Skip to main content
Job coaching meeting

‘Work first’ can work better

More people in employment is possible, with reforms to work and welfare policy.

Written by:
Dan Tomlinson
Date published:
Reading time:
31 minutes

'Work first' is a core idea that underpins the UK's employment and welfare systems, and effective ‘work first’ orientated systems have long-term, paid employment as the primary goal for people interacting with them. This is the right objective, but with work entry rates for unemployed benefit claimants falling, health-related inactivity rising sharply and millions of people not claiming benefits locked out of support, urgent improvement is needed to reach it.

For the unemployed, the Jobcentre has become a Universal Credit (UC) monitoring service rather than an employment service. By updating DWPs own figures, we estimate that the department now spends £350 million a year on monitoring claimants, the equivalent of over half of the annual spend on work coach salaries. The estimated 13 million hours a year that work coaches spend monitoring claimants crowds out the opportunity for support which could help people towards work.

The current approach is not leading to improved employment outcomes for people in receipt of unemployment-related benefits. We find that the proportion of unemployed benefit claimants who move into work each year has fallen from 30% in 2014-2015 to 20% in 2021-2022. Coupled with falling unemployment from 2014-2015 to the start of the pandemic, this means that whereas around 500,000 to 600,000 unemployed benefit claimants moved into work each year around the middle of the 2010s, the figures immediately pre-pandemic were around 250,000 a year.

Rather than double down on a compliance-led approach to 'work first', with the recent curtailment of ‘permitted periods’ in which claimants can look for work in their preferred sector or with the doubling of the sanction rate (from 3% to 6%) post-pandemic, a better approach would be to seek to reorientate Jobcentres around building productive, supportive, work-focused relationships between claimants and their work coaches.

Unacceptably low levels of benefits for claimants required to look for work, and the dominance of monitoring and compliance for unemployed claimants, may have also contributed to higher incapacity benefit caseloads in recent years, as people with health conditions seek to find more support elsewhere in the welfare system. There are now 2.5 million people in receipt of incapacity benefits who the DWP does little to engage with, its therefore unsurprising that numbers of people moving out of incapacity benefit are so low – less than 1% move off the benefit each month. Reform in this space should enable more engagement with this group, many of whom want to work, with support that could move them closer to employment.

As of early 2023, there were 3.9 million non-students who were inactive and not claiming any means-tested benefits – a large group, but one that is largely frozen out of support offered by the DWP. This wasn’t always the case, in the late 2000s around 100,000 people a week who were not in receipt of benefits were walking into the Jobcentre and finding support. Opening up the Jobcentre to non-benefit claimants so they could also receive help when needed should be part of any plans to raise labour market participation.


For the unemployed: replace the ‘claimant commitment’ with a ‘joint commitment’ that provides work coaches and the unemployed with a set of shared goals and expectations. Work coaches should be granted flexibility to increase or decrease the frequency of interaction with claimants, to allow a focus on those who need support. The permitted period within which claimants can apply for jobs in specific sectors should be returned to 3 months.

For people claiming incapacity benefit: increase proactive engagement from DWP with this group, many of whom would like to work. Provide consistency with dedicated support workers for people who wish to make progress towards work. Longer-term reforms of the Work Capability Assessment must be informed by the experiences of claimants, and should have maximising engagement with support as a primary objective.

For people not in paid work who don’t claim benefits: people that want help to find work but aren’t on benefits should be offered help at the Jobcentre if they seek it. Work coaches should be available to point people towards appropriate employment and broader support services.


After a decade in which an expanding workforce provided a tailwind to the UK economy, employment is only slightly higher than it was in early 2020 and ‘economic inactivity’ describes 730,000 more people than before the pandemic. This is of long-term concern, with negative implications for family living standards, economic growth and the public finances if recent trends cannot be reversed.

The government recognises this and made a package of reforms to benefits and employment support a key plank of last year’s Autumn Statement. But some of these measures only acted to reinforce the growing concern that policy in this space is too often led more by rhetoric and the need to sound ‘tough’ than an honest assessment of which approaches will maximise employment.

So what can be done to increase labour market participation, and support more inactive and unemployed people into sustainable employment? Some suggest that we should reconsider the long-standing 'work first' approach that has dominated in Whitehall for the best part of 3 decades. But, here rather than giving up on work entry as the objective, a better approach would be to re-wire the DWP policies and practices so that 'work first' becomes a reality. This is about not only increasing rates of work entry but doing so in a way that leads to better job matches and, ultimately, a higher productivity economy. What happens inside DWP should complement economic growth objectives, not undermine them by maintaining low-pay/no-pay cycles for many.

Despite the rhetoric that places a strong emphasis on work entry for the unemployed, the reality is that our employment support and welfare system is no longer performing as well as it could in terms of getting either the unemployed or ‘economically inactive’ into sustained employment. The proportion of unemployed benefit claimants who move into work each year has fallen from 30% in 2014-2015 to 20% in 2021-2022, and less than 1% of people receiving the higher rate of incapacity benefit move off the benefit each month.

Instead, layers of monitoring, gatekeeping and assessments stand in the way of the DWP being able to effectively support people into work. The intense focus on conditionality as the means to maximising employment has crowded out vital support and engagement. There is clearly space to re-balance resources towards support and away from monitoring within DWP – we estimate that the DWP spends £350 million a year on monitoring benefit claimants, equivalent to just over half the amount spent on work coach salaries. 

Introducing ‘‘work first’’

'‘Work first’' orientated systems have paid, sustained employment as their primary goal. Other potential goals, like improving health or education and skills development are secondary and largely understood through the lens of employment entry.

This should translate into a premium being placed on rapid job entry for the unemployed, alongside an acknowledgement that giving claimants some space to move into work that suits their skills and interests has benefits both to individuals and to the wider economy. And because it’s only once someone is looking for job that they can get work, ‘work first’ systems should also lead to the prioritisation of support and interventions that assist the ‘inactive’ population to get to a place where they can start looking for a job.

But in its current incarnation the ‘work first’ approach has moved beyond these core principles. Unemployment benefit claimants are mandated to take any job that they can after 4 weeks unemployment, with rapid job entry becoming the only thing that matters for this group. In addition, the route to getting a job is increasingly hampered by an excessive focus on monitoring and compliance, over support and engagement. The UK spends much less than comparator countries on active labour market programmes, and in all likelihood spends much more than others on administering and policing sanctions.

At the same time, among the inactive benefit claiming population there are now 2.5 million people with ill-health or disability who the DWP does little to engage with, and who are stuck in a system designed with the intention of maximising engagement with support, but which is failing in this task.

Making ‘work first’ work

After decades of 'work first' related policies, the debate on its efficacy – and what it should mean in practice – still rumbles on. The Government continues to pursue the path of more sanctions and more conditionality for some claimants. It announced at Autumn Statement 2023 that, alongside more promising reforms, sanctions would be increased on people who don’t take up jobs fast enough.

On the other side are assembled campaigners, researchers and claimants who point to a range of evidence showing that the 'work first' system as currently formulated leads to both emotional and financial strain for claimants, and doesn’t lead to sustained employment outcomes. For some, this evidence suggests that the system shouldn’t have work as its primary goal.

But although some of this evidence is compelling, it shouldn’t lead us to give up on the principles that underpin a work-orientated welfare system. The benefits to individual families and the broader economy of work entry are clear: 89% of people in the UK in destitution are workless. Someone in a working-age workless household is 50% more likely to move out of very deep poverty if someone in their household moves into work. 'Work first' has also existed alongside the UK’s relatively high labour market participation rates. And in fact, some of the programmes which are held up as being best in class, such as ‘individual placement and support’ and ‘supported employment’ are 'work first' programmes – finding the right job for people, not just any job.

The way forward is not to give up on 'work first' as a guiding light, but to reform the system so that the objective of maximised and sustained employment can be fully realised. This involves stepping away from the rhetoric-driven policies that have in all likelihood done more harm than good, and instead focusing squarely on the reforms that can and should be made to make 'work first' effective.

The reality of ‘work first’

To best make sense of the ways in which the ambition of a 'work first' orientated system isn’t being realised we can look at how the DWP interacts with different groups of workless people, for whom in different ways things aren’t operating as well as they should. For the unemployed, where the Jobcentre has become a UC monitoring service rather than an employment service; for millions of inactive people with ill-health who have been placed in a corner of the benefits system which the DWP has, until recently, done little to engage with; and those not claiming benefits, who are no longer provided with routes to engage with employment support.

Unemployed people

Over recent months, there have been around 1.4 million people in the ‘searching for work’ UC conditionality regime at any one time. This group – which is in constantly flux as people move in and out of work – is the core focus of the Jobcentre’s attention. Those in this category have been deemed able to work and are expected to focus their efforts on getting a job as quickly as possible. But although some of the practices and policies followed within Jobcentres will undoubtedly support claimants into work, the proportion of unemployed benefit claimants who move into work each year has been falling.

As Figure 1 shows, between 2014 and 2015, 30% of unemployed benefit claimants were employed a year later but this share has been steadily declining since, reaching 20% by 2021-2022. This compares to the entry rate for people not in receipt of unemployment-related benefits, which has remained between 40% and 50% for most of the decade to 2021-2022 – a finding in-line with that reported by Eurostat analysis of the Labour Force Survey. Coupled with falling unemployment from 2014-2015 to the start of the pandemic, the falling employment entry rate for unemployed benefit claimants implies that whereas around 500,000 to 600,000 unemployed benefit claimants moved into work each year around the middle of the 2010s, the figures immediately pre-pandemic were around 250,000 a year.

When it comes to DWP and Jobcentre policy, 2 issues that are worth focussing on by those wanting to reverse this trend are: the dominance of ‘compliance’ over ‘support’, and the exaltation of 'any job’ employment entry over helping claimants find sustained employment in productive job matches. In each instance, a desire to look tough has pushed politicians and policy makers to prioritise one side of a trade-off (compliance and 'any job') over the other (support and job matching). If 'work first' is to be as effective as possible the balance between these 2 sets of trade-offs should be reset, 

The first trade-off means that for claimants, the compliance-focused nature of the system is emphasised from day one. Before receiving the benefit each claimant must visit one of the 600 or so Jobcentres around the country and sign a ‘claimant commitment’. This sets out the expectations that must be met to avoid sanctions. These are lighter for some claimants, for example for those with young children, but many claimants report that they have insufficient involvement in drawing up the commitment. A DWP survey found that only half of claimants report that their personal circumstances were considered, and the independent Social Security Advisory Committee has questioned whether commitments are effectively tailored for claimants. The opportunity for jointly agreeing goals and plans that are tailored to individual circumstances isn’t always taken, nor is the opportunity to set out the expectations on the DWP as well as claimants themselves.

After signing their commitments, most unemployed claimants will then have weekly 10-minute appointments with work coaches for the first 13 weeks of their claim, which can reduce to fortnightly thereafter. But the amount of time that work coaches give to each claimant is limited: DWP’s planning assumption is for 125 claimants per work coach. Some of the small amount of time that a work coach has with each claimant will be spent, particularly early on in claims, on assisting claimants with issues relating to their benefit claims, but the bulk of it is spent on either monitoring claimants or supporting them to find a job.

The line between these 2 roles is of course blurred, but the DWP estimated in 2016 that it spent £204 million annually on ‘monitoring’ alone – which is equivalent to just over half of the amount we estimate it spent on work coach salaries in the same year. Uprating this figure to account for higher salaries and caseloads implies that the DWP may now be spending £350 million a year on monitoring, the equivalent of 13 million hours of work coach time. This must surely crowd out space for tailored support.

Further, there are also questions as to the effectiveness of a monitoring-led approach. Evidence suggests that the relationship between work coaches and claimants can be experienced as little more than a tick-box exercise, with both sides incentivised to make sure claimants are doing just enough to make sure the claimant commitment is met, rather than working together towards the shared goal of finding the claimant a job that is a good fit for them.

Sanctions, or the threat of sanctions, are a key component of the DWP's approach but this means that from the get-go the threat of sanctions for ‘non-compliance’ hangs over every interaction between a work coach and unemployed person, further limiting the prospects for a productive partnership between job seeker and Jobcentre from the start.

Even if work coaches spent less time checking up and more time job coaching, the primacy of sanctions as an enforcement device risk souring the relationship from the start. As Tom Pollard has articulated, few line managers would use their first meeting with a new recruit to set out potential scenarios that could lead to them being sacked. Although the bringing together of benefit administration and employment support was done with the best of intentions, it does seem misguided that we have designed a system in which the person whose job it is to help you find work is also involved (even if via a DWP ‘Decision Maker’) in taking away your income when you don’t keep to your claimant commitment.

The second unbalanced trade-off is that between the ‘any job’ approach to employment entry and the prospects for good, productive, job matches. Here the claimant commitment is also relevant, as it details precisely what is expected of jobseekers. Until recently jobseekers could spend the first 3 months of their claim looking for roles that were similar to their previous job. But as of 2022 this period has been reduced to 4 weeks, after which almost all claimants are expected to apply for any job that pays at least the minimum wage and is within a 90 minute commute. Since 2013, UC regulations have required too that claimants spend 35 hours a week applying for jobs.

Ministers have increasingly framed this direction of travel as an ‘any job’ approach to employment support, with less emphasis placed on the broader benefits of people finding roles that are a good fit for them and for the economy. This is in the context of an economy where many of those in low-paid jobs struggle to progress at the same time as DWP is now separately increasing requirements on in-work claimants to do so. Rather than aiming for sustained employment and better jobs (through wider labour market reform) the priority seems to be ramping up conditionality both on people out of work, so they get ‘any job’, and people in-work, so they take more hours.

There are trade-offs here. More time spent looking for a better job match implies more time in unemployment and higher benefit bills for the DWP in the short-term, but as recent work has shown, a tighter focus on encouraging people into work doesn’t necessarily lead to good work outcomes – even if it may lead to people being more likely to get a job quickly. In the UK context specifically, the IFS have found that extending conditionality to more single parents didn’t save the taxpayer money or lead to single parents entering good quality work, but did act to reduce the wellbeing of those on benefits.

An excessive focus on ‘any job’ is also limiting the Jobcentre’s ability to engage effectively with employers. Rather than fulfilling its traditional labour exchange role and helping form good matches between employees and employers, modern Jobcentres have become factories for large volumes of job applications of varying quality which employers are reluctant to wade through. The Institute for Employment Studies found that just 1 in 6 firms used the Jobcentre over the last 2 years to recruit. Employer engagement, though, is crucial if jobs are to be designed differently so more people can take them up and remain in them.

We should be clear-eyed too that the current approach may also be undermining wider economic goals. Both the Conservatives and Labour are prioritising their plans for growth in this election year, but any party concerned about ending the UK’s economic stagnation should care about what’s happening inside the DWP. An 'any job' approach could well act to undermine goals to increase productivity if it is facilitating low-wage employment and low-pay/no-pay cycles in the UK labour market.

People with limited capability to work

Separately from unemployed people, there are many people of working-age who have disabilities or health problems that limit their ability to work. In fact this group has become much larger as a share of people out of employment over time. For example: in mid-1996 the number of unemployed people was broadly the same as the number of people inactive with ill-health (2.3 million in both cases), but by the end of 2023 there were more than twice as many people inactive with ill-health as unemployed people (2.8 million compared to 1.3 million).

It’s right that the state provides an adequate and stable income for this group and doesn’t seek to force people who can’t work into employment. But it cannot be right that there are now 2.5 million people – and rising – who are not working with ill-health and receive incapacity benefit but receive little or no engagement from DWP that could help them move towards employment, particularly when there will be many thousands of people within this group would like to move into work or expect to do so in the future.

These 2.5 million are people who have been through the Work Capability Assessment (WCA) and been placed in the Limited Capability for Work and Work-related Activity (LCWRA) group, rather than the Limited Capability to Work (LCW) group. Those in the LCWRA group are provided with an additional £390 a month and expectations on their benefit claim are reduced to zero. By contrast, people placed in the LCW group receive no additional benefit income (after the LCW element, now worth £146 a month, was abolished for new claimants in 2017) and are required to carry out some activities, such as meeting their work coach at least every 3 months, to prepare for work.

The LCWRA group has increased in size from 1.7 million people in May 2019 to 2.5 million in May 2023. And over a longer time period, the proportion of people who are placed in this category as a result of a WCA has grown from 21% in 2011 to 65% in 2022.

There are a range of plausible explanations for this higher caseload, with a combination of the following factors all likely to have played a role: rising ill-health in the population; an easing of the WCA assessment regime; increased cost of living pressures colliding with the inadequacy of the basic rate of support for non-working people; and increased use of sanctions and conditionality for claimants with work search requirements.

The latter 2 factors are a direct consequence of the design of the benefits system for people expected to work. The basic rate of out-of-work benefit broadly kept track with prices rises from 2000-2015 but between 2015 and 2019 benefits were frozen and then benefit uprating lagged behind prices as the cost of living crisis escalated. This means that by September 2023, benefits had lost 12% of their value compared to April 2015. These sorts of decisions mean that pressure then builds on other parts of the system that provide more security, such as the LCWRA group for people with health conditions. Just as the current approach has caused unemployed claimants to disengage from the low-income/high-conditionality system and go through the motions of a tick-box conditionality regime, it has also pushed people with ill-health out of the system of support altogether.

Stepping back, we have a situation where the steady-state is one in which nearly 3-in-10 working-age adults who are out of work have been labelled incapable of work by the DWP, up from 22% pre-pandemic. The state has classed every one of these people as incapable of work, and so it is unsurprising that there is little engagement with support, and movement out of this group is so low, less than 1% move off the benefit each month. This situation isn’t right or sustainable. If an employment orientated system is to be effective for claimants with ill-health or disability, but with the potential to enter employment, then reform which opens the possibility of a system which engaged with more claimants must be the answer. One-fifth of those in the LCWRA group (or its equivalent in the previous system) have a desire to work and think they could work in the future, so the potential for higher work entry is there if it can be unlocked by the right reforms.

The Government recognises this and has committed to spending an extra £400 million a year over the next 4 years on additional work coach support for people who have been assessed as having ‘limited capability to work’, and on expanding a new voluntary ‘Universal Support’ employment support programme. And in the medium term, the Government plans to abolish the WCA entirely, with work coaches having discretion over the conditionality placed on claimants with disability or ill-health. Under these reforms eligibility for Personal Independence Payments (PIP) will determine receipt of the £390 LCWRA element. This big change comes with many risks, and there are alternative approaches – which we explore further below and in future JRF work.

People not working and not claiming benefits

The largest group of working-age adults out of work are those not claiming benefits: as of early 2023, 3.9 million people who are inactive (and not in education) don’t claim any means-tested benefits, 61% of the total non-student inactive population.

Many in this group are either early-retirees or have high wealth, and engagement with an employment service is unlikely to be the thing that helps them back into work. But, for example, JRF analysis of the Wealth and Assets Survey suggests that in 2018-20 there were 1.1 million non-students aged 16-49 who are inactive, not claiming means-tested benefits, and with below average wealth. This is a lot of people.

But despite this group’s size, and the potential for employment support to help some of this 1.1 million into work, they are largely excluded from the Jobcentre as it is currently designed. As discussed above, the Jobcentre has now effectively become a UC monitoring service and as such if someone not on UC arrived at its front door they may well be turned away, with no offer of support available. This contrasts with the Jobcentre of the late-2000s when, for example, around 100,000 people not in receipt of benefits were turning up each week at the Jobcentre without an appointment but were provided with a conversation with a Jobcentre staff member.

The lack of provision for those outside the benefit system has been recognised by the Government who are planning to make the Universal Support programme (discussed above) available to people not claiming benefits. This is a start, but it will likely take more than a programme bolted onto the side of the existing system to make a notable difference. 

What would a renewed ‘work first’ approach look like?

First, it’s worth saying that reform of the DWPs policies and practices for the 3 groups discussed above – the unemployed, the inactive with ill-health and the inactive not claiming benefits – are one strand of what’s needed to improve UK participation rates. Other changes are needed, from employers taking more steps to make roles inclusive for people with disabilities, to policies to improve the health of the population.

And if the reforms outlined below are to be effective then action will also be needed in 2 important cross-cutting areas: employment support and job quality.

The focus right now, from both parties, is often directed at the ‘how’ of employment support. Should it be devolved to regional or local authorities? What type of support works best? Which type of organisation should deliver it? These are good questions, getting the mechanisms, accountability and finance to the right places to deliver the right support is important. But hanging over all of them is the question of scale, there is just very little employment support delivered now compared to previous years. In the 5 years preceding the financial crisis the DWP spent an average of £1.3 billion a year on employment support. By contrast, in the 3 years prior to the pandemic it spent just £0.3 billion. This is in the context of large cuts to the DWP’s day-to-day spending budget, which halved in real terms between 2009-10 and 2019-20. Although reforms to the system to increase engagement with it would be beneficial without additional funding, to really unlock the potential of a reformed system will require additional funding for employment support.

In a similar vein, improving the quality of low-paid work in the UK should also be a priority – to make moving into employment a more attractive and sustainable proposition for people not working. The introduction and ambition of the National Living Wage has substantially reduced low-hourly pay in the UK, but despite some small progress in the last year (such as the introduction of a right for workers to request predictable shift patterns) there have been no big steps forward in labour market regulation and job quality in the UK for too long now. Lower-paid work has become more intense, more stressful and less secure in terms of the contracts or stability it offers, and yet the pace of reform to improve job quality is positively glacial. This too must change.

Unemployed people

The overarching objective for reform to support increased job entry by people who are unemployed should be for work coaches and Jobcentres to spend more of their time effectively supporting people to find sustainable employment, and less time either administering benefits or policing adherence to claimant commitments as currently designed. Putting conditionality in the background and bringing support to the fore.

To begin, the first conversation between claimants and work coaches should be designed around a ‘joint commitment’. This would be joint in 2 respects, both different from the current ‘claimant commitment’. First, it would set out the expectations on both claimants and the Jobcentre. Yes, what claimants should do to prepare for and search for work. But also detailing the options for support that claimants could engage with, be that high-quality personalised employment support programmes, a budget to be spent on items or training that would prepare someone for work they would like to do, or a guarantee of a paid job if you make a serious and persistent effort to find a job but one doesn’t materialise. This should be framed not as a set of things that each side ‘must’ do, but as a set of shared goals with sustained employment as the ultimate aim.

Second, the expectations for claimants should wherever possible be jointly agreed rather than imposed in a one-size-fits-all approach. If a claimant thinks they’ll manage 15 hours a week of applying for jobs, why put in their commitment that they must spend 35 hours doing so just for claimant and work-coach to engage in a tick-box exercise around whether this has been completed? Or if a claimant is keen to re-train, could this be acknowledged from the get-go in the commitment, both in terms of what training they would like to do and the extent to which the DWP will fund it? The imperative should be to start with the objective of agreeing a joint plan, rather than the imposition of one-sided expectations which are then forensically monitored in future conversations, at the expense of productive discussions of mutually agreed goals.

Work coaches should also have – and use – the flexibility to increase or decrease the frequency and nature of interactions with claimants early in their claim. The default is currently a weekly or fortnightly 10-minute appointment, but 10 minutes often doesn’t provide enough time to do more than perfunctory employment support. Currently, claimants with additional needs and mental health issues are often given longer appointments, but more flexibility on frequency of contact would free up space for some claimants to receive additional time from work coaches (or for the existing extended appointments to be longer/more frequent). For example, when a claimant is applying for jobs in a sector which they are keen to work in, and the prospects of them entering work without assistance seems high – why is the DWP checking-in on them every week? Why not loosen in-person contact requirements and monitoring of activities on these claimants to provide more time for contact with those who are less proactive in their work search?

What about when claimants don’t engage with the system or with reasonable requirements? To start, this should happen less often in a system where the expectations on claimants are set by goals that they had a hand in shaping, reducing the need for sanctions. But the big problem with sanctions is not that they exist – it’s right that with benefits comes a responsibility to look for work if you can – but that they are so punitive. There is growing evidence that low incomes can make employment entry more difficult, and as such loss of income that sanctions cause is likely to be making it more difficult for those who do want to find work to be able to do so.

Currently, when someone is sanctioned, they'll usually lose all their standard allowance for somewhere between 7 days and 6 months, depending on severity. This is a not a sensible way to design a system with engagement at its core. The financial difficulty that sanctions can cause must surely set people back a long way in their ability to effectively search for work. Particularly troubling is that sanction rates are now double pre-pandemic rates, with over 6% of ‘sanctionable claimants’ being subject to a sanction at any given point in 2023 – compared to less than 3% in 2019. If more sanctions were the answer, work entry would have surged of late.

Longer sanctions incentivise disengagement as many claimants will see little point in returning to the Jobcentre if they aren’t receiving some or all of their benefit. As such, the DWP should consider reform of sanctions so they vary in terms of percentage of the standard allowance deducted, instead of by length. Alternatively, sanctions could be lifted once someone re-engages with the DWP, rather than remain in place for the initially specified period.

When it comes to unwinding the ‘any job’ focus, one simple but important policy change would be to increase the ‘permitted period’ within which claimants can restrict their job applications to sectors of interest back to 3 months. Work coaches should be given the discretion to extend this period by a further 3 months if they think a claimant has a reasonable chance of getting a job in their preferred sector in a role that they are more likely to have sustained employment in.

Policy makers should consider too whether outside of ‘permitted periods’ claimants should have to apply for any job whatsoever. If people express a reasonable preference for a particular type of contract or job security, it should be considered when drawing up joint commitments. If someone doesn’t want a zero hours contract then mandating them to take up a job with that insecurity is not the route to assisting people into a job match that will last.

Finally, and zooming out, there is a question to be answered about the structure of the Jobcentre. The reforms outlined in this section can happen within the Jobcentre as currently formulated, but there is clearly a case for going further and re-focussing the Jobcentre around its traditional employment service role, separating out the DWP administered social security from a Jobcentre returned to its former status as an Executive Agency (it was brought ‘in-house’ in 2011). Issues with benefit claims could be sorted by dedicated ‘UC advisors’ so work coaches can focus almost entirely on support and guidance, rather than getting stuck in benefit administration. This could be coupled with a rebranding to clearly demonstrate to benefit claimants that the operation and culture of the Jobcentre was undergoing a transformation.

People with limited capability to work

The set of proposals outlined above would help to rebalance the relationship between the unemployed and the Jobcentre, away from compliance monitoring and the ‘any job’ approach towards a system that supports people into finding sustained employment.

Reforms in this vein would also support 'work first' objectives by unwinding some of the changes that have pushed people away from engaging with the DWP, either through not claiming benefits at all or by moving into categories that do not mandate work search related activities. We should be designing a system that people want to engage with, not one that instils fear and mistrust.

When it comes to people inactive due to sickness or disability the objective of reform to boost participation should be, first, engagement with activities and support that can move people who can or want to work in the future closer to employment.

This is about both the quality and quantity of support available. On quality, the Government is moving in the right direction with its ‘Universal Support’ programme which will extend personalised employment support to more of the people in the LCWRA group. It’s also welcome that Government is also planning additional funding for work coaches so that they can spend more time to engaging with this group. On quantity, it is welcome that Autumn Statement 2023 announced the doubling of the capacity of the Universal Support programme to 100,000 people a year by 2025-26 and the expansion of Individual Placement and Support. This though is in the context of DWP employment support funding having fallen by 75% over the 2010s. The Government should be rapidly evaluating these interventions and, if they are seen to be effective, increasing the funding for them.

But the knottier question, which underpins the potential rationale for larger reform of the design of the WCA system, is how to engage people with such support and interventions. Currently, the 2.5 million people placed in the LCWRA group have very little engagement with any support that could help them towards employment, and yet a third of this group are interested in receiving support.

The evidence suggests that one of the biggest barriers to engagement currently is that people do not know that support even exists, so at minimum Jobcentre staff should be required to speak to claimants at regular intervals about what is available (this should be included as part of ‘joint commitments’ discussed above). There should also be an expectation on the Jobcentre that if someone in the LCWRA group expresses an interest in being supported to improve their health, work-readiness or employability that the appropriate programmes or connections are made available. DWP should also seek to provide claimants who are seeking to move into employment with a dedicated support worker who can help them navigate job entry and be relied upon to assist if someone does have to leave work.

With better information and more proactive DWP engagement, alongside the recent clarity from Government that benefits will not be taken away if people find work, we may start to see moves into work from the LCWRA group increase. But the question of whether the broader WCA system is fit for purpose remains. Government proposals for a radical overhaul will now be taking shape inside the DWP and any reform should be informed by the experiences of claimants. To that end later this year JRF will be publishing work that draws on the experiences of incapacity benefit claimants to set out how we think reform should proceed in this space. In broad terms, the goal is to design a system that seeks to maximise the chances that employment is seen as an achievable and desirable goal for as many people as possible – rather than labelling people as incapable of work and cornering them in a part of a system in which the prospect for engagement with support is limited, even for those who would like to work or expect to do so in the future.

People not working and not claiming benefits

Finally, there needs to be an offer for people not currently on benefits but out of work. Traditionally, HM Treasury has been concerned about the ‘deadweight loss’ of focussing on people outside the benefit system, but of late both Government and opposition have recognised that there is merit to providing support to non-claimants too.

There’s a strong case for going further and opening up the Jobcentre or an equivalent organisation to all people out of work. Reformulating the Jobcentre as a stand-alone employment service, as discussed above, should not just involve changing the approach to benefit claimants, but also act as a reset in the approach to people not receiving benefits. People that want support to work but aren’t on benefits should not be turned away, but instead be offered help if they need it. That could be digital assistance, possibly just access to computers to apply for work, but should also involve the possibility of face-to-face support. Work coaches should be available to point people towards appropriate employment support programmes and other services that can assist people to move closer to employment.


'Work first' is the right approach, for individuals, their families, and the wider country. The more people in sustained employment the better. But with work entry rates for unemployed benefit claimants falling, health-related inactivity rising sharply and millions of people not claiming benefits locked out of support, it’s clear that the system isn’t operating as well as it should.

Reform to put support for unemployed people, and engagement with economically inactive people, at the heart of the system should lead to improved employment outcomes for claimants, lower benefit bills for the government and, if more care is taken to get people into sustained employment, not only a higher participation economy but a high productivity one too. This is what making 'work first' effective looks like, and what it could achieve.

Music producer in a music studio sat on a chair.

This briefing is part of the work topic.

Find out more about our work in this area.

Discover more about work