill Adults (End of Life) Bill III.

[73] D.A. Jones, ‘Do we want assisted death for people who feel they are “only a burden”?’ Parliament Politics Magazine (22 July 2024).

[74] John Pring, ‘Sinister’ government analysis of assisted dying bill adds weight to fears of financial incentives for deaths’, Disability News Service (8 May 2025).