Independent report

Keep Britain Working: Technical Note

Published 5 November 2025

This document describes the methodological approach to key estimates produced for the Keep Britain Working Independent Review.

Calculating the current cost of ill-health

The annual cost to employers of poor workplace health is enormous – around £85 billion. A summary breakdown is set out below:

• £10 billion – Direct costs from Statutory and Occupational Sick Pay
• £47 billion – Lost output when employees cannot work
• £21 billion – Lost productivity from presenteeism
• £7 billion – Conflict resolution, litigation, and recruitment

The direct costs of sick pay to employers are estimated using the same methodology as set out within the Statutory Sick Pay (SSP) Regulatory Impact Assessment^{[footnote 1]}.

The lost output when employees cannot work due to sickness is the mid-point of the estimate of the cost of working age ill-health^{[footnote 2]}.

The estimated lost productivity from presenteeism uses estimated days lost due to presenteeism per worker from the UNUM study that suggests between 4 and 9 days are lost per year per employee^{[footnote 3]}. This is then combined with the total number of UK employees, and the average wage per week assuming productivity is equal to wage to get an estimate of the total cost of presenteeism.

The total cost of health-based conflict is estimated from government statistics on the number of employment tribunals^{[footnote 4]}, combined with Advisory, Conciliation and Arbitration Service (Acas) figures^{[footnote 5]} of the total number of jurisdictional complaints that were health or disability related.

Calculating the costs and benefits of Workplace Health Provision

The Workplace Health Provision (WHP) is envisaged to be a new type of work-focused, non-clinical service which supports both employers and employees across all stages of the Healthy Working Lifecycle. WHP is expected to act as a trusted case manager, and facilitator, helping employees and employers navigate the often complex and fragmented relationship between work and health or disability.

Unlike clinical roles, WHP would not diagnose or treat medical conditions. Instead, it would focus on how best to support a disabled employee, or employee with a health condition, to thrive in work, as well as considering an individual’s ability to work while ill or recovering, in partnership with the employer. The WHP is expected to provide personalised guidance, and to co-develop structured, functional plans such as Stay-in-Work and Return-to-Work, Plans. Crucially, WHP operates as an independent, third-party provision, positioned outside the employer’s direct management structure.

We estimate the cost of Workplace Health Provision for all target cohorts and corresponding benefits, in line with the recommendations from the Keep Britain Working Independent Review.

Results

The specifics of how WHP would be delivered to and implemented in workplaces would be determined through testing during the Vanguard Phase. While there are several different examples of other types of workplace health services (for example, Occupational Health, EAPs, WorkWell) which contain elements of the envisaged WHP service, these do not provide direct comparators. Modelling the costs and benefits of WHP at this stage is challenging as it relies on a number of broad benchmarks and assumptions. There are significant uncertainties around how WHP would compare to other types of workplace health services, and we would look to develop this evidence base further as we gather more insight throughout the Vanguard Phase.

On this basis we have also not taken the full Green Book methodology to appraising policies but estimate monthly and annual individual and aggregate costs and benefits of WHP at full take-up levels, i.e. that all UK employees have access to some WHP in their workplace.

The service specification used to develop the analysis is based on current understanding of the WHP but will change and evolve through vanguard testing. We have assumed a “core” set of services, i.e. the minimum expected requirement. There will be many businesses that choose to go above this level of service.

As benefits are presented on an annual basis, they do not fully consider longer-term impacts of keeping someone in work – for example, if one early intervention is able to keep someone in work for the next two or three years.

Summary results

Costs (total)[footnote 6]	Lower Estimate	Central Estimate	Higher Estimate
Total annual cost	£2.0 billion	£4.1 billion	£6.1 billion

Direct

Annual Benefits	Lower Estimate	Central Estimate	Higher Estimate
Employers	£2.8 billion	£3.9 billion	£7.4 billion
Government	£0.14 billion	£0.2 billion	£0.3 billion
Employees	£0.3 billion	£0.5 billion	£0.9 billion
Total	£3.2 billion	£4.6 billion	£8.6 billion

Potential additional benefits from substantially improving the system

Annual Benefits	Lower Estimate	Central Estimate	Higher Estimate
Employers	£3.8 billion	£3.8 billion	£5.2 billion
Government	£0.8 billion	£0.8 billion	£1.5 billion
Employees	£1.5 billion	£1.5 billion	£2.6 billion
Total	£5.8 billion	£5.8 billion	£9.3 billion

Annual Benefits	Lower Estimate	Central Estimate	Higher Estimate
Total Benefits	£9.0 billion	£10.3 billion	£17.9 billion

A further breakdown of the benefit calculated here can be found in the diagram at the end of this document.

Methodology

We have taken a bottom-up approach to estimating both costs and benefits. These are built using “cohorts” of employees based on the Healthy Working Lifecycle described in the main report. We have made assumptions around the size of each cohort of employees based on latest data available.

Phase	This is when	Volume assumption (annual)
Healthy in work	The employee is generally healthy and working effectively.	Around 19 million employees.
Unwell in work	The employee is struggling with health issues but still at work.	Around 11 million employees.
Absence and return to work	The employee is signed off but expected to return.	Up to 4.4 million employees.
Redeployment/exit and re-employment	The employee can no longer perform their role.	120,000 people

Costs of Workplace Health Provision

The average monthly cost of Workplace Health provision is likely to be in the region of £5 and £15 per employee (£60 to £180 per annum) – this is based on benchmarks we have from providers and validated using a bottom-up methodology. The top-down figures are used to avoid spurious accuracy. On an annual basis this is equivalent to c. £2 billion to £6 billion across all employees. It is expected that costs will primarily fall to the employer, although there may be a level of direct or indirect cost sharing with employees.

Methodology and assumptions

Estimates have been built from the estimated cost of providing services and treatment to each of the cohort groups described above. This includes the cost of labour of the WHP including time spent on case management and any further support on adjustments, Return to Work, redeployment, etc., cost of referrals and cost of limited early treatments.

Benchmarks are used to determine the per-hour cost of time invested by the WHP in activities within the scope of provision, and the costs of referrals and any early treatments offered (for example, counselling and physiotherapy). These are paired with estimates of time required for each cohort, varying in the intensity of support required. At this stage of the analysis, non-salaried costs/overheads like office rent, office equipment used, etc. have been included as 20% of the total costs of delivery.

This is likely to overestimate the long-term cost reduction potential for the WHP, as some benchmarks will double count profit margins, it does not include the benefits of economies of scale, and it does not account for potential efficiencies related to digital and other innovation.

Benefits of WHP

Analysis estimates that the likely direct benefits as a result of implementing the WHP would be between £3.2 and £8.6 billion overall per year. This is estimated to be between £130 and £350 per employee. This is the overall benefit per employee, including business, government and wider societal benefits. The split across business, government and societal are set out in the table below.

Benefit split by recipient	Lower estimate of benefit per employee (annual)	Central estimate of benefit per employee (annual)	Higher estimate of benefit per employee (annual)
Business	£115	£160	£305
Government	£5	£10	£10
Individual/Society	£10	£20	£40

Methodology and assumptions

Benefit estimates are considered for each cohort individually based on relevant evidence from similar interventions. We also consider some wider indirect impacts including the preventative benefits of reducing the likelihood an individual needs to go on long-term sick leave (rather than, for example, staying in working with additional adjustments), and also the likelihood that some of those currently economically inactive due to long-term sickness would be able to return to employment more easily.

Direct Impacts

Redeployment/exit and re-employment

An estimate for the cohort of employees that would access the redeployment element of WHP is assumed to be the number of employees who fall out of work each year due to a long-term sickness absence. From DWP Employment of Disabled People publication, in 2024 this was estimated to be around 120,000^{[footnote 7]}.

We have assumed a central assumption that 5% of in-work employees would be prevented from falling out of work as a result of the WHP, this is based on evidence from the work aspirations study suggesting 5% of health and disability benefit customers felt they could work given the right job or support was available^{[footnote 8]}. We have also modelled a low and high estimate of 3% and 7%^{[footnote 9]} respectively.

This volume of employees prevented from leaving work are then assumed to have a number of benefits as a result of this, set out below:

• Exchequer benefits (as a result of reduced reliance on UC/other benefits, increased fiscal contribution and reduction in NHS costs as a result in moving an individual with health conditions from inactivity to activity^{[footnote 10]})
• Employee benefits (as a result of increased earnings and wellbeing impacts of remaining in work)
• Employer benefits (as a result of increased output and prevented recruitment costs from an employee remaining in work)

These benefits are monetised using the DWP Social Cost-Benefit Analysis model^{[footnote 11]}, estimates for full and part-time medium pay of employees combined with labour vs capital share of Gross Value Add (GVA). Combining these assumptions, the overall benefit of preventing this cohort of employees from falling out of work to be between £400 million and £900 million per year.

Absent and return to work

An estimate for the cohort of employees that would access the absent from work element of WHP is assumed to be the total number of fit note episodes per year. A fit note episode refers to the entire period of sickness that one or more fit notes cover for an individual, and therefore in this case is assumed to equate to number of employees^{[footnote 12]}. We then use the NHS Fit Note Statistics to look at the number of fit notes in a fit note episode to distinguish between a cohort more likely to have short term absences versus long term absences.

Short-term absences

We use the number of fit note episodes with only 1 fit note as a proxy for the number of employees with a short-term sickness absence. From the latest NHS fit note statistics^{[footnote 13]} this is around 2.2 million employees.

Anecdotally^{[footnote 14]} we know that around 2 minutes of a GP appointment is spent discussing/signing the fit note. We have assumed this time would be saved as our central assumption if fit notes are no longer required through GPs and instead routed through WHP. We have also assumed that in some cases, the full 10-minute GP appointment time would be saved. We have assumed this to be around 7% of cases, based on the 7% of Fit Notes that are currently ‘May Be Fit For Work’.

This benefit is monetised using the average cost of a GP appointment^{[footnote 15]}.

Overall, this is estimated to save between £21 million and £52 million per year.

We have also assumed employees in this cohort see a reduction in sickness absence as a result of WHP intervention. We have used the NHS fit note statistics to determine the average duration of Fit Note episodes under 4 weeks to be around 14 days. Our central assumption is an 8% reduction in sickness absence based on evidence from a Musculoskeletal Study of Work and Pain (SWAP) trial^{[footnote 16]} and judgement based on intensity of support.

We also model a reduction in presenteeism to all employees within this cohort. From Unum’s Health, Happiness and Productivity publication^{[footnote 17]}, 35% of employees who have access to health and wellbeing support provided by the employer would be more likely to report being happy at work. We assume that this is the proportion of employees within this cohort who would reduce the amount of times they go into work when poorly. We then take a mid-point of estimates from the Unum study of those who reported presenteeism, what levels of productivity individuals were working at. This was estimated to be around 30%. We also use Unum’s study to understand how many potential current presenteeism days could be replaced by 100% productivity days. Our central assumption is 1 day with a higher assumption of 4 days.

Together this means 35% of employees could see on average a reduction in presenteeism days by between 1 and 4 days, and that these days would see an increase in productivity from 30% to 100% in order to reflect full productivity. This is monetised assuming wage is equal to productivity.

In total, the combined benefits of prevented short term sickness absence days, GP savings and presenteeism days prevented are estimated to be between £500 million and £900 million.

Long-term absences

We use the number of fit note episodes with between 2 and 5+ repeat fit notes as a proxy for the number of employees with a long-term sickness absence. From the latest NHS fit note statistics this is around 2.2 million employees.

We are aware impacts of WHP may be different based on different health conditions faced by employees. A study of the effect of mental and physical health problems on sickness absence^{[footnote 18]} suggests that recovering from a mental health condition would reduce absence rates by almost double that of recovering from a physical condition. There is also some evidence that seeing an OH psychologist can reduce sickness absence due to mental disorders by 9 days per year^{[footnote 19]}.

Based on evidence from an Musculoskeletal Study of Work and Pain (SWAP) trial^{[footnote 20]}, our central assumption is a 16% reduction in sickness absence as a result of WHP intervention. We have also modelled a lower and higher estimate of 8% and 32% respectively.

This reduction in sickness absence is then applied to the average number of days taken per employee with a long-term sickness absence. Looking at the Employment of Disabled People 2024 publication^{[footnote 21]}, the median number of days taken for those with long- term sickness is around 43 days.

The benefits resulting from a reduction in sickness absences are modelled as:

• Employer profits (employees are more productive as fewer days off sick) – around £120 in employer profit per day
• Increased wages (employees receive more in work than off sick when eligible for SSP only^{[footnote 22]}) – around £15 per day

These benefits are monetised using estimates for full and part-time medium pay of employees combined with labour vs capital share of GVA. This captures both the output change and change in wage paid to an employee when an employee is in work and off sick. The monetary benefit to employees in receipt of SSP only is much greater when in work than off sick.

Additionally, we again assume that there will be some GP savings for this cohort as a result of routing Fit Notes through the WHP system and apply the same assumptions as set out above.

In total, the combined benefits of both prevented sickness absence days and GP savings are estimated to be between £1.1 billion and £4.4 billion. This is the greatest driver of overall benefits and is dependent on the assumption that sickness absences reduce by between 8% and 32%.

Stay in work/Unwell in Work

We use the number of employees with a disability and/or health condition as a proxy for the cohort of employees that would access the Stay in work element of the WHP. From the latest DWP Employment of Disabled People 2024 publication^{[footnote 23]} and the Labour market status of disabled people publication^{[footnote 24]} in combination this is estimated to be around 15 million employees. We then remove those who are absent from work (around 4.4 million as above) to avoid double counting and so the final cohort is assumed to be around 11 million employees.

We then apply a take-up assumption of 10% to this cohort as it is unlikely that all of this cohort will require a more intensive support package. The average Employee Assistance Programme usage is around 5%^{[footnote 25]}, we have estimated take-up will be higher in WHP due to ease of access and greater employer connectivity.

Based on evidence from an MSK SWAP trial and judgement based on intensity of support required in this cohort, our central assumption is an 8% reduction in sickness absence as a result of WHP intervention. We have also modelled a higher estimate of 16%.

This reduction in sickness absence is then applied to the average number of days taken per employee with a disability which is 12.6 days^{[footnote 26]}.

The benefits resulting from a reduction in sickness absences are modelled as:

• Employer profits (employees are more productive as fewer days off sick)
• Increased wages (employees receive more in work than off sick when eligible for SSP only)

These benefits are monetised using estimates for full and part-time medium pay of employees combined with labour vs capital share of GVA. This captures both the output change and change in wage paid to an employee when an employee is in work and off sick.

We also model a reduction in presenteeism to all employees within this cohort using the same methodology as set out for employees with short term absences. We apply a marginally higher assumption of between 4 and 9 days presenteeism to this group.

Overall benefits to this cohort in the form of prevented sickness absences and presenteeism days of between £800 million and £1.8 billion.

Healthy in work

We assume the cohort likely to access the Healthy in work element of WHP are the remaining UK employees after considering the above cohorts. This is estimated to be around 19 million employees.

Whilst this group are the cohort likely to require less intensive and direct support, we have assumed given a likely shift towards greater employer awareness and access of health and wellbeing support, there would be some reduction in presenteeism days for the remaining UK employees. We have used the above UNUM report as set out above and assumed based on the intensity of support available a 1-day reduction in presenteeism days for the Healthy in Work cohort. This suggests presenteeism benefits of around £300 million per year.

Wider benefits – HR/Conflict benefits

We make an additional assumption that overall the WHP would lead to a reduction in formal litigation that is in line with the proportion of employees prevented from falling out of the workplace (low estimate 3%, central estimate 5%, higher estimate 7%). Combining assumptions from government tribunal statistics and Acas, we have assumed that the overall employers’ costs associated with health-based conflict in the workplace is around £7 billion. Breaking this down, the total cost associated with health-based formal conflict is around £4.3 billion. Applying the above percentages, we have therefore assumed a monetised benefit of reducing formal conflict of between £100 million and £300 million per year.

Potential additional benefits from substantial system improvements

As adoption increases, and the WHP aims to drive further improvements in the employment landscape, there is the potential of possible further benefits in addition to the more direct impacts. We have also assumed some potential impacts of WHP in considering how the flow of people moving into long-term sickness absence may be reduced, whether any people currently inactive due to long term sickness could move into employment such as increasing the number of disabled people in employment, and whether overall preventative effects will be stronger to assume potential overall reductions in sickness absence days for all employees.

Stemming the flow into long-term sickness absence

We have assumed that a potential additional impact of WHP is that it would reduce the number of employees falling into long-term sickness absence. Whilst evidence is limited in this space across similar interventions, we have modelled the impact of both a 5% and 10% reduction in the size of this cohort. To model this, we have assumed a reduction in this cohort would be equivalent to employees’ absence length reducing from the median number of days of long-term sickness absence (43 days) to the average number of sick days for individuals with a long-term health condition (8 days). We then monetise these sickness absence days saved in the same methodology as set out above which estimates benefits between £550 million and £1.1 billion per year.

Individuals moving from inactivity into employment

Currently, 3.8% of people that are economically inactive move into employment over a 4-year period^{[footnote 27]}. Additionally, a DWP work aspirations study^{[footnote 28]} finds that 5% of health and disability customers felt they would be able to work if the right job or support was available. We have therefore assumed the marginal potential benefit in moving people from inactivity to employment is around 0.3%^{[footnote 29]}. We have applied this to the 2.8 million people currently inactive due to long-term sickness absence which estimates around 8,000 individuals. We have then monetised using the DWP Social Cost Benefit Analysis (SCBA)^{[footnote 30]} model as set out above which estimates benefits per year of around £1 billion.

We have also modelled what the benefits would be if we were to see a 0.5% to 1% increase in the number of disabled people in employment. This is estimated to be an increase of between 28,000 to 56,000 individuals per year. We then monetise this using the DWP SCBA model as set out above to estimate the benefit in returning an employee to work. This applies assumptions around the proportion of employees that return to work both full and part-time. This is estimated to provide benefits of between £3 billion and £6 billion per year.

Reducing overall sickness absence days for those healthy in work

We have made an assumption to understand what the additional benefits would be of reducing the total number of sickness absence days by 0.5 days for employees healthy in work. The benefits resulting from a reduction in sickness absences are modelled as set out above in the form of:

• Employer profits
• Increased wages

Overall, benefits are estimated to be around £1 billion per year.

Increasing the employment impact of WHP to an upper of 10%

Our central estimate is that WHP will be effective in preventing 5% of employees who leave work as a result of a long-term sickness absence. We have also modelled the benefits that 10% of employees are prevented from leaving work as a result of a long-term sickness absence. This is then modelled in the same way as set out above using the DWP SCBA model as set out above to estimate the benefit in retaining an employee in work.

Overall, potential additional benefits are expected to be in the range of £6 billion to £9 billion per year.

Potential additional benefits from substantially improving the system

	Lower Estimate	Central Estimate	Higher Estimate
Employers	£3.8 billion	£3.8 billion	£5.2 billion
Government	£0.8 billion	£0.8 billion	£1.5 billion
Employees	£1.5 billion	£1.5 billion	£2.6 billion
Total	£5.8 billion	£5.8 billion	£9.3 billion

Calculating the lifetime impacts of economic inactivity

It is impossible to precisely model the complexity and nuances of an individual’s life. We have therefore used a stylised persona to produce estimates of the lifetime impact of falling out of work to reach an indicative view. This includes the individual is:

• full-time employed until state pension age for their age bracket – i.e. they do not take periods of absence due to job loss, children, career breaks etc
• single, does not have children, and lives alone
• saves the default minimum for their pension when earning (4%, plus tax relief) – with no further savings or wealth included

We also do not account for any significant macroeconomic or social policy shifts in the future that may significantly alter the income profile in either scenario. These simplifying assumptions allow us to construct estimates of earning and benefit payments that are logically comparable.

When the individual is in work they are assumed to be:

• earning the median pay for their age. This profile grows until their 40s, when it reduces again (see ASHE 2024 earnings profiles). Real-term median pay for each age bracket grows at an average rate of <1% per annum, in line with average annual income growth over the last 20 years
• paying taxes for their earnings bracket, including both income tax and NIC contributions
• paying into their pension, and the employer contributes the minimum level (3%). The pension pot real annual rate of return is conservatively assumed to be 1%

When they are out of work, they:

• receive Personal Independence Payments (PIP) – as this is not a work-related benefit, it is conservative to only include this in the out-of-work scenario
• receive Universal Credit, plus additional support for healthcare payments and housing
• receive benefits that are adjusted in line with inflation, that is zero real-term growth over time

We have included more peripheral benefit payments, for example on prescription charges.

These assumptions produce the estimate that an individual that leaves work at the age of 22 loses over £1 million in earnings over their lifetime.

This is robust to sensitivity testing, such as higher potential benefit payments, council tax reductions, and student loan repayments for the income earner. If there is no real average annual growth in earnings, the lifetime loss is £0.7 to 0.8 million. Including more years out of work due to other reasons also closes the gap.

On the other hand, more optimistic views of income growth or pensions rate of return widen the gap. We have not included consideration of how the individual spends or saves their additional income that could also have a material impact on lifetime finances.

A similar exercise for the estimate of lifetime impact on an individual in their early 50s, based on average income, pension wealth and the high likelihood of home ownership status.

Together this analysis provides an indicative and stylised case study of lifetime earnings impacts.

Further Breakdown of benefits

1. Impact assessment: Improve access Statutory Sick Pay removing Lower Earnings Limit and waiting period ↩

2. The cost of working age ill-health and disability that prevents work – GOV.UK ↩

3. Health-Happiness-Productivity.pdf ↩

4. Tribunal Statistics Quarterly: April to June 2024 – GOV.UK ↩

5. Estimating the costs of workplace conflict – Acas ↩

6. These costs are not necessarily additional to what employers are already spending on workplace health provision, such as Occupational Health. ↩

7. The employment of disabled people 2024 – GOV.UK ↩

8. The work aspirations and support needs of health and disability customers: Interim findings – GOV.UK ↩

9. Pathways to Work impact evidence – Archived content ↩

10. The cost of working age ill-health and disability that prevents work – GOV.UK ↩

11. The DWP Social Cost-Benefit Analysis framework (WP86) – GOV.UK ↩

12. This assumption has potential to be a slight over-estimate as we know a small proportion of Fit Notes are used for benefit purposes. Data is not available as to the scale of this. ↩

13. Fit Notes Issued by GP Practices – NHS England Digital ↩

14. Evaluation of the 2022 Fit Note Reforms – GOV.UK ↩

15. Unit Costs of Health and Social Care 2024 Manual – Kent Academic Repository ↩

16. Effectiveness and costs of a vocational advice service to improve work outcomes in patients with musculoskeletal pain in primary care: a cluster randomised trial (SWAP trial ISRCTN 52269669) – PubMed ↩

17. Health-Happiness-Productivity.pdf ↩

18. The effect of mental and physical health problems on sickness absence – PMC ↩

19. Seeing an occupational health psychologist reduces sickness absence due to mental disorders: A quasi-experimental study – ScienceDirect ↩

20. Effectiveness and costs of a vocational advice service to improve work outcomes in patients with musculoskeletal pain in primary care: a cluster randomised trial (SWAP trial ISRCTN 52269669) – PubMed ↩

21. The employment of disabled people 2024 – GOV.UK ↩

22. We assume 25% of employees receive SSP only and 75% of employees receive Occupational Sick Pay and therefore we assume receive the same pay when in work and off sick. ↩

23. The employment of disabled people 2024 – GOV.UK ↩

24. A08: Labour market status of disabled people – Office for National Statistics ↩

25. 53 employee assistance programme (EAP) statistics for 2024 ↩

26. Sickness absence in the UK labour market – Office for National Statistics ↩

27. tables-keep-britain-working-2015-2024.ods – Table KBW009 ↩

28. The work aspirations and support needs of health and disability customers: Interim findings – GOV.UK ↩

29. 5% to 3.8% over a 4 year period is around 0.3% per year. ↩

30. The DWP Social Cost-Benefit Analysis framework (WP86) – GOV.UK ↩