BSA 43 | Who supports Reform?
Report
BSA 43 | Attitudes to welfare
Have attitudes to disability benefits changed?
This chapter analyses public attitudes towards spending on disability benefits and relevant trends in attitudes to welfare spending more broadly.
Summary
The level of spending on disability benefits is often said to be ‘unsustainable’. Yet the Starmer government’s attempt in 2025 to cut disability benefits was opposed by much of the public and partially thwarted by a rebellion of Labour MPs. However, new proposals for reform are in the pipeline. This chapter analyses the public’s support for spending on disability benefits, how views have changed over time, and how the public might react to future changes. It focuses in particular on whether shifts in attitudes are best understood as:
- a reaction to changing levels of spending on welfare,
- the result of changes in the perceived ‘deservingness’ of disability
benefit recipients, or - a reflection of a political polarisation of attitudes.
Spending on disability benefits is relatively popular
- Support for the government spending more on ‘welfare benefits for the poor’ is comparatively low. The proportion who are opposed is 15 points higher than the proportion in favour.
- However, 49% back more spending on benefits for disabled people who cannot work, while just 7% are opposed.
- No less than 63% say that benefits for disabled people are a priority for extra spending on social benefits – more than say this of any other benefit including retirement pensions (53%)
- Even among those who oppose the government spending more on welfare benefits for the poor, just 11% favour spending less on disabled people.
- Nevertheless, support for spending more on disability benefits for those who cannot work is at its lowest level since the question was first asked in 1998.
Views on spending are politically more polarised
- In 1998, those supporting left-leaning parties were 35 points more likely than those backing right-leaning parties to support spending more on welfare benefits. Now the gap is 47 points.
- Those who support left-leaning parties were only six point more likely than those backing right-leaning parties to support spending more on disabled people who cannot work. The difference now is 25 points.
- Nevertheless, even among supporters of right-leaning parties, the balance of opinion is in favour of spending more on benefits for this group. Only 11% of Reform supporters favour less spending as do just 7% of Conservatives.
It is difficult to forecast whether any particular proposal to reduce disability benefits will be popular or not. However, it is clear that this is very challenging terrain for a government to be confident of taking the public with them.
Introduction
In recent years, it has become commonplace for politicians and the media to describe spending on disability benefits as ‘unsustainable’ – a view often presented as reflecting wider public opinion. Despite this, the Starmer Government’s attempt to cut disability benefits in 2025 was opposed by a majority of the population,1 and was (at least partly)2 thwarted by a rebellion of Labour MPs. Given that both the Milburn Review and the Timms Review3 are making new proposals for reform this autumn, this is a good time to try to shed light on this apparent contradiction. The existing literature on the evolution of attitudes in this area suggests there are three ways in which we might expect attitudes to disability benefits to have changed – in reaction to changes in levels of spending, in line with changes in public perceptions of disability benefit recipients, and as a result of a political polarisation in views. In this chapter, we use data from four decades of the British Social Attitudes (BSA) survey to analyse how public attitudes towards spending on disability benefits have changed in practice and how the trends we have identified relate to attitudes to welfare spending more broadly. What do our findings tell us about public attitudes towards disability benefit reform, and how might the public react to future proposed changes?
How would we expect attitudes to disability benefits to have changed?
We begin by outlining, on the basis of the current literature, how we might expect attitudes to spending on disability benefits to have changed in recent years. Broadly speaking, there are three ways in which we might expect attitudes to spending on disability benefits to change – as a reaction to changes in levels of spending, in response to changing perceptions (or prototypes) of disability benefit recipients, and as a consequence of political polarisation. We begin by examining the evidence for these three mechanisms in turn, before considering which are most likely to have driven changing attitudes; given recent trends in welfare policy and rhetoric generally, and in terms of disability specifically.
Reactions to spending changes
One of the core findings from previous analyses of BSA surveys has been that attitudes to benefits spending in general are essentially ‘thermostatic’ (Curtice, 2010). This is the relatively straightforward idea that, as welfare spending increases, the public will notice, and support for yet more spending will erode (and vice-versa). This is similar to how a thermostat operates: as people feel warmer, they become more likely to say that they want to be colder, but this does not mean that their ideal temperature has changed. Alternatively, people may not have a clear ‘ideal’ level of benefits spending, but will still be more likely to say they want more spending if they perceive that it has recently been cut (and vice versa).
A simple expectation in terms of disability benefits would therefore be that, as spending on this area of welfare has risen, public support for cuts to disability benefits should have risen in tandem (and support for more spending should have declined). However, the relationship between public perceptions and actual spending levels is likely to be significantly more complicated than this simple story suggests.
First, increased spending on disability benefits is not part of an overall increase in welfare spending. Disability benefits spending has genuinely been rising, both in real terms and as a proportion of GDP (as shown in Figure 1). However, proportional to GDP,4 overall levels of welfare spending on working-age adults and children have been relatively flat. Hence, rising disability benefits spending has been compensating for cuts elsewhere rather than driving the overall (non-pensioner) welfare bill to ‘unsustainable’ new heights. If this balance were accurately perceived by the public, it could limit their appetite for disability benefit cuts. However, it seems more likely that the public thinks that increased disability benefits spending has contributed to a growth in the overall welfare bill.5 In this case, we would expect public attitudes to respond thermostatically to the perceived trend (that spending on both disability benefits and on welfare in general has risen), rather than the real, more nuanced, one.
Second, levels of spending do not simply demonstrate the generosity of the system – they also reflect changes in levels of need. When we look at trends in welfare spending as a percentage of GDP (see Figure 1), we see a very sharp rise during the 2008-2009 financial crisis – which reflects the sharp decline in employment that occurred during that period, rather than changes in entitlement to benefits. The COVID-19 pandemic saw a similar spike in spending. The effect on public attitudes of spending increases that are clearly driven by a need is unclear – with the public likely making some allowances for spending in response to a well-recognised external shock. We explore this idea later in the chapter in terms of the impact of COVID-19 on support for spending on unemployment versus disability benefits.
Third, some areas of spending appear resistant to thermostatic trends. The most notable example here is spending on pensioners. Pensioner benefits have become more generous due to the ‘triple-lock’, which ensures that state pension payments increase each year in line with whichever is the highest out of inflation, average earnings, or by 2.5%. Spending on pensioners (as a proportion of GDP) has therefore risen - and is projected to rise further. However, even if support for pensioner spending has fallen (as we discuss below), it is still strongly supported - and attempts to cut pensioner benefits have struggled. Notably, the Starmer Government’s attempt – which targeted the Winter Fuel Allowance, a universal lump sum payment to help older people with winter fuel costs – ended in failure, leading to both messy backtracking and sustained public disenchantment with the government.
Finally, the fiscal context also matters. There is currently a widespread perception that government spending is significantly under pressure, with many competing demands – including, most recently, the pressure to increase defence spending to meet NATO targets. This is likely to reduce support for welfare spending across the board, and potentially especially in areas where the public perceive spending levels to have risen.6
Taken together then, the effect of spending changes on public attitudes to spending on disability benefits are not certain. We may still expect support for benefits spending to have fallen (particularly for disability benefits, where spending unequivocally has risen) – but if this has taken place, it is not an automatic public response to how spending has changed. Instead, it would be a response to a combination of changes in benefits spending, changes in the fiscal context, and (most important of all) changes in how politicians, the media and the public at large talk about these issues.
Changing prototypes of disability
Attitudes to spending on disability benefits may have also been influenced by attitudes to disability specifically. Decades of research have shown that a central driver of people’s welfare attitudes is the extent to which claimants are perceived to deserve state support. Perceived ‘deservingness’ is based on factors like blame (whether people are at fault for their circumstances), gratitude, level of need, and whether recipients are perceived to be part of a similar in-group or a dissimilar outgroup (van Oorschot and Roosma, 2017). Within this framework, disabled claimants (and also pensioners) have often been seen as more deserving – certainly compared to unemployment benefit claimants (e.g. Blekesaune and Quadagno, 2003; van Oorschot, 2000; 2006). On this basis, we might expect cuts to disability benefits to face considerable inherent opposition.
However, a more subtle account is clearly necessary, given that many countries have successfully cut disability benefits and/or made them harder to access (Böheim and Leoni, 2018; Geiger, 2017). The way these reforms have been justified is by drawing distinctions between different groups of disabled people, and by claiming that reforms will only affect ‘less deserving’ disabled people (e.g. Geiger, 2017; Morris, 2016; Mays, 2012; Pennings, 2011; Soldatic and Pini, 2009). We have seen similar rhetoric in the UK,7 and it chimes with British public attitudes. One of the few papers to explore this issue in detail, found that people make sharp distinctions about the deservingness of different disabled benefit claimants, primarily based on a hierarchy of disability (with wheelchair users and people with schizophrenia at the top, and people with depression and asthma at the bottom), but also based on blameworthiness, medical certificates, and ethnicity/race (Geiger, 2021).
As a result, the prototypes of disability benefit claimant adopted by the public matter – something made even more important because ‘disability’ is poorly understood and covers a highly diverse group of people. The public will largely defend the entitlements of some prototypical disability benefit claimants, but be much more willing to support cuts for others. Attitudes are perhaps more likely to change according to perceived changes in the character of those claiming disability benefits, rather than because people have changed their attitudes on the principle of who should receive support. We can see this in how the Starmer Government’s attempted cuts to PIP in 2025 played out. While a number of factors contributed to widespread opposition8, news coverage emphasised how the PIP cuts would disproportionately affect older people with musculoskeletal conditions9, contradicting the government’s claim about the need to reform benefits for young people with mild anxiety/depression.
This leads us to several hypotheses about recent attitude change. Attitudes to disability benefit claimants will still be more positive than attitudes to unemployment benefit claimants, because there is very strong support for disability benefit claimants that are perceived to be ‘genuine’. However, this may have lessened over time because the prototypical disability benefit claimant has been presented as ‘less deserving’. It also seems likely that public support would be highly sensitive to the prototypical recipient of disability benefits or the prototypical target of reform. It would therefore be unsurprising if, a cut to disability benefits initially had popular appeal, but resulted in widespread opposition when it became clearer who was actually affected.
Political polarisation
The final factor which may explain changes in attitudes to spending on disability benefits is political polarisation – that is, that the attitudes of those who support different parties have moved apart – which may have affected both attitudes to welfare spending in general, and to spending on disability benefits in particular.
Attitudes to welfare in Britain have undergone two major periods of change since the late 1990s, (Geiger et al., 2023). First, there was a rise in anti-welfare sentiment through the late 1990s and 2000s (Curtice, 2010, Hills, 2002), followed by a decline in anti-welfare attitudes from the early/mid-2010s (Curtice, 2022; Kelley et al., 2018; Park et al, 2013). While these changes were partly thermostatic, it is also clear that political debate played a part – particularly the sharp rightward turn in New Labour’s rhetoric on benefits in the late 1990s, alongside newspaper headlines fixated on ‘scroungers’ (Baumberg, 2012; Hills, 2002; O’Grady, 2022; Sefton 2009). As a result, the views of Labour Party supporters towards benefit recipients became remarkably less favourable than they had been in the past, with the result that Labour and Conservative supporters were less distinct in their views. More recently, COVID-19 unexpectedly failed to increase solidarity with welfare claimants (de Vries et al., 2025), and there have been recent signs that support for welfare has begun to ebb further (Geiger et al., 2023; Curtice et al 2025).
For disability benefits specifically, there are plausible arguments in both directions about whether we would expect to see more or less political polarisation than in respect of wider attitudes towards welfare. Some have argued that political polarisation is strongest for groups of benefit claimants who are seen as less deserving, with much more consensus about the treatment of claimants seen to be more deserving, such as disabled people (Aarøe and Petersen, 2014; Jensen and Petersen, 2017; Petersen, 2012). However, we find it more plausible that attitudes towards disability benefits will be politically polarised. As discussed previously, not all disability benefit claimants are perceived by everyone to be equally deserving; it is also the case that right-wing people’s support for welfare is more contingent on deservingness cues than is the case for left-wing people (Geiger 2021). Moreover, we would expect the perceptions of prototypical disability benefits claimant (as discussed above) to be different across the political spectrum, particularly given evidence that negative media coverage affects people’s judgements about the ‘deservingness’ of disabled people that they know (Geiger 2025).
In conclusion, we would hypothesise that, while there is likely to have been some society-wide shift in attitudes towards disability benefits, there may have also been some political polarisation in attitudes. While politicians on both the left and right have been talking about the need to reform benefits for people with mental ill-health,5 different politicians have told different stories in a different way, which is likely to be mirrored in increasingly divergent public attitudes across the political spectrum.
How have attitudes to disability benefits spending changed?
Having set out our hypotheses regarding how attitudes to spending on disability benefits might have changed, we next examine four decades of data from the BSA survey10, to establish what changes have happened in practice. We begin by examining how far support for spending on welfare benefits overall has varied in relation to actual spending levels, to establish how ‘thermostatic’ this support has been. Second, we examine trends in support for spending on different groups of benefits claimants – including unemployed people, disabled people who cannot work, and pensioners – and consider how changing levels of support reflect the hypotheses set out previously. Are changes in support for different groups purely thermostatic - or is support for some groups more resilient than others, as suggested by theories of deservingness? Third, we examine data collected for the first time in 2025 on attitudes towards ‘extra cost’ disability benefits (i.e. benefits that are not linked to working status, but exist to support disabled people with the extra costs they face due to their impairments). How does the public feel about these types of benefits? What do these attitudes tell us about potential public support for changes to PIP? Finally, we examine political polarisation. Have attitudes towards welfare spending overall become more polarised over recent years? Have attitudes towards groups perceived as more deserving (e.g. disabled people) resisted polarisation, or have they followed broader trends? How does support for ‘extra cost’ benefits vary across the political spectrum, and what does this tell us about the likely prospects for reform?
Overall support for welfare spending
We begin by examining overall support for welfare spending, in order to contextualise our later findings on support for spending on specific groups of welfare recipients. To measure support for welfare spending overall, we present respondents with the following item and ask them if they agree or disagree with it:
The government should spend more money on welfare benefits for the poor, even if it leads to higher taxes.
To place these data in the context of attitudes to wider spending on the welfare state, respondents have historically also been asked the following question:
Suppose the government had to choose between the following three options.
Which do you think it should choose?
Reduce taxes and spend less on health, education and social benefits
Keep taxes and spending on these services at the same level as now
Increase taxes and spend more on health, education and social benefits
Figure 2 presents trends in ‘net support’ for these two measures, alongside actual levels of spending on benefits (as a percentage of GDP) from 1987 to 2025. ‘Net support’ represents the difference between the proportion of respondents who support more spending and the proportion who support less spending.11 Thus, a value of -15 for the first question, as in 2025, indicates that the proportion who agree that the government should spend more is 15 percentage points lower than the proportion who disagree with this view.
The data on which Figure 2 is based can be found in Appendix Table A.1 of this chapter.
Over the course of nearly four decades, we can see that the contours of public opinion are partially – but only partially – thermostatic (for a detailed discussion, see Geiger et al., 2023). In summary, between 1989 and 2009, there was a sharp decline in support for welfare spending (Curtice, 2010; Hills, 2002), even though support for wider government spending initially remained stable’, but then later alongside declining wider support. These two trends were mostly not thermostatic. There was a rise in welfare spending after the 1991 recession, and a smaller rise in welfare spending in 2003, when the Labour Government brought in tax credits. There was also a broad decline in welfare spending between 1995 and 1999 and between 2003 and 2009. From 2009 to 2019, there was a rise in support for welfare spending (until 2017, alongside a rise in support for spending on the wider welfare state) (Kelley et al., 2018; Curtice, 2022). This latter change did track thermostatically with austerity reductions in spending.
Since COVID-19, however, support for welfare spending (and wider welfare state spending) has fallen sharply. The pandemic did not increase solidarity with those reliant on government support (de Vries et al., 2025). Welfare spending has been fairly flat over this period, making this trend unlikely to be strictly thermostatic. However, as we discussed above, the public may be reacting to wider perceptions that welfare spending is increasing, allied to cost-of-living increases which make tax-supported spending less appealing.
Support for spending on different groups of welfare recipients
Support for more spending
As noted previously, support for spending on different groups of welfare recipients may vary over time, as a result of perceptions of their comparative ‘deservingness’. To explore this issue, we have regularly asked respondents the following question about benefits for six groups of welfare recipients: benefits for unemployed people, for disabled people who cannot work, for single parents, for parents on very low incomes, for retired people, and for people who care for those who are sick or disabled:
For each of the following groups, please say whether you would like to see more or less government spending on them than now? Bear in mind that if you want more spending, this would probably mean that you would have to pay more taxes. If you want less spending, this would probably mean paying less taxes.
Respondents are invited to answer for each group on a five-point scale, ranging from “spend much more” to “spend much less”, with “spend the same as now” as the mid-point. It is important to note that, prior to 2025, the question did not include the stipulation that more spending would necessitate more taxes. A further discontinuity is that, in the 2024 survey, the question was only asked in relation to “disabled people who cannot work” – not for any other group. We have presented changes in support for disabled people between 2021-2024 and 2024-2025 as dotted lines to highlight this discontinuity.
Figure 3 shows trends in support for spending on the six different groups of welfare recipients, with support for spending on disabled people who cannot work and unemployed people highlighted as coloured lines. As we have noted, unemployed people – as other primary group receiving ‘out of work’ benefits – are a key comparator for disabled people. As was the case in Figure 2, the lines in the chart depict the net support for more spending on the target group – i.e. the percentage point difference between support for more spending and support for less spending. Hence, negative values indicate that more people support less spending than support more spending.
The data on which Figure 3 is based can be found in Appendix Table A.2 of this chapter.
If we look at the most recent data in Figure 3, we see that support for more spending on all groups has fallen since 2021. We should not over-interpret this trend, not least because the statement that more spending would lead to higher taxes, introduced in 2025, would logically be expected to dampen support for extra spending. Nevertheless, the falls in net support for extra spending here match the drops shown above in support for spending on welfare overall (where there was no corresponding change in question wording) – making it more likely that the recent declines in support shown in Figure 3 reflect a genuine change in attitudes.
Focusing on benefits for “disabled people who cannot work” (depicted in purple), we see that net support for spending is down between 2021 and 2025 (from 51% to 43%). The size of this drop is similar to the falls for other groups (with the exception of retired people, who saw a smaller decline). The inclusion of data from 2024 (shown as a dotted purple line), suggests an increase in support for spending on disabled people between 2024 and 2025 (from 33% to 43%)12. However, it should be borne in mind that, because respondents were not asked about spending on other groups in 2024, an implicit comparison with groups generally perceived as less deserving had been removed, potentially making it less likely that respondents would express support for spending on disabled people who cannot work. Taking this into account, the 2025 figure appears less like a sharp uptick, and more like a continuation of the downward trend from 2021. Nevertheless, it is also possible that people were particularly supportive of disability benefits at the time of the 2025 survey (late August to early October), as this was shortly after the Government’s failed attempt to implement unpopular cuts to extra cost disability benefits (PIP).
In the longer-term, support for spending on disabled people who cannot work – and for carers, parents on very low incomes, and retired people – is now lower than at any point since these questions were first asked in1998. Yet despite this, support for spending on all these groups remains quite high. Today, only 7% of people say there should be less spending on disability benefits, and 49% say there should be more spending (even if it would probably lead to higher taxes). This is consistent with the hypothesis that disabled people’s high level of perceived ‘deservingness’ means that support for cuts to disability benefit spending is enduringly low.
It is however difficult to know exactly which disabled people are being considered when people are answering this question. The question specifically references “disabled people who cannot work”. This may be interpreted by respondents as “disabled people who genuinely cannot work” and therefore excludes any possible softening of support due to people focusing on prototypical disability benefit claimants that they believe are ‘less deserving’. It is also unclear how far people distinguish between disability benefits for people who cannot work (usually referred to as incapacity benefits – currently primarily the Limited Capability for Work-Related Activity (LWCWRA) group in Universal Credit), versus benefits focused on the extra costs of disability (PIP, DLA and Attendance Allowance)13. We discuss these issues later in the chapter.
Finally, Figure 3 shows that support for spending on unemployed people is consistently lower than for all other groups. At no point since the question was first asked in 1998 has a greater proportion of people supported more, rather than less, spending on benefits for unemployed people – whereas for all other groups, net support has consistently been positive. From this lower base, support for more spending on unemployed people has generally tracked support for more government spending overall. One exception is a rise in support from 2017 to 2021, when support for spending on other groups (and for spending overall) was generally declining. In our view, this is likely due to an emphasis during COVID-19 – both in policy and public discourse – on non-disabled people who had lost jobs due to the pandemic. Nevertheless, this did not cause levels of support for more spending on this group to rise to anywhere near the levels for the other groups, and support has fallen sharply since.
Where does the public want cuts to fall?
The above analyses present an apparent disconnect between overall levels of welfare support, and support for spending on disability benefits in particular. While support for more spending on disability benefits has trended down in tandem with support for welfare spending overall, the latter is currently strongly ‘underwater’ – with more people supporting cuts than supporting increases. In contrast, support for cuts to disability benefits remains robustly low. This raises the question of the extent to which support for welfare spending overall is driven by attitudes towards different groups of claimants. Do people primarily support cutting spending on welfare overall because they want cuts to benefits for non-disabled unemployed people specifically, or because they want cuts across a broader variety of groups?
To address this question, we focus specifically on the group pf respondents who disagree that spending on “welfare benefits to the poor” should be increased (and could therefore be interpreted as supporting a cut to welfare spending broadly conceived). Figure 4 shows, over time, the proportion of this group who support less spending on unemployed people, disabled people, single parents, parents on very low incomes, retired people, and carers.
The data on which Figure 4 is based can be found in Appendix Table A.3 of this chapter.
This chart shows that, among those who disagree with increasing overall spending on “welfare benefits for the poor”, support for reductions is predominantly concentrated on benefits for unemployed people. Consistent majorities of those don’t want benefits spending to increase support spending less on benefits for unemployed people. Levels of support for reductions among this group vary from lows of 50%-60% (in the late 1990s and mid 2000s) to highs of over 70% from the 2008 financial crisis to the early 2010s.
However, even among this group of respondents who do not want increases to benefits, backing for reductions to spending on disabled people and other groups (carers, parents on low pay, and retired people) is consistently low. Support for cuts to spending on these groups has been on an upward trajectory since around 2017, but remains low even now. Only 11% of those who oppose welfare increase also favour spending less on disabled people, while 67% support reductions to spending on unemployed people. These figures suggest that opposition to welfare spending primarily reflects opposition to spending on unemployment benefits (and to a lesser extent, single parents), with spending on disability benefits being strongly insulated.
Priorities for extra government spending
The BSA survey also measures support for spending on different potential welfare recipients by asking which kinds of benefits should be a priority for extra government spending – which helps to remove the effect of overall tax/spend preferences, Respondents are asked:
Thinking now only of the government's spending on social benefits, which, if any of the following would be your highest priority for extra spending?
- Retirement pensions
- Child benefits
- Benefits for the unemployed
- Benefits for disabled people
- Benefits for single parents
Respondents are then asked about their second priority for extra government spending.
Figure 5 depicts the proportions selecting each group as their first or second priority for extra government spending at each point at which the question has been asked since 1986 (to facilitate comparison with the trends presented in Figure 3). Reflecting the focus of this chapter, benefits for the unemployed and benefits for disabled people are presented as solid lines.
As we would expect, people are consistently and substantially more likely to prioritise extra government spending on retirement pensions and disabled people than on unemployed people and single parents. The most striking change in the most recent data is that benefits for disabled people are now the public’s top priority for extra spending, as they have been for most of the period since 2015. This is partly because of declining support for extra spending on pensions, which are still widely supported, but not to the same extent as they were in the mid-2000s (potentially a thermostatic response to the perceived increase in spending on pensioners and the ‘triple-lock’). It is also because of a rebound after the sharp increase in the prioritisation of unemployed people in 2020 (which came alongside a sharp decrease in the prioritisation of disabled people). This is consistent with both our framing of these groups as, in a sense, ‘direct competitors’ for out-of-work benefits spending, and with the interpretation of the effect of COVID-19 described above. By 2025, though, priorities had rebounded in such a way that the prioritisation of disabled people for extra spending is at an almost record high, while support for prioritising unemployed people is at a near record low.
It is worth noting that this question refers to “benefits for disabled people” in general terms (without the specification that they “cannot work”). Hence, in policy terms, it would encapsulate both incapacity benefits and extra cost disability benefits like PIP. Again however, it is not clear how much this distinction matters in terms of driving public spending preferences.
Summary
Taken together, the above results suggest that support for spending on groups that are perceived as more deserving – including retirees and disabled people – is reactive to wider conditions. Spending levels can thermostatically push support up or down, and events like the pandemic can cause fairly dramatic shifts. However, at other times attitudes seem to change in ways that are not thermostatic – potentially due to political and media discourse, as we discuss below. Irrespective of the direction of changes, support for more spending on these groups remains robustly high, supporting the hypotheses about the generally strong support for disabled people and retirees.
There is little sign that the public has an appetite for cuts. As we show above in Figure 3; even at its current nadir, 49% of respondents say they want more spending on benefits for disabled people, and 57% support more spending on retirement pensions. The proportion supporting less spending on benefits for disabled people “who cannot work” (or for retirement pensions) has never exceeded 7%.14 Even among people who oppose more welfare spending overall, only small minorities (10% or less) support cuts to spending on disabled people who cannot work. Disabled people are also the highest priority for extra welfare spending, followed by pensioners. The absence of detailed data on what people mean by “disabled people who cannot work”, means it is possible that there is support for cuts in benefits to particular groups of disabled people, – but overall, while support for welfare spending has gone down, there is also strong and continued support for spending on disability benefits.
Attitudes to extra cost disability benefits
Since reforms to PIP have been at the centre of political debate, and are currently under active review by the Government, we included three questions on the 2025 survey which were designed to measure the public’s view on the level of impairment required to qualify for extra costs disability benefits such as PIP. Respondents were informed that:
Some people with a disability get extra money from the government to help them with the additional costs they have because of their disability. How much money they get depends on the severity of their disability.
Respondents were then asked how much help someone with a disability should need with the each of the following tasks in order to qualify for extra money:
- Preparing or cooking a simple meal.
- Washing themselves.
- Making decisions about their budget
A set of response options was tailored to each of these three types of help, in ascending order of impairment. These response options were designed to align with the PIP scoring system, which gives two points for the mildest level of impairment, four points for the intermediate level, and eight points for the highest level. For example, in the case of cooking, respondents were offered the following answer options:
- They can cook for themselves, but they need to use an aid or appliance when cooking, such as a stool to sit on or lightweight pots and pans (2 points)
- They can cook for themselves, but they need someone to supervise or help them (4 points)
- They need someone to cook for them (8 points)
- N/A – nobody should get additional money to help with this task
Under the current PIP system, claimants are assessed against 10 daily living activities and two mobility activities simultaneously. In order to qualify for the minimum PIP payment rate, claimants must reach a total score of eight points across all activities (12 points would entitle them to the enhanced rate). These points can come from a severe impairment within one activity, or less severe impairments across multiple activities. The Starmer Government’s failed attempts to reform PIP did not change the scoring of the descriptors or the 8/12 point cut-offs. Instead, the Government attempted to introduce a requirement that, as well as scoring eight points in total, claimants would also need to score four points under at least one activity to qualify (rather than, e.g. getting two points under four different activities).
Table 1 presents the proportions of people who feel that a disabled person should need a particular level of help in order to quality for additional money.
| Cooking | % |
| They can cook for themselves, but they need to use an aid or appliance when cooking, such as a stool to sit on [2 points] | 19 |
| They can cook for themselves, but they need someone to supervise or help them [4 points] | 32 |
| They need someone to cook for them [8 points] | 32 |
| N/A . nobody should get additional money to help with this task | 9 |
| Don't Know/Refused | 7 |
| Budgeting | |
| They can make easy decisions about budgeting, but need help or advice from someone else with more difficult ones [2 points] | 14 |
| They need help or advice from someone else to make any decision about budgeting [4 points] | 30 |
| They need someone else to make decisions about budgeting for them, because they cannot do it themselves [8 points] | 36 |
| N/A . nobody should get additional money to help with this task | 12 |
| Don't Know/Refused | 8 |
| Washing | |
| They can wash themselves, but they need to be reminded to get washed [2 points] | 14 |
| They can wash themselves, but they need help with washing between their shoulders and their waist [4 points] | 29 |
| They need someone to wash their entire body [8 points] | 44 |
| N/A . nobody should get additional money to help with this task | 6 |
| Don't Know/Refused | 8 |
| Unweighted base | 1002 |
Bearing in mind the PIP system currently in operation and the failed reform by the Starmer government, there are three key messages we might take from Table 1. First, an overwhelming majority of people (83% for cooking, 80% for budgeting, and 78% for washing) support providing financial support to disabled people to help with the extra costs associated with their impairments. Only 3% of respondents chose “nobody should get any additional money” for all three tasks.
However, the data cannot be neatly mapped onto the failed PIP reforms proposed by the Starmer Government. It is possible to make two contrasting arguments about what Table 1 shows in this regard. First of all, it can be argued that the reforms are not supported: many people would actually be more generous than the current PIP system allows. If we look at the cooking activity for example, 19% of people think that an extra cost disability benefit should be given to people who need to use an aid or appliance when cooking, and 32% think it should be given to people who need someone to supervise or help them cook – neither of which would currently qualify someone for PIP on their own, without scoring points under other activities.
Looked at from another angle, it could be argued that the proposed reforms are supported: clear majorities of the public (73% for cooking, 78% for budgeting, and 70% for washing) think that a score of four or higher should be necessary to receive support in relation to each individual task (or that no-one should receive support at all). This may suggest broad support for a system that would be at least as strict, if not stricter, than the one the Starmer government failed to push through (which would have required people to score four points under one activity, as well as eight points overall).
Above all, these conflicting interpretations point towards a theme that runs throughout this chapter – that is the inherent difficulty of asking survey questions to measure public attitudes to disability benefits – both because of varied understandings of ‘disability’ and the highly complex nature of the assessment processes currently in place.
Political differences
Having presented an overview of public attitudes to disability benefits in general and analysed how far these behave thermostatically or reflect perceptions of deservingness, we now turn to how politically polarised these attitudes are, and whether the degree of polarisation has changed over time. Whilst we focus on the period since 1998 (as above), we should be aware of how support for welfare declined among Labour voters earlier, following a sharp rightward turn in New Labour rhetoric on benefits in the late 1990s alongside newspaper headlines that focused on ‘scroungers’, (Baumberg, 2012; Hills, 2002; O’Grady, 2022; Sefton 2009).
In examining trends in political polarisation, we primarily focus on how these have varied by people in Left and Right ‘voting blocs’. As recently as the 2019 election, 76% of voters cast their ballot for either the Conservatives or Labour; only 3% voted for the Green Party, and 2% for the Brexit Party. By 2024, the vote share of the two main parties was down to 58%, with 7% voting Green, and 14% voting Reform. Support for the two main parties has since eroded even further. As of June 2026, political polling of voting intention put Reform ahead of all the other parties on 25%. Labour and the Conservatives were tied in second place on only 19%, with the Green Party not far behind on 14%15 (although these levels of support continue to evolve). It is therefore difficult to make comparisons over time solely focusing on Labour and Conservative supporters. Instead, when examining trends in welfare attitudes over time, we organise the parties into Left and Right voting blocs as set out in Table 2.
| Left | Right |
|---|---|
| Labour | Conservative |
| Liberal Democrat / Alliance (1983-1990) | UK Independent Party (UKIP) (from 2004) |
| SDP / Social Democrat (1983-1990) | Brexit Party / Reform UK (from 2019) |
| Scottish National Party (SNP) | British National Party (BNP) / National Front |
| Plaid Cymru | |
| Green Party (from 1989) | |
| Your Party (from 2025) | |
| RESPECT / Scottish Socialist Party / Socialist Party |
Polarisation in overall welfare spending support
In Figure 6, we show trends in net support for welfare spending (as measured by our general question asking whether the government should spend more money on welfare benefits for the poor, even if it leads to higher taxes) by voting bloc, with the dotted line depicting the difference between right and left-leaning voters. On the one hand, attitudes among right and left-leaning voters largely run on parallel tracks, with support for welfare spending lower now among both blocs than in the late 1990s (the last attempt by the Labour Party to enact significant welfare reform). On the other hand, support for more spending has clearly become more polarised over time – with the gap between the two blocs widening from around 35 percentage points in 1998 to 47 points in 2025. This widening primarily occurred under the Conservative-led Government between 2010 and 2015. There has, if anything, been a slight decline in polarisation since 2023 – mostly because net support for welfare spending has fallen more among Labour Party supporters.
The data on which Figure 6 is based can be found in Appendix Table A.5 of this chapter.
While it is helpful to study trends across voting blocs for the reasons outline above, Table 3 provides a more detailed breakdown of support for welfare spending across political parties in 2025. As expected, Green and Labour Party supporters are the most in favour of more spending, Conservative and Reform voters are the least supportive, and Liberal Democrats sit in the middle. However, one notable finding is that, as suggested in the ‘Who Supports Reform’ chapter in this report (Curtice et al., 2026), Reform voters are not substantially more anti-welfare than Conservative voters. Net support for more spending is in fact substantially higher among Reform voters than among Conservative voters. It would therefore be a mistake to automatically consider Reform voters as further right than Conservatives on all issues.
| Agree | Neither agree nor disagree | Disagree | Net agree (Agree – Disagree) | Unweighted base | |
|---|---|---|---|---|---|
| % | % | % | pp | ||
| Reform UK | 17 | 30 | 52 | -35 | 823 |
| Conservative | 12 | 26 | 62 | -50 | 640 |
| Liberal Democrat | 30 | 32 | 38 | -8 | 384 |
| Labour | 39 | 25 | 35 | +4 | 1,035 |
| Green Party | 46 | 26 | 28 | +18 | 395 |
Polarisation in attitudes to benefits for disabled people
When examining the attitudes of the public as a whole earlier, we demonstrated that support for more spending on both unemployed people and disabled people who cannot work has declined since 2020. However, the public remains substantially supportive of – and very strongly opposed to cuts to – spending on the latter group. Prioritisation of disabled people for extra spending has remained consistently high, and the relative prioritisation of this group compared with unemployed people has actually increased since prior to the pandemic.
The question we address in this section is the extent to which these patterns are distinct across the political spectrum. Has wider political polarisation meant that the left and right voting blocs are farther apart in their support for both unemployed and disabled people? Or are attitudes towards disabled people, traditionally perceived as more deserving, resistant to polarisation?
Figure 7 shows the trends in net support for more spending on unemployed people and disabled people who cannot work in the left and right voting blocs, with the dotted lines illustrating the magnitude of the difference between the two.
For government spending on unemployed people, the lines depicting net support largely run in parallel, though with some notable deviations. In particular, support for more spending on unemployed people increased dramatically among left-leaning voters between 2018 and 2021, with no substantial corresponding increase among right-leaning voters. We would interpret this as a likely pandemic effect – with left-leaning voters being much more supportive of increased government help for the unemployed during and immediately after the pandemic. However, support among both groups declined precipitously afterwards – and was accompanied by a slight decline in polarisation.
The data on which Figure 7 is based can be found in Appendix Table A.6 of this chapter.
For spending on disabled people who cannot work, there has been a general increase in polarisation between the late 1990s and today. We have gone from a situation in which (unlike spending on unemployed people), support for more spending on disabled people who cannot work was almost identical among left and right-leaning voters, to a situation in 2017 when levels of support in the two blocs were almost 20 percentage points apart. Polarisation had increased by 2021 as support among right-leaning voters declined more sharply than it did among left-leaning voters. However, this trend has somewhat reversed by 2025 as support among left-leaning voters began to erode more quickly.
If we compare these levels of polarisation with those for other types of benefit claimant (shown in the chapter appendix – Table A.7), we see increased polarisation for all claimant types between 2011 and 2025 – with support for spending on pensioners showing a particularly interesting trend. In 2011, left-wing voters were more supportive than right-wing voters of spending on retired people – but by 2025 this pattern had reversed. This may be because of perceptions of these claimant groups as competitors for scarce government resources – with pensioners sometimes seen by the left as being over-rewarded (for example, through the ‘triple lock’) compared with other groups.
Overall, these results suggest that support for spending on disability benefits has polarised sharply, particularly in the late 2000s and the late 2010s. It is not the case that attitudes towards spending on disability benefits are resistant to wider trends in polarisation – even if the level of polarisation is lower than that observed in relation to support for spending on unemployed people. However, it is worth noting that, even among right-leaning voters, support for spending on disabled people remains high. As many as 40% of right-leaning voters support more spending on benefits for disabled people who cannot work, and only 10% support less spending.
As a final note, we should be clear that even within these voting blocs, Table 4 indicates that supporters of different political parties often have dissimilar views – albeit the major divergences of view are between, rather than within, these blocs. But what is most striking is that opposition to cuts to spending on disability benefits runs across party lines. Reform and Conservative voters are most likely to support spending less on benefits for disabled people who cannot work, but even here this is very much a minority position. Only 7% of Conservative voters and 11% of Reform voters support cutting disability benefits spending. The real party splits are between spending more versus keeping spending levels the same.
The picture for unemployment benefits is almost the reverse. Support for spending more is the minority position for supporters of all parties. Even among Green Party voters – who are otherwise the most supportive of additional welfare spending – similar numbers support spending less as support spending more.
| Spend more | Spend same | Spend less | Net spend (Spend more – Spend less) | Unweighted base | |
|---|---|---|---|---|---|
| % | % | % | pp | ||
| Unemployed people | |||||
| Reform UK | 8 | 31 | 61 | -54 | 547 |
| Conservative | 6 | 33 | 61 | -54 | 402 |
| Liberal Democrat | 14 | 51 | 34 | -20 | 245 |
| Labour | 22 | 48 | 29 | -6 | 716 |
| Green Party | 31 | 40 | 28 | +3 | 249 |
| Disabled people who cannot work | |||||
| Reform UK | 38 | 50 | 11 | +27 | 548 |
| Conservative | 43 | 50 | 7 | +35 | 402 |
| Liberal Democrat | 54 | 43 | 3 | +51 | 243 |
| Labour | 57 | 38 | 5 | +53 | 716 |
| Green Party | 63 | 35 | 1 | +62 | 250 |
Polarisation in attitudes to extra cost benefits
Finally, we examine whether there are any differences between supporters of different political parties in their support for extra cost benefits for disabled people. To facilitate this comparison, we calculated a total ‘severity’ score across the three items introduced earlier (cooking, budgeting, and washing) by first assigning each response category a score of 0-3 (from the least severe impairment being sufficient qualification for support, to no level of severity being sufficient). We then summed these scores across the four items to create a 0 to 9 scale,16 where higher scores indicate that people consider a more severe impairment to be a necessary qualification for support.
Table 5 shows that, while there are differences between the supporters of different parties in their attitudes towards extra costs benefits for disabled people, these are not large. Right-wing voters are less likely to support extra-cost benefits for the mildest impairments (as shown in the ( ‘1-3’ column), but they are just as unlikely as left-leaning supporters to endorse only providing support for the most severe impairments, or not providing support at all. The public is essentially united in thinking that extra cost support is sometimes warranted for disabled people who have difficulty with everyday tasks. This suggests that the conditions under which disabled people should receive such extra-costs support is currently to some extent ‘beyond politics’, at least in its current form.
| % with a severity score of… | Mean | Unweighted base | ||||
|---|---|---|---|---|---|---|
| 0 Minimum | 1-3 | 4-6 | 7-9 | |||
| Reform UK | 3.5 | 23.3 | 53.2 | 13.9 | 4.8 | 146 |
| Conservative | 9.5 | 21.8 | 49.6 | 9.9 | 4.2 | 132 |
| Liberal Democrat | 1.1 | 38.2 | 41.8 | 10.1 | 4 | 91 |
| Labour | 5 | 40.9 | 39.2 | 5.1 | 3.9 | 258 |
| Green Party | 6.2 | 38 | 44.5 | 8.5 | 3.9 | 93 |
Conclusion
In the context of a public debate about whether spending on disability benefits is ‘unsustainable’ (which includes some apparent misperceptions about how exactly benefits spending has changed), one of the key questions for this chapter was whether the public has an appetite for cuts to disability benefits. While our analysis shows that support for reducing disability benefits is enduringly low, the data do not support a simple ‘yes’ or ‘no’ answer to this question.
First, it is clear that overall support for more spending on ‘welfare benefits for the poor’ (and more broadly, for spending on education/health/welfare) has gone down sharply since 2019. While supporters of left-wing parties have always been more likely to support more spending than supporters of right-wing parties (a pattern that deepened in the 2010s), today even supporters of left-wing parties are almost as likely to disagree as to agree that spending should be increased.
If we focus on disability benefits, then this picture changes significantly. It is true that support for spending on disabled people who cannot work does change over time – partly reacting to perceived levels of spending, but also to external events and political discourse. But even after recent declines in support, 49% of the public back more spending on benefits for disabled people, far outweighing the 7% who would prefer cuts. Disabled people are also the highest priority for extra welfare spending. Attitudes to spending on pensioners show a similar pattern, with 57% of the public still supporting more spending on retirement pensions, and pensioners being the second higher priority for extra spending.
Attempts by the current government to cut spending will be in the face of 57% support among Labour voters for more spending on disability benefits, with only 5% favouring a reduction. Polarisation of attitudes towards disability benefits has increased, and backing among right-wing voters is lower – but even a Conservative or Reform Government would face very weak support for spending less on benefits for disabled people who cannot work (11% and 7%, respectively). Any attempted cuts to PIP would also be made in the context of widespread, cross-party support for extra-cost benefits for disabled people with impairments that affect their daily living. This is particularly notable given Reform’s recently announced plans to “modify or withdraw” PIP payments from large numbers of claimants.17
Although these figures would suggest significant public opposition to disability benefit cuts, there are likely to be substantial battles over who the public has in mind as a prototypical ‘disability benefit claimant’. We know from wider research that some types of disabled people are seen as ‘deserving’ by the public, and others less so (particularly but not exclusively by supporters of right-wing parties), and we can clearly see how political battle lines are being drawn around the ‘genuineness’ of some health conditions. Public attitudes to disability benefits therefore crucially depend on which disabled people are in the public’s mind, and it would be useful to both track these prototypes and to look at attitudes to pen portraits of particular disabled people (which would also help us to study how far the PIP eligibility criteria track public opinion).
Still, there is a clear message to policymakers here: public attitudes to disability benefits are a double-edged sword. There is a line that portrays disability recipients as undeserving that sometimes resonates – yet equally, cuts to disability and pensioner benefits can easily be portrayed as attacks on deserving people who are overwhelmingly supported by all parts of the public, and particularly by Labour voters. To make policymakers’ lives harder, a policy that is initially viewed positively may later be viewed negatively, after it becomes clearer who will be affected. We therefore cannot say in advance whether a particular cut to disability benefits will be supported or opposed by the public, but it is clear that this is very challenging terrain for a government to be confident of taking the public with them.
References
Aarøe, L. & Petersen, M. B. (2014), Crowding out culture: Scandinavians and Americans agree on social welfare in the face of deservingness cues. The Journal of Politics, 76, 03: 684-697.
Baumberg, B. (2012), Three ways to defend social security in Britain. Journal of Poverty and Social Justice, 20, 2: 149-161.
Baumberg, B., Bell, K. & Gaffney, D. (2012), Benefits Stigma in Britain. London, Elizabeth Finn Care/Turn2us.
Blekesaune, M. & Quadagno, J. (2003), Public Attitudes toward Welfare State Policies: A Comparative Analysis of 24 Nations. European Sociological Review, 19, 5: 415-427.
Böheim, R. & Leoni, T. (2018), Sickness and disability policies: Reform paths in OECD countries between 1990 and 2014. International Journal of Social Welfare, 27, 2: 168-185.
Curtice, J. (2010), Thermostat or weathervane? Public reactions to spending and redistribution under New Labour. In: A. Park, J. Curtice, K. Thomson, M. Philips, E. Clery & S. Butt (eds), British Social Attitudes: the 26th Report, London, Sage.
Curtice, J. (2022), Taxation, welfare and inequality: The shape of post-COVID public attitudes. British Social Attitudes: the 39th Report, London, NatCen Social Research.
Curtice, J., Sivathasan, C. & Morton, G. (2025), Repairing Britain: Attitudes towards the economy, taxation and public services. British Social Attitudes 42, London, NatCen.
de Vries, R., Geiger, B. B., Scullion, L., et al. (2025), Welfare attitudes in a crisis: How COVID exceptionalism undermined greater solidarity. Journal of Social Policy, 54, 714-733.
Geiger, B., de Vries, R., O'Grady, T. & Summers, K. (2023), The rise and fall of anti-welfare attitudes across four decades: politics, pensioners and poverty. British Social Attitudes 40.
Geiger, B. B. (2017), Benefits conditionality for disabled people: Stylized facts from a review of international evidence and practice. Journal of Poverty and Social Justice, 25, 2: 107-128.
Geiger, B. B. (2021), Disabled but not deserving? The perceived deservingness of disability welfare benefit claimants. Journal of European Social Policy, 31, 3: 337-351.
Geiger, B. B. (2025), Suspicious Minds? Media effects on the perception of disability benefit claimants. Journal of Social Policy, 54, 693-713.
Hills, J. (2002), Following or Leading Public Opinion? Social Security Policy and Public Attitudes Since 1997. Fiscal Studies, 23, 4: 539-558.
Jensen, C. & Petersen, M. B. (2017), The Deservingness Heuristic and the Politics of Health Care. American Journal of Political Science, 61, 1: 68-83.
Kelley, N., Warhurst, C. & Wishart, R. (2018), Work and welfare: The changing face of the UK labour market. British Social Attitudes: the 35th Report, London, NatCen Social Research.
Mays, J. (2012), Australia's disabling income support system : tracing the history of the Australian disability income support system 1908 to 2007 : disablism, citizenship and the basic income proposal [PhD thesis]. Queensland University of Technology.
Morris, Z. (2016), Constructing the need for retrenchment: disability benefits in the United States and Great Britain. Policy & Politics, 44, 4: 609-626.
O'Grady, T. (2022) The Transformation of British Welfare Policy: Politics, Discourse, and Public Opinion, Oxford University Press.
Park, A., Bryson, C., Clery, E., Curtice, J. & Phillips, M. (2013), British Social Attitudes: the 30th Report. London, NatCen Social Research.
Pennings, F. (2011), The New Dutch Disability Benefits Act: the link between income provision and participation in work. In: S. Devetzi & S. Stendahl (eds), Too Sick to Work? Social security reforms in Europe for persons with reduced earnings capacity, The Netherlands, Kluwer Law International.
Petersen, M. B. (2012), Social Welfare as Small-Scale Help: Evolutionary Psychology and the Deservingness Heuristic. American Journal of Political Science, 56, 1: 1-16.
Sefton, T. (2009), Moving in the right direction? Public attitudes to poverty, inequality and redistribution. In: J. Hills, T. Sefton & K. Stewart (eds), Towards a more equal society? Poverty, inequality and policy since 1997, Bristol, Policy Press. pp. 223-244.
Soldatic, K. & Pini, B. (2009), The three Ds of welfare reform: disability, disgust and deservingness. Australian Journal of Human Rights, 15, 1: 77-95.
van Oorschot, W. (2000), Who should get what, and why? On deservingness criteria and the conditionality of solidarity among the public. Policy & Politics, 28, 1: 33-48.
van Oorschot, W. (2006), Making the difference in social Europe: deservingness perceptions among citizens of European welfare states. Journal of European Social Policy, 16, 1: 23-42.
van Oorschot, W. & Roosma, F. (2017), The Social Legitimacy of Targeted Welfare and Welfare Deservingness. In: W. Van Oorschot, F. Roosma, B. Meueleman & T. Reeskens (eds), The Social Legitimacy of Targeted Welfare: Attitudes to Welfare Deservingness, Cheltenham, UK, Edward Elgar. pp. 3-36.
Appendices
Table A 1 provides the data on disability benefit and other welfare spending visualised in Figure 1.
| Table A 1. Trends in spending on working-age adult and child welfare benefits as a share of GDP, 1978-2025 | ||
| Disability spending | Other spending | |
| Year | % of GDP | % of GDP |
| 1978 | 1 | 2.4 |
| 1979 | 0.9 | 2.5 |
| 1980 | 0.9 | 2.7 |
| 1981 | 0.9 | 3.2 |
| 1982 | 0.9 | 3.5 |
| 1983 | 0.9 | 3.8 |
| 1984 | 0.9 | 3.8 |
| 1985 | 0.9 | 3.8 |
| 1986 | 1 | 3.7 |
| 1987 | 0.9 | 3.3 |
| 1988 | 0.9 | 2.8 |
| 1989 | 1 | 2.5 |
| 1990 | 1 | 2.6 |
| 1991 | 1.2 | 3.1 |
| 1992 | 1.5 | 3.5 |
| 1993 | 1.7 | 3.5 |
| 1994 | 1.8 | 3.3 |
| 1995 | 2 | 3.1 |
| 1996 | 1.9 | 2.9 |
| 1997 | 1.9 | 2.7 |
| 1998 | 1.9 | 2.6 |
| 1999 | 1.8 | 2.6 |
| 2000 | 1.8 | 2.6 |
| Table A 1. (continued) | ||
| Disability spending | Other spending | |
| Year | % of GDP | % of GDP |
| 2001 | 1.8 | 2.6 |
| 2002 | 1.8 | 2.6 |
| 2003 | 1.8 | 3 |
| 2004 | 1.7 | 3.1 |
| 2005 | 1.6 | 3.1 |
| 2006 | 1.6 | 3 |
| 2007 | 1.6 | 3 |
| 2008 | 1.6 | 3.3 |
| 2009 | 1.8 | 3.9 |
| 2010 | 1.8 | 3.9 |
| 2011 | 1.8 | 3.9 |
| 2012 | 1.8 | 3.8 |
| 2013 | 1.8 | 3.4 |
| 2014 | 1.9 | 3.1 |
| 2015 | 1.9 | 2.9 |
| 2016 | 1.9 | 2.7 |
| 2017 | 1.9 | 2.5 |
| 2018 | 2 | 2.4 |
| 2019 | 2 | 2.3 |
| 2020 | 2.3 | 3.1 |
| 2021 | 2.2 | 2.5 |
| 2022 | 2.2 | 2.6 |
| 2023 | 2.4 | 2.8 |
| 2024 | 2.7 | 2.1 |
| 2025 | 2.8 | 2 |
Table A2 provides the BSA data visualised in Figure 2:
- Net agreement (agree/strongly agree minus disagree/strongly disagree) that the government should spend more money on welfare benefits for the poor, even if it leads to higher taxes
- Net support for increasing taxes and spending more on health, education, and social benefits (support minus support for reducing taxes and spending less on health, education, and social benefits)
| 1987 | 1989 | 1990 | 1991 | 1993 | 1994 | 1995 | |
| Government should spend more on welfare – net support (ppt) | 33 | 45 | 40 | 33 | 27 | 27 | |
| Unweighted base | 1279 | 2595 | 2474 | 2561 | 2924 | 3130 | |
| Government should increase taxes and spend more on health, education, and benefits | 47 | 54 | 51 | 62 | 59 | 55 | 56 |
| Unweighted base | 2842 | 3019 | 2783 | 2905 | 2931 | 3455 | 3623 |
| 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 | |
| Government should spend more on welfare – net support (ppt) | 17 | 17 | 12 | 7 | 17 | 16 | |
| Unweighted base | 3077 | 2524 | 2448 | 2695 | 2509 | 2586 | |
| Government should increase taxes and spend more on health, education, and benefits | 56 | 58 | 60 | 54 | 46 | 57 | 58 |
| Unweighted base | 3605 | 1355 | 3136 | 3141 | 2071 | 2958 | 3092 |
| 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | |
| Government should spend more on welfare – net support (ppt) | 18 | 4 | 5 | 6 | -2 | 0 | -16 |
| Unweighted base | 873 | 2606 | 2699 | 2817 | 2670 | 2985 | 967 |
| Government should increase taxes and spend more on health, education, and benefits | 45 | 43 | 39 | 41 | 35 | 31 | 26 |
| Unweighted base | 3272 | 2142 | 2165 | 3235 | 3091 | 2220 | 1139 |
| 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | |
| Government should spend more on welfare – net support (ppt) | -9 | -11 | 2 | 4 | -9 | 8 | 5 | 17 |
| Unweighted base | 2785 | 2841 | 2851 | 2829 | 2373 | 2779 | 2396 | 3255 |
| Government should increase taxes and spend more on health, education, and benefits | 22 | 29 | 28 | 29 | 30 | 41 | 44 | 56 |
| Unweighted base | 3289 | 3305 | 3242 | 3238 | 2870 | 3262 | 2934 | 2960 |
| 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | |
| Government should spend more on welfare – net support (ppt) | 15 | 18 | 6 | 3 | 4 | -7 | -14 | -15 |
| Unweighted base | 3059 | 2633 | 3931 | 6250 | 6644 | 5128 | 3248 | 4444 |
| Government should increase taxes and spend more on health, education, and benefits | 52 | 48 | 45 | 46 | 47 | 38 | 27 | 17 |
| Unweighted base | 2878 | 3218 | 3931 | 2073 | 2415 | 1018 | 3248 | 4444 |
The data on working age and child welfare spending figures visualised in Figure 2 are the same as in Figure 1 (combining disability benefits and other benefits), and are shown in Table A.1 above.
Table A3 shows net support for more spending (support for more spending minus support for less spending) on claimant groups, as visualised in Figure 3.
| 1998 | 1999 | 2002 | 2004 | 2006 | 2008 | 2011 | |
|---|---|---|---|---|---|---|---|
| Claimant group | |||||||
| Unemployed people | -14 | -8 | -17 | -29 | -29 | -40 | -36 |
| Disabled people | 70 | 70 | 67 | 60 | 59 | 57 | 47 |
| Single parents | 13 | 11 | 20 | 16 | 18 | 20 | 8 |
| Parents on low pay | 65 | 66 | 65 | 58 | 62 | 64 | 53 |
| Retired people | 69 | 67 | 70 | 71 | 70 | 70 | 54 |
| Carers | 81 | 81 | 81 | 80 | 81 | 83 | 73 |
| Unweighted base | 3136 | 3141 | 3092 | 3193 | 3235 | 3340 | 3305 |
| 2013 | 2015 | 2017 | 2021 | 2024 | 2025 | |
|---|---|---|---|---|---|---|
| Claimant group | ||||||
| Unemployed people | -33.9 | -28.4 | -17.4 | -5.5 | -27.6 | |
| Disabled people | 49.9 | 58.4 | 63.2 | 51.4 | 33.1 | 42.5 |
| Single parents | 11.5 | 19.3 | 31.4 | 27.4 | 16.3 | |
| Parents on low pay | 54.5 | 55.7 | 61.2 | 56.2 | 39.3 | |
| Retired people | 41.4 | 42.4 | 39.6 | 35.8 | 34 | |
| Carers | 71.2 | 73.3 | 76.5 | 69 | 50.6 | |
| Unweighted base | 3238 | 3262 | 2960 | 1065 | 2918 |
Table A4 shows support for less spending on claimant groups among those who disagree (or disagree strongly) with spending more money on welfare benefits for the poor, as visualised in Figure 4.
| 1998 | 1999 | 2002 | 2004 | 2006 | 2008 | 2011 | |
|---|---|---|---|---|---|---|---|
| Claimant group | |||||||
| Unemployed people | 58 | 51 | 53 | 61 | 63 | 75 | 70 |
| Disabled people | 4 | 3 | 5 | 7 | 7 | 7 | 8 |
| Single parents | 36 | 36 | 25 | 28 | 28 | 23 | 30 |
| Parents on low pay | 6 | 4 | 7 | 8 | 9 | 7 | 8 |
| Retired people | 5 | 4 | 3 | 4 | 3 | 3 | 5 |
| Carers | 1 | 1 | 1 | 1 | 2 | 1 | 2 |
| Unweighted base | 624 | 680 | 738 | 792 | 799 | 994 | 1067 |
| 2013 | 2015 | 2017 | 2021 | 2024 | 2025 | |
|---|---|---|---|---|---|---|
| Claimant group | ||||||
| Unemployed people | 71 | 68 | 56 | 60 | 67 | |
| Disabled people | 7 | 5 | 7 | 10 | 21 | 11 |
| Single parents | 31 | 28 | 19 | 30 | 28 | |
| Parents on low pay | 9 | 10 | 7 | 15 | 18 | |
| Retired people | 10 | 9 | 10 | 13 | 13 | |
| Carers | 2 | 2 | 2 | 5 | 10 | |
| Unweighted base | 871 | 842 | 648 | 300 | 1246 |
Table A5 shows the percentage of respondents listing each benefit as one of their top two priorities for extra spending, as visualised in Figure 5.
| 1996 | 2000 | 2001 | 2003 | 2005 | 2007 | 2010 | |
|---|---|---|---|---|---|---|---|
| Target group | |||||||
| Retirement pensions | 71 | 74 | 76 | 78 | 80 | 78 | 72 |
| Child benefits | 30 | 33 | 35 | 38 | 39 | 41 | 42 |
| Benefits for the unemployed | 26 | 13 | 11 | 10 | 8 | 7 | 11 |
| Benefits for disabled people | 54 | 62 | 58 | 54 | 52 | 53 | 53 |
| Benefits for single parents | 12 | 15 | 15 | 16 | 15 | 15 | 14 |
| Unweighted base | 3605 | 3094 | 2958 | 3272 | 3191 | 3091 | 3289 |
| 2012 | 2014 | 2016 | 2018 | 2020 | 2025 | |
| Target group | ||||||
| Retirement pensions | 70 | 67 | 59 | 55 | 54 | 53 |
| Child benefits | 38 | 37 | 35 | 35 | 44 | 43 |
| Benefits for the unemployed | 13 | 13 | 13 | 16 | 24 | 9 |
| Benefits for disabled people | 58 | 60 | 65 | 63 | 52 | 63 |
| Benefits for single parents | 15 | 17 | 17 | 21 | 17 | 24 |
| Unweighted base | 3242 | 2870 | 2934 | 2878 | 3931 | 2918 |
Table A6 shows net support for more spending on welfare benefits for the poor (agree/strongly agree minus disagree/strongly disagree) for voters of left and right-wing parties, as visualised in Figure 6.
| 1998 | 1999 | 2000 | 2001 | 2002 | 2003 | 2004 | |
|---|---|---|---|---|---|---|---|
| Voting bloc | |||||||
| Left wing voters | 29 | 23 | 21 | 27 | 28 | 37 | 22 |
| Right wing voters | -5 | -12 | -12 | -3 | 0 | -8 | -21 |
| Difference (ppt) | 35 | 35 | 33 | 30 | 28 | 44 | 42 |
| Unweighted base | 2524 | 2448 | 2976 | 2793 | 2896 | 873 | 2606 |
| 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | |
|---|---|---|---|---|---|---|---|
| Voting bloc | |||||||
| Left wing voters | 16 | 21 | 9 | 21 | -2 | 4 | 7 |
| Right wing voters | -14 | -16 | -24 | -20 | -34 | -34 | -36 |
| Difference (ppt) | 29 | 36 | 32 | 41 | 32 | 37 | 42 |
| Unweighted base | 2699 | 2817 | 2670 | 2985 | 967 | 2785 | 2841 |
| 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | |
|---|---|---|---|---|---|---|---|
| Voting bloc | |||||||
| Left wing voters | 20 | 20 | 15 | 35 | 31 | 40 | 36 |
| Right wing voters | -25 | -21 | -37 | -19 | -23 | -14 | -16 |
| Difference (ppt) | 44 | 42 | 52 | 54 | 54 | 54 | 53 |
| Unweighted base | 2851 | 2829 | 2373 | 2779 | 2396 | 3255 | 3059 |
| 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | |
|---|---|---|---|---|---|---|---|
| Voting bloc | |||||||
| Left wing voters | 39 | 31 | 27 | 29 | 21 | 12 | 7 |
| Right wing voters | -7 | -24 | -28 | -33 | -40 | -44 | -42 |
| Difference (ppt) | 45 | 55 | 54 | 61 | 60 | 56 | 49 |
| Unweighted base | 2633 | 3931 | 6250 | 6644 | 5547 | 3849 | 4444 |
Table A.7 shows recent trends in net support for more spending on different groups of claimants, split by whether people support a left-wing or right-wing political party.
| 2011 | 2017 | 2021 | 2025 | 2025 vs. 2021 | 2021 vs. 2011 | ||
|---|---|---|---|---|---|---|---|
| Unemployed people | |||||||
| Left | -22 | -3 | 19 | -7 | -26 | 41 | |
| Right | -57 | -36 | -33 | -54 | -21 | 25 | |
| Difference | 35 | 33 | 52 | 47 | -5 | 17 | |
| Disabled people who cannot work | |||||||
| Left | 55 | 71 | 67 | 55 | -12 | 12 | |
| Right | 37 | 54 | 34 | 31 | -3 | -3 | |
| Difference | 18 | 18 | 33 | 25 | -9 | 16 | |
| Single parents | |||||||
| Left | 19 | 45 | 48 | 33 | -15 | 30 | |
| Right | -17 | 10 | -2 | -5 | -4 | 15 | |
| Difference | 35 | 35 | 50 | 39 | -11 | 15 | |
| Parents on low pay | |||||||
| Left | 59 | 71 | 71 | 57 | -14 | 11 | |
| Right | 43 | 48 | 37 | 24 | -13 | -6 | |
| Difference | 17 | 23 | 34 | 33 | -1 | 17 | |
| Retired people | |||||||
| Left | 54 | 42 | 34 | 28 | -6 | -20 | |
| Right | 49 | 30 | 34 | 39 | 5 | -15 | |
| Difference | 6 | 12 | 0 | -10 | -11 | -6 | |
| Carers | |||||||
| Left | 77 | 78 | 75 | 61 | -14 | -2 | |
| Right | 69 | 77 | 64 | 45 | -19 | -6 | |
| Difference | 8 | 1 | 11 | 16 | 5 | 4 | |
Publication details
Clery, E., Curtice, J. and Jessop, C. (eds.) (2026)
British Social Attitudes: The 43rd Report.
London: National Centre for Social Research
© National Centre for Social Research 2026
First published 2026
You may print out, download and save this publication for your non-commercial use. Otherwise, and apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, this publication may be reproduced, stored or transmitted in any form, or by any means, only with the prior permission in writing of the publishers, or in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the National Centre for Social Research.
-
For polling on the 2024 PIP reforms, see href="https://www.moreincommon.org.uk/research/public- opinion-ahead-of-the-welfare-vote/">https://www.moreincommon.org.uk/research/public- opinion-ahead-of-the-welfare-vote/ and href="https://yougov.com/en-gb/articles/52470-britons-are- opposed-to-cuts-to-disability-benefits-as-starmer-bows-to-backbenchers">https://yougov.com/en-gb/articles/52470-britons-are- opposed-to-cuts-to-disability-benefits-as-starmer-bows-to-backbenchers
↑ -
In fact, a major cut to health/disability-related benefits did take place, with the halving of work- related disability benefits on Universal Credit (UC). It was the separate cuts to extra-cost disability benefits (PIP) that were withdrawn – see below.
↑ -
The Milburn Review is an independent government review to investigate why a growing number of young people are not in education, employment or training (NEET). The Timms Review is the first comprehensive UK government evaluation of Personal Independence Payment (PIP) since the benefit was introduced in 2013. For transparency: the second author of this chapter (BBG) is on the steering group of the Timms Review. For this reason, we do not make any recommendations in this chapter on what should happen to PIP, nor should the analysis and discussion here be taken to represent the thinking of the Timms Review.
↑ -
Some people believe that we should describe spending trends purely in inflation-adjusted (‘real’) terms. This compares benefits spending to what people can materially afford – in terms of food, clothes, housing, and so on. But in practice, what is considered ‘the bare minimum’ in terms of living standards increases as a country gets richer and as society changes more broadly (a smartphone is now close to essential, for example). The fairest way to look at trends is therefore to examine welfare spending as a percentage of GDP, which compares benefits spending to the wealth of the country as a whole. If we do this, then spending levels are now higher than 2019, but similar to 2015 or 2008, and much lower than in 2010.
↑ -
For example, a straw poll of the Financial Times office conducted by Chris Giles around the publication of his article “No need for a moral panic about the welfare system” (October 15th 2025) found that even this relatively well-informed group felt that welfare spending had been rising sharply (private communication).
↑ -
See, for example, YouGov polling on areas of spending Britons are willing to cut in order to increase defence spending: href="https://yougov.com/en-gb/articles/54790-what-are-britons-willing-to- do-to-increase-defence-spending">https://yougov.com/en-gb/articles/54790-what-are-britons-willing-to- do-to-increase-defence-spending
↑ -
E.g. on the political left by arguing that benefits are not the best response for some groups of disabled people (e.g. the current Secretary of State Pat McFadden MP saying that “simply writing a cheque” is not enough), or on the political right by claiming that the people affected are not ‘really’ disabled at all (e.g. then-Work and Pensions Secretary Mel Stride MP’s claim that “mental health culture has gone too far”; see href="https://www.theguardian.com/society/2026/jul/15/labour-stop-writing-cheque-benefit-claimants- pat-mcfadden-welfare-reform and https://www.telegraph.co.uk/politics/2024/03/20/mental-health- culture-has-gone-too-far-says-mel-stride/">https://www.theguardian.com/society/2026/jul/15/labour-stop-writing-cheque-benefit-claimants- pat-mcfadden-welfare-reform and https://www.telegraph.co.uk/politics/2024/03/20/mental-health- culture-has-gone-too-far-says-mel-stride/).
↑ -
E.g. the presentation of the reforms was poor from the outset (with the planned cuts increased to meet a last-minute fiscal hole, undermining any moral case for reform), and was made more challenging by strong assumptions by the Office for Budget Responsibility about behavioural responses, meaning that the apparent cuts were almost double the size of the estimated cuts in practice.
↑ -
See e.g. “Older PIP claimants with bad backs to lose out most in reforms”, Times 20/5/2025, href="https://www.thetimes.com/uk/politics/article/older-pip-claimants-affected-reforms-ggsnvllsd">https://www.thetimes.com/uk/politics/article/older-pip-claimants-affected-reforms-ggsnvllsd; and “Elderly with arthritis to be hardest hit by Labour benefits cuts”, Telegraph 9/5/2025 href="https://www.telegraph.co.uk/business/2025/05/09/elderly-with-arthritis-hardest-hit-labour- benefits-cuts/">https://www.telegraph.co.uk/business/2025/05/09/elderly-with-arthritis-hardest-hit-labour- benefits-cuts/
↑ -
In these analyses we focus on data from respondents in Great Britain, excluding respondents from Northern Ireland where welfare spending is administered differently. All analyses also exclude respondents under age 18.
↑ -
For the first question, this is the difference between the proportion who agree and the proportion who disagree; for the second question, this is the difference between the proportions who select “Increase taxes” and “reduce taxes”.
↑ -
This change in net support was driven by a modest increase in both the percentage of respondents supporting more spending (from 44% to 49%) and a decrease in the percentage supporting less spending (from 12% to 7%).
↑ -
PIP = ‘Personal Independence Payment’ (for working-age people), DLA = ‘Disability Living Allowance’ (for under-16s).
↑ -
The exception for disability benefits being 2024, when 11% of people endorsed spending less. However, this is likely to be an outlier due to a change in question wording, as discussed above.
↑ - ↑
-
Responses to these items were highly correlated, and so the overall scale has good reliability (Cronbach’s α = 0.89).
↑ - ↑
