Consultation outcome

Impact assessment (accessible)

Updated 12 February 2024

This was published under the 2022 to 2024 Sunak Conservative government

Title: Minimum service levels (MSL) consultation

IA number: HO0429

RPC reference number: N/A

Other departments or agencies: Department for Business and Trade, Department for Health and Social Care, Department for Transport

Date: 21 February 2023

Stage: Consultation

Intervention: Domestic

Measure: Secondary legislation

Enquiries: [email protected]

RPC opinion: Not applicable

Business impact target: Non-qualifying regulatory provision

Cost of preferred (or more likely) option (in 2022/23 prices)

Net present social value (NPSV) (£m): N/A

Business net present value (BNPV) (£m): N/A

Net cost to business per year (EANDCB) (£m): N/A

What is the problem under consideration? Why is government intervention necessary?

Currently, fire and rescue services (FRS) during strike action rely on cover from individuals not undertaking strike action or external parties (military personnel or contracted support). FRS should have business continuity plans in place, and often return to work agreements, however these do not provide sufficient assurances to fully mitigate the risk posed by strike action to the lives and welfare of the public. Minimum service levels (MSL) need to be able to mitigate as much risk to life and limb as possible and ensure an appropriate level of staffing is upheld during strike action. Currently, there is no legislation that introduces statutory MSL during periods of strike action.

What is the strategic objective? What are the main policy objectives and intended effects?

Strategic objective

Improve public safety by limiting the impacts of firefighter strike action whilst balancing this with the ability for FRS employees to strike.

Policy objective

The legislation aims to ensure a minimum service is provided by FRS to the public during strike action while maintaining the ability for strike action to take place. It also aims to mitigate the risk that FRS could be overwhelmed by demand during strike periods, reduce the potential impact of major incidents during strike action, improve public safety and the safety of firefighters, and reduce the costs associated with developing business continuity plans

What policy options have been considered, including any alternatives to regulation? Please justify preferred option (further details in evidence base)

  • Option 0: Take no action and make no legislative changes (do-nothing).

  • Option 1: Set an MSL such that staffing numbers must never go below a certain level of attendance.

  • Option 2: Set an MSL so staffing levels are geared to respond to specific risks, and major incidents.

  • Option 3: Set individual MSL following local leader/organisational input in collaboration with Home Office

  • Option 4: Set a national MSL, but allow local leader and their organisation to decide specifics

  • Option 5: Set an MSL so cover is maintained on high-risk days or at high-risk times.

Main assumptions/sensitivities and economic/analytical risks

Discount rate (%): 3.5%

The best available data is used in the analysis, informed by experience and expertise. Several high impact assumptions are made which are significant determinants of the NPSV of these policies. These include the amount of strike action that this legislation would prevent, actual and potential strike turnout, and the monetised value of a firefighter’s work. However, relevant analysis and data is put forward along with sensitivity testing to help understand the uncertainties involved. The overall net social present value could be higher or lower depending on the final policy design.

Will the policy be reviewed?

It will be reviewed. If applicable, set review date: February 2028.

I have read the impact assessment and I am satisfied that, given the available evidence, it represents a reasonable view of the likely costs, benefits, and impact of the leading options.

Signed by the responsible minister:

Date: 20/02/2023

Policy option 1

Summary: analysis and evidence

Description: Set an MSL such that staffing numbers must never go below a certain level of attendance.

Full economic assessment

Year(s): Price base 2022 to 2023 PV base 2023 to 2024 Appraisal 10 Transition 1
Estimate of NPSV (£m)           Estimate of BNPV (£m)    
Low: -2.2 High: 182.0 Best: 29.3 Best BNPV -0.0  
Costs (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 0.5 0.0 0.5 0.1 0.0
High 2.2 0.0 2.2 0.2 0.0
Best estimate 1.1 0.0 1.1 0.1 0.0

Description and scale of key monetised costs by ‘main affected groups’

The only monetised costs of this option are from familiarisation to trade unions (£3,000 to £12,000, central estimate £6,000), FRS staff (£0.5m to £1.9m, central estimate £0.9m) and senior FRS management (£0.1 to £0.3m, central estimate £0.1m).

Other key non-monetised costs by ‘main affected groups’

A number of potential costs have not been monetised in this IA, including potential enforcement costs, costs from the potential changing nature of strike action, and any operational costs to employers and unions. Some consideration is also made of the impact of MSL on FRS staff.

Benefits (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 0.0 0.0 0.0 0.0 0.0
High 0.0 182.5 182.5 3.5 0.0
Best estimate 0.0 30.4 30.4 21.2 0.0

Description and scale of key monetised benefits by ‘main affected groups’

The monetised benefits of this policy arise from the increase in hours worked by firefighters on strike days and from cost savings to FRS and central government from not having to provide additional resilience cover on these days. This totals £0 to £182.5m, best estimate £30.4m.

Other key non-monetised benefits by ‘main affected groups’

There are also considerable public safety benefits from an improvement in fire response on strike days, and from an increase in other activity undertaken on strike days (for example, fire safety activity and non-fire response). This could have wider economic output benefits too from reduced fire severity and a reduction in the fear of fire individuals may have during strike days.

Business assessment (option 1)

Direct impact on business (equivalent annual) £m

Cost (£m): 0.0

Benefit (£m): 0.0

Net (£m): 0.0

Score for business impact target (qualifying provisions only) £m

N/A

Is this measure likely to impact on trade and investment?

No

Are any of these organisations in scope?

Micro: Yes

Small: Yes

Medium: Yes

Large: Yes

What is the CO2 equivalent change in greenhouse gas emissions? (Million tonnes CO2 equivalent)

Traded: N/A

Non-traded: N/A

People and specific impacts assessment (option 1)

Are all relevant specific impacts included?

Yes

Are there any impacts on particular groups?

Yes

Policy option 2

Summary: analysis and evidence

Description: Set an MSL such that staffing levels must be geared to respond to specific risks, including a minimum standard to respond to a major incident.

Full economic assessment

Year(s): Price base 2022 to 2023 PV base 2023 to 2024 Appraisal 10 Transition 1
Estimate of NPSV (£m)           Estimate of BNPV (£m)    
Low: 3.8 High: 375.6 Best: 117.5 Best BNPV -0.0  
Costs (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 0.5 0.0 0.5 0.1 0.0
High 2.2 0.0 2.2 0.2 0.0
Best estimate 1.1 0.0 1.1 0.1 0.0

Description and scale of key monetised costs by ‘main affected groups’

The only monetised costs of this option are from familiarisation to trade unions (£3,000 to £12,000, central estimate £6,000), FRS staff (£0.5m to £1.9m, central estimate £0.9m) and senior FRS management (£0.1 to £0.3m, central estimate £0.1m).

Other key non-monetised costs by ‘main affected groups’

A number of potential costs have not been monetised in this IA, including potential enforcement costs, costs from the potential changing nature of strike action, and any operational costs to employers and unions. Some consideration is also made of the impact of MSL on FRS staff.

Benefits (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 0.0 6.0 6.0 0.7 0.0
High 0.0 376.1 376.1 43.7 0.0
Best estimate 0.0 118.6 118.6 13.8 0.0

Description and scale of key monetised benefits by ‘main affected groups’

The monetised benefits of this policy arise from the increase in hours worked by firefighters on strike days and from cost savings to FRS and central government from not having to provide additional resilience cover on these days. This totals £6.0m to £376.1m, best estimate £118.6m.

Other key non-monetised benefits by ‘main affected groups’

There are also considerable public safety benefits from an improvement in fire response on strike days, and from an increase in other activity undertaken on strike days (for example fire safety activity and non-fire response). This could have wider economic output benefits too from reduced fire severity and a reduction in the fear of fire individuals may have during strike days.

Business assessment (option 2)

Direct impact on business (equivalent annual) £m

Cost (£m): 0.0

Benefit (£m): 0.0

Net (£m): 0.0

Score for business impact target (qualifying provisions only) £m

N/A

Is this measure likely to impact on trade and investment?

No

Are any of these organisations in scope?

Micro: Yes

Small: Yes

Medium: Yes

Large: Yes

What is the CO2 equivalent change in greenhouse gas emissions? (Million tonnes CO2 equivalent)

Traded: N/A

Non-traded: N/A

People and specific impacts assessment (option 2)

Are all relevant specific impacts included?

Yes

Are there any impacts on particular groups?

Yes

Policy option 3

Summary: analysis and evidence

Description: Set individual MSL following local leader/organisational input in collaboration with Home Office/Secretary of State (for example, not a national level, but based on local priorities and pressures).

Full economic assessment

Year(s): Price base 2022 to 2023 PV base 2023 to 2024 Appraisal 10 Transition 1
Estimate of NPSV (£m)           Estimate of BNPV (£m)    
Low: -5.2 High: -1.4 Best: -2.8 Best BNPV -0.0  
Costs (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 1.4 0.0 1.4 0.1 0.0
High 5.2 0.0 5.2 0.5 0.1
Best estimate 2.8 0.0 2.8 0.3 0.0

Description and scale of key monetised costs by ‘main affected groups’

The monetised costs of this option are the same familiarisation costs as in options 1 to 3. There are also additional set-up costs for trade unions (£15,000 to £46,000, central estimate £31,000), FRS staff (£0.5m to £1.9m, central estimate £0.9m) and senior FRS management (£0.4m to £1.2m, central estimate £0.8m).

Other key non-monetised costs by ‘main affected groups’

A number of potential costs have not been monetised in this IA, including potential enforcement costs, costs from the potential changing nature of strike action, and any operational costs to employers and unions. Some consideration is also made of the impact of MSL on FRS staff.

Benefits (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 0.0 0.0 0.0 0.0 0.0
High 0.0 0.0 0.0 0.0 0.0
Best estimate 0.0 0.0 0.0 0.0 0.0

Description and scale of key monetised benefits by ‘main affected groups’

There are no monetised benefits of this option at this stage due to uncertainties on the MSL that would be set in FRS.

Other key non-monetised benefits by ‘main affected groups’

There are also considerable public safety benefits from an improvement in fire response on strike days, and from an increase in other activity undertaken on strike days (for example fire safety activity and non-fire response). This could have wider economic output benefits too from reduced fire severity and a reduction in the fear of fire individuals may have during strike days.

Business assessment (option 3)

Direct impact on business (equivalent annual) £m

Cost (£m): 0.0

Benefit (£m): 0.0

Net (£m): 0.0

Score for business impact target (qualifying provisions only) £m

N/A

Is this measure likely to impact on trade and investment?

No

Are any of these organisations in scope?

Micro: Yes

Small: Yes

Medium: Yes

Large: Yes

What is the CO2 equivalent change in greenhouse gas emissions? (Million tonnes CO2 equivalent)

Traded: N/A

Non-traded: N/A

People and specific impacts assessment (option 3)

Are all relevant specific impacts included?

Yes

Are there any impacts on particular groups?

Yes

Policy option 4

Summary: analysis and evidence Policy Option 4

Description: Secretary of State/Home Office set a national MSL, and Chief Fire Officers and their organisation decide specifics for local area.

Full economic assessment

Year(s): Price base 2022 to 2023 PV base 2023 to 2024 Appraisal 10 Transition 1
Estimate of NPSV (£m)           Estimate of BNPV (£m)    
Low: -5.2 High: 205.1 Best: 38.8 Best BNPV -0.0  
Costs (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 1.4 0.0 1.4 0.1 0.0
High 5.2 0.0 5.2 0.5 0.1
Best estimate 2.8 0.0 2.8 0.3 0.0

Description and scale of key monetised costs by ‘main affected groups’

The monetised costs of this option are the same familiarisation costs as in options 1 to 3. There are also additional set-up costs for trade unions (£15,000 to £46,000, central estimate £31,000), FRS staff (£0.5m to £1.9m, central estimate £0.9m) and senior FRS management (£0.4m to £1.2m, central estimate £0.8m).

Other key non-monetised costs by ‘main affected groups’

A number of potential costs have not been monetised in this IA, including potential enforcement costs, costs from the potential changing nature of strike action, and any operational costs to employers and unions. Some consideration is also made of the impact of MSL on FRS staff.

Benefits (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 0.0 0.0 0.0 0.0 0.0
High 0.0 206.5 206.5 24.0 0.0
Best estimate 0.0 41.6 41.6 4.8 0.0

Description and scale of key monetised benefits by ‘main affected groups’

The monetised benefits of this policy arise from the increase in hours worked by firefighters on strike days and from cost savings to FRS and central government from not having to provide additional resilience cover on these days. This totals £0 to £206.5m, best estimate £41.6m.

Other key non-monetised benefits by ‘main affected groups’

There are also considerable public safety benefits from an improvement in fire response on strike days, and from an increase in other activity undertaken on strike days (for example fire safety activity and non-fire response). This could have wider economic output benefits too from reduced fire severity and a reduction in the fear of fire individuals may have during strike days.

Business assessment (option 4)

Direct impact on business (equivalent annual) £m

Cost (£m): 0.0

Benefit (£m): 0.0

Net (£m): 0.0

Score for business impact target (qualifying provisions only) £m

N/A

Is this measure likely to impact on trade and investment?

No

Are any of these organisations in scope?

Micro: Yes

Small: Yes

Medium: Yes

Large: Yes

What is the CO2 equivalent change in greenhouse gas emissions? (Million tonnes CO2 equivalent)

Traded: N/A

Non-traded: N/A

People and specific impacts assessment (option 4)

Are all relevant specific impacts included?

Yes

Are there any impacts on particular groups?

Yes

Policy option 5

Summary: analysis and evidence

Description: Set an MSL so that staffing levels ensure that cover is maintained on high-risk days or at high-risk times.

Full economic assessment

Year(s): Price base 2022 to 2023 PV base 2023 to 2024 Appraisal 10 Transition 1
Estimate of NPSV (£m)           Estimate of BNPV (£m)    
Low: -2.2 High: 242.0 Best: 37.0 Best BNPV -0.0  
Costs (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 0.5 0.0 0.5 0.1 0.0
High 2.2 0.0 2.2 0.2 0.0
Best estimate 1.1 0.0 1.1 0.1 0.0

Description and scale of key monetised costs by ‘main affected groups’

The only monetised costs of this option are from familiarisation to trade unions (£3,000 to £12,000, central estimate £6,000), FRS staff (£0.5m to £1.9m, central estimate £0.9m) and senior FRS management (£0.1 to £0.3m, central estimate £0.1m).

Other key non-monetised costs by ‘main affected groups’

A number of potential costs have not been monetised in this IA, including potential enforcement costs, costs from the potential changing nature of strike action, and any operational costs to employers and unions. Some consideration is also made of the impact of MSL on FRS staff.

Benefits (£m) Transition
Constant price
Ongoing
Present value
Total
Present value
Average/year
Constant price
To business
Present value
Low 0.0 0.0 0.0 0.0 0.0
High 0.0 242.5 242.5 28.2 0.0
Best estimate 0.0 38.1 38.1 4.4 0.0

Description and scale of key monetised benefits by ‘main affected groups’

The monetised benefits of this policy arise from the increase in hours worked by firefighters on strike days and from cost savings to FRS and central government from not having to provide additional resilience cover on these days. This totals £0 to £242.5m, best estimate £38.1m.

Other key non-monetised benefits by ‘main affected groups’

There are also considerable public safety benefits from an improvement in fire response on strike days, and from an increase in other activity undertaken on strike days (for example fire safety activity and non-fire response). This could have wider economic output benefits too from reduced fire severity and a reduction in the fear of fire individuals may have during strike days.

Business assessment (option 5)

Direct impact on business (equivalent annual) £m

Cost (£m): 0.0

Benefit (£m): 0.0

Net (£m): 0.0

Score for business impact target (qualifying provisions only) £m

N/A

Is this measure likely to impact on trade and investment?

No

Are any of these organisations in scope?

Micro: Yes

Small: Yes

Medium: Yes

Large: Yes

What is the CO2 equivalent change in greenhouse gas emissions? (Million tonnes CO2 equivalent)

Traded: N/A

Non-traded: N/A

People and specific impacts assessment (option 5)

Are all relevant specific impacts included?

Yes

Are there any impacts on particular groups?

Yes

Evidence base

A. Strategic objective and overview

A.1 Strategic objective

1. Problem under consideration: Currently, fire and rescue services (FRS) activity during strike action relies on cover from individuals not undertaking strike action or from external parties (military personnel or contracted support). All FRS should have business continuity plans (BCP) in place, and sometimes return to work agreements. However, these do not provide sufficient assurances to fully mitigate against the risk posed by strike action. Minimum service levels (MSL) need to be able to mitigate as much risk to public safety as possible and ensure an appropriate level of staffing during strike action. Currently, there is no legislation that introduces statutory MSL during periods of strike action, so government intervention is necessary.

2. Strategic objective: Improve public safety by limiting the impact of firefighter strike action whilst balancing this with the ability for FRS employees to strike.

A.2 Background

Policy background

3. The government intends to legislate to formalise a MSL provided by FRS to the public during strike action whilst maintaining the ability for strike action to take place. The Strikes (Minimum Service Levels) Bill that was introduced to Parliament on 10 January will, if passed by Parliament, amend the Trade Union and Labour Relations (Consolidation) Act 1992 Act to:

  • restrict the protection that is provided to trade unions and employees in respect of strikes where provision has been made in regulations for minimum levels of service

  • enable employers to issue work notices to unions and those persons required to maintain those minimum service levels.

4. The Bill provides that work notices are the mechanism that puts MSL into practice for particular strikes in relevant services. Under the Bill, work notices may be given by the employer to trade unions and must identify the persons required to work during the strike in order to maintain MSL and specify the work to be carried out by them during the strike in order to achieve MSL. A work notice must not identify more persons than are reasonably necessary for the purpose of providing the levels of service under the minimum service regulations nor can the employer have regard to whether the person is or is not a member of a trade union (or a particular trade union). There is a requirement for the employer to consult with the union over the number of persons to be identified and the work to be specified in the work notice. The employer must have regard to any views of the union in response to this consultation. Work notices must be issued at least 7 days in advance on the strike date, or later if agreed by the union. Work notices can also be varied up until the fourth day prior to the strike date, or later if agreed by the union.

5. The Bill would, if passed:

  • provide a power for the Secretary of State (for FRS, it would be the Home Secretary) to make regulations providing for minimum levels of service where there is strike action in relevant services – referred to as minimum service regulations

  • provide for a further power for the Secretary of State (for FRS, it would be the Home Secretary) to specify in regulations the relevant services for which minimum service level regulations may be made. This power to make regulations specifying relevant services is limited to the following categories: health services; fire and rescue services; education services; transport services; decommissioning of nuclear installations and management of radioactive waste and spent fuel; and border security

6. The Secretary of State (for FRS, it would be the Home Secretary) is required to consult before making regulations setting MSL and specifying relevant services to which they are to be applied to. The Bill provides that the consultation process may be carried out before (and after) the Bill receives Royal Assent. The accompanying consultation document starts the process of consultation on both the ‘essential services’, within FRS services, to which MSL should apply, the scope of staff who should be included and possible MSL options. The aim of the consultation is to better understand the framework in which this policy is being developed, compare various models, and draw out advantages and disadvantages of each option.

Fire and rescue background

7. The role of emergency services is to keep the public safe. While there remains importance in maintaining the ability for FRS employees to strike, this should be balanced against the need to protect the public from disproportionate negative impacts.

8. Fires poses a significant public safety risk, and the destruction they can cause has disproportionate effects on the most vulnerable in society. There are still significant concerns around a number of multi-occupied residential properties with dangerous cladding. The government is taking action to make these buildings safer, as highlighted by its commitment to provide funding for remediation of cladding.[footnote 1] In addition, the government has committed to reforming the FRS following recent challenges. The proposed reforms were set out in the Fire Reform White Paper in May 2022 and the government response to that consultation is set to be published in due course.

9. FRS take on other essential services beyond that of firefighting. The proposed MSL would incorporate this activity to ensure suitable cover is also available for, but not limited to, rescues (including but not limited to, those on the road network, water rescues or rescues at height). This includes actions to avoid further harm such as rectifying potentially hazardous situations to avoid future risk of fire and rescue, for example clearance of debris on motorways and major roads, dangerous substance clean-up, ability to maintain crewing of national resilience assets, services necessary to carry out the above, including for example control room activities.

10. Against this background and given the risk that fires can spread rapidly, it is vital that FRS maintain a minimum level of emergency response and service during periods of strike action, protecting the public, and the places they live and work.

11. The consultation[footnote 2] considers the balance between public protection and workers’ rights. The ability for FRS employees to strike is an important part of industrial relations in the UK, protected by law. Any strike action will inevitably cause some form of disruption. Therefore, it is important to strike a reasonable balance between public protection and workers’ rights. To support all life-saving activity conducted by FRS, the legislation will not be limited to firefighting. It will include all reasonable rescue activity covered by FRS in a proportionate way with the majority of focus on first responders.

12. Last month (January 2023) the Fire Brigades Union (FBU) received the outcome of a ballot, with members voting in favour of strike action on pay. The outcome of the ballot was followed by a revised pay offer from employers, which is now out for a consultative ballot. The Executive Council of the FBU have unanimously agreed to recommend that members vote to accept the revised pay offer, and have postponed the announcement of strike dates pending the outcome of the ballot - results are due on 6 March. Not all staff within FRS will be members of a union, however, the FBU does have considerably higher membership rates compared to other FRS unions, as well as other unions across public sector.

13. The International Labour Organisation, which is an agency of the United Nations, has stated that minimum service levels are justifiable for the following services:

  • services the interruption of which would endanger the life, personal safety or health of the whole or part of the population (essential services in the strict sense of the term)

  • services which are not essential in the strict sense of the term but where the extent and duration of a strike might be such as to result in an acute national crisis endangering the normal living conditions of the population, and in public services of fundamental importance

14. FRS would likely be in scope of these definitions as strikes within these services would cause interruption which would endanger public safety, and FRS are of fundamental importance.

15. The consultation seeks to understand, through responses, the implications of setting MSL in FRS in England, Scotland, and Wales. Therefore, the scope of this impact assessment (IA) is Great Britain.

A.3 Groups affected

16. The proposed legislation would affect the following groups:

  • Fire and rescue authorities (FRA): a fire authority or fire and rescue authority is a statutory body made up of a committee of local councillors, which oversees the policy and service delivery of a fire and rescue service. They will have responsibility for issuing work notices under MSL proposals.

  • Fire and rescue services (FRS): responsible for delivery of fire and rescue services, reporting to fire and rescue authorities, and will have a role in drafting work notices.

  • FRS employees: the main focus of the MSL proposals will be on firefighters and control room staff.

  • General public: they will be impacted in terms of changes to public safety and FRS response during strike action.

  • Unions and other membership organisations: unions and similar organisations will have varying negotiation and consultation duties.

A.4 Consultation

Within UK government

  • Department for Business and Trade (DBT)

  • Department for Transport (DfT)

  • Department of Health and Social Care (DHSC)

  • Department for Levelling Up, Housing and Communities (DLUHC)

  • Cabinet Office (CO)

  • Border Force leads within Home Office (HO)

  • HM Treasury

  • Department for Education

  • Department for Environment, Food & Rural Affairs

  • Wales Office

  • Scotland Office

18. The Home Office are also consulting with the Welsh and Scottish governments.

19. Trade union legislation is applicable to Great Britain only. However, the Home Office will maintain an open dialogue with the Northern Ireland Office and the Northern Ireland Executive.

Public consultation

20. This IA forms part of the public consultation. The consultation will be a central part of the considerations on how to take forward regulations that will set out the detail of MSL to be applied to FRS.

21. The government is keen to seek the views of those affected by the proposals contained in this IA to better understand the impact they could have, and what the most appropriate next steps are. The consultation is open to the public over a 12-week period.

B. Rationale for intervention

22. Strike action in public services such as ambulance and fire can put lives and welfare at risk. More generally, it can lead to adverse personal impacts for users of these services, as well as generating wider social, economic, and environmental impacts on the UK and its economy. Whilst a substantial number of users and economic agents bear the impact of strike action, they are neither party to any dispute nor have any avenue to have their interests formally represented. The impact of strike action on these parties represents a negative externality which is not reflected in the interests of employers and trade unions.

23. Currently, all FRS should have BCP in place, and sometimes return to work agreements, however, these do not provide sufficient assurances to fully mitigate against the risk posed by strike action. MSL need to, in the first instance, be able to mitigate against risk to life and limb by formalising an appropriate level of staffing during strike action to respond to these fire incidents. MSL could also be set to further improve public safety by including all reasonable rescue activity covered by FRS in a proportionate way with the majority of focus on first responders. This would have benefits on the public, whether that be through increased public safety, reduced public costs, or reduced economic costs from fire (for example from property damage or environmental damage).

24. Currently, there is no legislation that introduces statutory MSL during periods of strike action, so government intervention is necessary.

C. Policy objective

5. There are a number of policy objectives:

  • Limit the impacts of strike action on the lives and livelihoods of the public and to strike a balance between the ability of unions and their members to strike with the need for the wider public to be able to access FRS during strikes.

  • Mitigate the risk that fire and rescue services could be overwhelmed by demand during strike periods. Business continuity plans prepared by FRS are rigorously assessed and stress tested before any period of strike action, but there remains a risk that these will not be sufficient and not as effective as using professional firefighters.

  • Reduce the potential impact of major incidents during strike action. Unions and FRA may choose to negotiate a return-to-work protocol for major incidents, but this is on a strike by strike basis and is not a formalised or statutory level of service.

  • Ensure public safety: MSL aim to improve public safety during strike action. Even relatively small fires can potentially spread rapidly and quickly develop into significant incidents, so secured MSL which ensure faster response times and improved weight of response (number of appliances) could improve public safety.

  • Reduce the public cost of FRA arranging contracts with private contractors in order to bolster their BCP. To ensure appropriate resources are in place, external support may be required to support BCP which can come at a significant cost. MSL provide the ability to help reduce this cost.

  • Remove the need for military assistance during strike action. Ahead of strike action, contingency plans are developed in England in case FRS are overwhelmed on strike days. These plans can include the option of drawing on military aid to the civil authorities. However, this is not a long-term or sustainable solution as military support does not provide the same depth of expertise and experience as professional civilian firefighters. Drawing on military aid to the civil authorities also incurs costs as determined by the Ministry of Defence and takes military personnel away from their own duties.

  • Provide certainty and allow local leaders to plan for periods of strike action by giving a clear indication of how many staff they will have available on strike days.

  • Ensure responsible crewing levels to maintain firefighter safety during strike action.

D. Options considered and implementation

26. A non-regulatory approach to encouraging MSL to be put in place by FRS and unions without legislation would not meet the government’s objectives, so has not been considered.

27. As set out in the consultation document, there are two levels to the options being proposed. The first is who would set the MSL, and how they would set it; and the second level is what level the MSL would be set at.

Option 0: Take no action and make no legislative changes (do-nothing)

28. Under Option 0 there would be no legislative changes to formalise MSL and current local arrangements during strike action, including business continuity plans and potential contractor and armed forces support would remain.

Option 1: Set MSL such that staffing numbers must never go below a certain level of attendance

29. Under Option 1, an MSL would be set based on a certain number of appliances (fire engines) and/or staff being available. This could be based on holiday staffing levels (akin to lower levels expected during the summer holidays), bank holiday staffing levels, a percentage based on how many people are required to operate lifesaving initiatives; or an appliance availability based MSL – that is a set number of fire appliances required, which would then require a minimum level of firefighters to crew those.

Option 2: Set MSL such that staffing levels must be geared to respond to specific risks, including a minimum standard to respond to major incidents

30. Under Option 2, MSL would be set so that each FRS must be able to respond to a major incident if one were to occur. The exact definition of this major incident is being consulted on and will require further work. For the purposes of this analysis, major incidents across three incident types are considered, to understand what an equivalent MSL could be so that FRS could staff an adequate response to such an incident. For each of these incident types, the maximum number of distinct pumping appliances mobilised to these incidents[footnote 3] are taken for each FRS, as per incident recording system (IRS) data. These are then compared to the number of pumping appliances FRS have available[footnote 4], to work out the proportion of their pumping appliances (a proxy for staff) would be required to respond to the most major incident in terms of mobilisation the FRS has faced. These percentages are calculated for each FRS, and then averaged, to work out the average ‘maximum’ number of pumping appliances mobilised to each major incident type. The incident types, with their estimated percentage levels, are below:

i) dwelling fires in purpose-built flats (10+ storeys): 22%

ii) a major wildfire (proxied as fires in grassland, woodlands, and crops): 52%

iii) a major ‘other building’ fire: 65%

31. FRS also have other responsibilities, beyond firefighting, and attend different types of non-fire incidents such as rescues. It is possible that some of this activity would be incorporated into the MSL for option 2, and so for completeness, the same analysis was completed for non-fire incidents attended. This showed that, on average, the maximum number of distinct pumping appliances mobilised to non-fire incidents was approximately 27% across all FRS. This falls in the range proposed above, and is close to the dwelling fire scenario (i) presented.

32. It is expected, that because of the nature of ‘major incidents’ that setting an MSL using this process would lead to a significantly higher MSL than option 1, especially because many FRS are already staffed for this capability in ‘normal’ times.

33. Using the examples above under paragraph 30 i)-iii) and adding the appliance availability requirements from option 1 of 25% to deal so that staffing numbers must never go below a certain level of attendance, it is assumed that under this option the MSL could be set at approximately 47% to 90% for each FRS, with a central estimate of 77%.[footnote 5] This would lead to increased public safety benefits as, unlike option 1, it would further enable an effective FRS response to be mounted during any period of potential strike action, whereas option 1 would likely still require effective return to work agreements in case of a major incident.

Option 3: Set individual MSL following local leader/organisational input in collaboration with Home Office/Secretary of State (for example, not a national level, but based on local priorities and pressures)

34. Under option 3 there would not be a national MSL set, it would instead be based on local priorities and pressures. This would mean that chief fire officers (CFO), local leaders, and their organisations would be asked to input and provide evidence to determine what a local level MSL should look like, and there would be a local process with employees and trade union officials to determine this level. This would incur additional costs than options 1 and 2.

35. This would allow an MSL to be set based on local risks, however, as the power to set MSL will rest with the Secretary of State of the relevant department, further consideration needs to be given on how this would work in practice. There is a risk that this option could lead to significant variations between areas (potentially creating perverse FRS staff incentives and divergent levels of public safety across the country) and would be much more resource intensive to administer compared to other approaches. It is likely that for this to be viable the government would set some parameters (the MSL would be bounded between certain percentage levels) with the final approval having to be made by the Secretary of State of the Home department through regulations.

36. The variable mechanisms for how these MSL would be set, and at what level they would be set, means that there is significant uncertainty associated for them. At this stage, it is not possible to estimate what percentage level these MSL would likely be set at, as depending on local input, they could be higher or lower than MSL proposed in other options. Therefore, no estimation for the percentage level of the MSL in this option is made in this IA, and no benefits are monetised for this option at this stage. The estimated level, and impact of this option, will be further refined following the consultation using any relevant evidence gathered.

Option 4: Secretary of State/Home Office set a national MSL, and chief fire officers and their organisation decide specifics for local area

37. Under option 4, there would be a national MSL in place but with flexibility for individual areas to adapt to their needs. This adaptability would be decided by CFO and their organisation. For example, a Metropolitan FRS may be required to be equipped to fight high-rise fires and a rural FRS may be required to be staffed to respond fully to flood risk. The incidents in-scope would be decided locally.

38. It is likely that this option would be combined potentially with elements from other options, such as a minimum percentage staffing level. For the purposes of this IA, it is modelled in the same way as option 1 for total costs, but with a 5 percentage point higher MSL level (to account for specific risks). The range is 30 to 55% with a central estimate of 42.5%.

Option 5: Set an MSL so that staffing levels ensure that cover is maintained on high-risk days or at high-risk times

39. Under option 5, MSL would have the additional requirement that FRS would need to maintain and provide normal business as usual arrangements at high-risk times or dates. This would essentially prevent FRS undertaking strike action at certain times. This would have public safety benefits, as it would ensure cover was adequate when FRS response is more likely to be required, and so could control against inadequate cover being provided by FRS. This means that it would likely give greater public safety benefits than option 1 (depending on if strike action was just displaced), however it would not provide as much public safety benefit as option 2, because it is not possible to predict when every major incident will occur.

40. Option 5 could be added onto any other option as an additional requirement, and so could be modelled alongside any option 1 to 4. For the purposes of this IA, it is assumed that option 5 is modelled as an addition to option 1, with 3 different scenarios presented for the MSL set. These make up the low, central, and high-cost estimate.

41. Low scenario - Strikes cannot take place on certain days such as Bonfire Night and surrounding days when incidents are high, and other public holidays such as Christmas Day or New Years Eve. This is modelled by assuming a 25% MSL (option 1 low scenario) normally, and then accounting for 7 out of 365 potential strike days (2%) having a 100% MSL instead of a 25% one. For the purposes of monetising impact in this IA, this is equated to an annual overall MSL level of approximately 26%.[footnote 6] This assumes that strike action is equally likely to happen on any day of the year, which may not be the case, and has been completed for simplicity in this IA so that benefits can be calculated and monetised. With this assumption, there are additional hours saved over option 1. It is important to note that this may not have any hours saved benefits over option 1 if strike action is just displaced. It may have some public safety benefits though, as strike action would happen on less risky days.

42. Central scenario - Strikes must be retracted in times of unprecedented weather, that is extreme flooding or wildfires. This is modelled by assuming a 37.5% MSL (option 1 central scenario) normally, and then accounting for 20 out of 365 potential strike days (a broad proxy for number of days with adverse weather per year, 5%) being deemed as unprecedented weather days and so strike action not being possible. For the purposes of monetising impact in this IA, this is equated to an overall annual MSL level of approximately 41%.[footnote 7] This assumes that strike action is equally likely to happen on any day of the year, which may not be the case, and has been completed for simplicity in this IA so that benefits can be calculated and monetised. With this assumption, there are additional hours saved over option 1 because during these times there would essentially be a 100% MSL applied, and no strike action could happen. This would have public safety benefits over current return to work arrangements, as these strike prevention measures would happen before strike action could happen, important to note that this may not have any hour saved benefits over option 1 if strike action is just displaced. It may however have public safety benefits because these are times when FRS are more likely to respond to incidents.

43. High scenario - Strikes cannot take place during peak hours (assumed to be 4pm to 10pm – 25% of the day).[footnote 8] This is modelled by assuming a 50% MSL (option 1 high scenario) normally, but also accounting for there being 25% of the year where strikes are not possible. For the purposes of monetising impact in this IA, this is equated to an overall MSL of approximately 63%.[footnote 9] With this assumption, there are additional hours saved over option 2 because 25% of the time that strike action would occur in, is met with a 100% MSL. As with the low and centrals scenario, it is important to note that this may not have any hour saved benefits over option 1 if strike action is just displaced. It would likely though have some public safety benefits because this is the time of day when the most fire incidents occur.

Preferred option and implementation date

44. This is a consultation IA and there is no preferred option at this stage. The purpose of this IA is to show the potential impacts of the policy, and the purpose of the accompanying consultation is to seek views from those likely to be affected by these proposals. These views will be used to further the government’s understanding and inform future policy considerations. The consultation is seeking meaningful engagement and views on what is the most effective option is to implement MSL.

45. Subject to the outcome of the consultation, it is expected that the measures set out in this IA will require secondary legislation. The consultation closes in May 2023, and any secondary legislation will follow this (accompanied with a final stage IA), subject to the parliamentary timetable.

E. Appraisal

46. The following sections present analysis of the costs and benefits of the options in the consultation compared to the ‘do-nothing’ option.

General assumptions and data

47. The best available data has been used for this IA. Costings for the appraisal section are based on data primarily from the Home Office, Department for Business and Trade (DBT) and National Fire Chiefs Council (NFCC). Further evidence will be gathered ahead of a final stage IA.

48. The appraisal period for measuring the impact of the MSL proposals is ten years in line with HM Treasury Green Book (2022) guidance.[footnote 10] A social discount rate of 3.5% is used to discount future values to present values. All costs and benefits are in 2022 to 2023 prices price base year (PBY), with a 2023 to 2024 present value base year (PVBY).

49. Transition/set-up costs are assumed to occur in year one only, and ongoing costs and benefits are expected to occur from year one of the policy onwards. It is hoped that the consultation may provide further data and information to refine the estimates of costs and benefits presented here.

50. The main assumptions used in this IA are listed in table 1 below.

Table 1: Assumptions

ID Assumption description
1.1 It is assumed that the total staff numbers for wholetime firefighters, on-call firefighters and control staff remain constant over the 10-year appraisal period because of a lack of data on how firefighter numbers may change over the next 10 years. This will be further assessed following the consultation. Total staff numbers are assumed to be 32,148 (FTE) and 36,117 (headcount) in England[footnote 11], 7,102 (headcount) in Wales[footnote 12] and 6,057 (FTE) and 6,463 (headcount) in Scotland[footnote 13], giving a total of 45,307 (FTE) and 49,682 (headcount).    
1.2 FRS staff can either be wholetime firefighters, on-call firefighters, control staff or support staff. The definitions for these roles (as per HO statistical releases[footnote 14] are:

Wholetime firefighter: A full-time firefighter.

On-call firefighter: A firefighter responding when required during their ‘on-call’ hours, sometimes called retained duty system.

Fire control: A uniformed member of staff working in a fire and rescue service control centre to answer emergency calls and deal with mobilising, communications, and related activities, regardless of rank.

Support staff: A member of staff employed by the FRA who is not a firefighter or in fire control. It includes, for example, administrative roles, clerical support, analytical support, finance, vehicle maintenance, property management etc.

For the purposes of monetising benefits in this IA, it is assumed that only wholetime firefighters and control staff of the following job ranks would receive work notices and be impacted by MSL: watch manager, crew manager and non-managerial. This is a broad assumption and is made because these are most likely to undertake strike action and assumed to be the most likely to be receive work notices to meet MSL as they have the most directly operational response roles. This assumption will be further reviewed following the consultation, as it is still possible in the legislation that individuals of other ranks could receive work notices. This assumption is also tested in the sensitivity analysis (section G).

On-call firefighters are currently excluded from the benefit analysis due to uncertainties around how many of these individuals are usually on shift at a given time, how many would likely undertake strike action or be included in BCPs, and how work notices would be applied to them.

For the remainder of the IA, this group of firefighters and control staff are referred to as the ‘in-scope workforce’. For the benefits appraisal, only the in-scope workforce is analysed. Given that all firefighters and control staff, regardless of rank would be expected to familiarise themselves with the MSL legislation and any MSL that is set after FRS senior leadership input, wholetime and on-call firefighters of all job ranks, as well as control staff of all job ranks, are included in the estimation of these familiarisation costs.
   
1.3 Assuming that the average shift pattern of wholetime firefighters and control staff is 2-2-4 (2 day shifts, 2 night shifts, 4 days off) it is assumed that, on average, a full-time firefighter/control staff member will work 42 hours a week (48 hours per every 8 days), which is 6 hours per day on average.    
1.4 Assumption 1.3 means that, on average, each wholetime firefighter and control staff will work on average 25% of the time over an average week.[footnote 15]

This means the in-scope workforce for this policy is 6,616 individuals (using the rank assumption in 1.2, and FTE workforce numbers outlined in assumption 1.1).
   
1.5 The turnout for strike action (the proportion of the FTE workforce in scope that would strike) is assumed to be 76.1%, with a low estimate of 64.2% and a high estimate of 88.0%. These estimates are based on the results, received in January 2023, of the FBU ballot to strike: there was a 73% turnout and 88% of those who voted, voted to reject the offer.[footnote 16] Therefore, the high estimate for strike turnout is the proportion of voters who voted for strike action, the low estimate is the proportion of eligible FBU members that voted for strike action[footnote 17], with the central estimate being the midpoint of these 2 figures. It is assumed that turnout is the same in all FRS in Great Britain, with this assessed in Section E: Place-based analysis.    
1.6 One single industrial dispute period is estimated to equate to 295.5 hours of national strike action. This means that, for every period of strike action, there are 295.5 hours of national action in which a proportion of the in-scope workforce are on strike. This estimate is derived from analysis into the amount of action during the 2013 to 2015 firefighter industrial dispute (17 months), which is the last occurrence of national strike action. The total number firefighter/control staff hours this impacts are determined by the number of incidents of strike action that occur over the 10 year appraisal period, the proportion of the in-scope workforce that strike in the do-nothing scenario (option 0), and the level at which the MSL is set.
br>For the purposes of monetising benefits in this IA, in the low estimate scenario, it is assumed that one period of national strike action occurs in the 10 year appraisal period. In the high estimate scenario, it is assumed that 5 periods of national strike action occur in the 10 year appraisal period, and in the central scenario, it is assumed that 3 periods of national strike action occur in the appraisal period. Therefore, it is assumed that 29.6 hours per year of national strike action occur in the low estimate scenario, compared to 147.8 hours of national strike action per year in the high estimate scenario, and 88.7 hours of national strike action per year in the central estimate scenario.[footnote 18]

An example of how these assumptions are used to estimate the number of firefighter/control staff hours saved by the MSL options is given in paragraph 85.
   
1.7 The benefit to society of firefighter/control staff member’s work is proxied by the value a HFSC (low estimate) and responding to non-fire incidents (high estimate), with the midpoint of these two figures being used as the central estimate. The estimate for the value of an hour of firefighter’s labour is derived from an NFCC report[footnote 19] and the calculation for this estimate is explained in paragraphs 70 to 84. These values are assumed to remain constant over the appraisal period.

In the low estimate, it is assumed that any firefighter hours that are saved because of the MSL are used for prevention and protection work (with HFSC being used as a proxy for all prevention and protection work). In the high estimate, it is assumed that any firefighter hours that are saved because of the MSL work will be used to respond to non-fire incidents, with the additional assumption that non-fire incidents will occur at the same rate as they did in the years 2016 to 2017 to 2018 to 2019.

In the appraisal, an hour of control staff time is assumed to have the same benefit as an hour of firefighter’s time. It is also assumed that the benefit of an hour of firefighter/control time is the same for all job ranks in the FRS.

It is assumed that, in the do-nothing approach, all fire incidents will be responded to because of the existence of business continuity plans.
   
1.8 The benefits of the policy are assumed to be spread evenly amongst each of the 10 years of the appraisal period, because it is not possible to predict exactly when and what scale of strike action could occur.    
1.9 For appraisal purposes, it is assumed that all firefighters ranked brigade manager or area manager will be involved in the decision of what level to set the MSL at in options 4 and 5.    
1.10 The labour costs used in the appraisal are shown in table 2 below. The wage costs for non-London firefighters and control staff ranked area manager and below are sourced from the 2021 pay settlement data.[footnote 20] Wage costs for London staff and brigade managers is taken from the Fire Reform White Paper Impact Assessment[footnote 21], and uprated by the 1.5% agreed pay rise in 2021 to bring them to 2022 to 2023 levels. A 30% uplift is applied to FRS wages to account for non-wage labour costs, and the union official labour costs are updated by 17.9%.[footnote 23] The wages for union officials assumed to be the same for London and non-London based officials.    
1.11 There are 4 trade unions for firefighters and control staff. It is assumed that one general secretary and four senior directors from each trade union will need to familiarise themselves with the legislation and be involved in negotiating the MSL for options 4 and 5.[footnote 23]    
1.12 For the purposes of estimating the benefits, the weighted average hourly wage (uprated to account for non-wage labour costs) of watch manager, crew manager and non-managerial firefighters and control staff (the in-scope labour force, across Great Britain) is calculated to be £20.51. The sources for the workforce numbers used to calculate this weighted average are listed in the footnotes to assumption 1.1, and the wages are in Table 2.    

Table 2: Labour costs used in appraisal

Hourly wage, non-London Hourly wage, London Hourly uprated labour costs, non-London Hourly uprated labour costs, London
Firefighters        
Brigade manager £58.13 £58.13 £75.57* £75.57
Area manager £26.44 £41.56 £34.37 £54.03
Group manager £22.52 £31.60 £29.28 £41.08
Station manager £19.56 £27.72 £25.43 £36.04
Watch manager £17.14 £19.59 £22.28 £25.47
Crew manager £16.32 £18.73 £21.22 £24.34
Non-managerial £14.72 £17.16 £19.14 £22.31
Control staff        
Area manager** £26.44 £41.56 £34.37 £54.03
Group manager £21.40 £31.60 £27.82 £41.08
Station manager £18.58 £27.72 £24.15 £36.04
Watch manager £16.28 £19.59 £21.16 £25.47
Crew manager £15.51 £18.73 £20.16 £24.34
Non-managerial £13.99 £17.16 £18.19 £22.31
Union officials        
General Secretary £30.83 £30.83 £36.35*** £36.35
Other senior official £30.83 £30.83 £36.35 £36.35

*Firefighter and control staff wages are uplifted by 30% to account for non-wage labour costs.

**Data on area manager wages for control staff is not published, so the hourly wage for this category of worker is assumed to be the same as that of a firefighter area manager.

***Wages uplifted by 17.9% to account for non-wage labour costs for union officials.[footnote 24]

Appraisal - costs

Option 0: Take no action and make no legislative changes (do-nothing)

51. This is the do-nothing option and so no costs have been monetised. For option 0, no legislation is undertaken and so there is no impact of the proposals. This is the baseline against which all options are measured.

52. There are costs associated from option 0 if looking at it in isolation, because in this option there would a higher level of strike action than in the other options. These costs are instead accounted for in the benefits of each individual option, and option 0 is assumed to be the counterfactual option with no costs or benefits beyond it.

Option 1: Set a MSL such that staffing numbers must never go below a certain level of attendance

Set-up costs for option 1

53. There will be set-up costs in year 1 of the appraisal period from trade unions, FRS staff and employers familiarising themselves with the policy.

Familiarisation – trade unions

54. It is expected that trade unions will have to familiarise themselves with the legislation and any relevant guidance produced to support the policy. It is assumed that it would take between half a day (4 hours) and 2 days (16 hours) in meetings for the union general secretary and four other senior directors, with a best estimate of 1 day (of 8 hours), to familiarise themselves with the proposed policy. This is based on the trade union familiarisation estimates provided in the DBT MSL Bill IA.[footnote 25] Given that the hourly labour cost of union officials is £36.35, it is assumed that 5 union officials per union will need to familiarise themselves with the legislation (see assumption 1.11), and there are four firefighter trade unions, the familiarisation cost is estimated to be between £3,000 and £12,000, with a central estimate of £6,000.

Familiarisation – FRS staff

55. It is expected that all firefighter and control staff in FRS up to group manager level will need to familiarise themselves with the legislation and any relevant guidance produced to support the policy, however this expectation will be further assessed following the consultation. It is assumed that each firefighter (on-call and wholetime) and member of control staff up to this rank will take between 30 minutes and 2 hours, with a central estimate of 1 hour, to familiarise themselves with the proposed policy. The familiarisation cost is calculated by multiplying familiarisation time by the labour cost of each firefighter/control staff member and headcount staff numbers.[footnote 26][footnote 27] This gives an estimate of the familiarisation cost of between £0.5 million and £1.9 million, with a central estimate of £0.9 million.

Familiarisation – FRS senior leadership teams and local employers

56. It is expected that senior FRS staff (brigade and area managers) will have to do the same level of familiarisation as trade union officials (see paragraph 54), so in a range of 4 to 16 hours, central estimate 8 hours. This is because there are similar responsibilities placed on employers and unions by this policy. There may also be a level of familiarisation to local authorities, however this has not been costed at this stage because they are likely to be interested parties as opposed to being directly involved in setting or ensuring MSL and because of uncertainties regarding the number of local authority individuals in scope. This familiarisation cost has been calculated in the same way as the familiarisation costs for other FRS staff, but accounting for the longer familiarisation times.[footnote 28] This gives an estimate of the familiarisation cost of between £0.1 million and £0.3 million, with a central estimate of £0.1 million.

Total set-up costs for option 1

57. Total set-up costs are presented in table 3. The costs are estimated to be between £0.5 million and £2.2 million, with a central estimate of £1.1 million.

Table 3: Total set-up costs option 1, to the nearest million (2022 to 2023 prices)

Cost area Low Central High
Familiarisation – trade unions £0.0 £0.0 £0.0
Familiarisation – FRS staff £0.5 £0.5 £1.9
Familiarisation – FRS senior leadership £1.9 £0.1 £0.3
Total £0.5 £1.1 £2.2

Option 2: Set an MSL so staffing levels are geared to respond to specific risks, and major incidents

58. The set-up costs for option 2 are the same as option 1 as the level of familiarisation is assumed to be the same.

Option 3: Set individual MSL following local leader/organisational input in collaboration with Home Office/Secretary of State (for example, not a national level, but based on local priorities and pressures)

Set-up costs for option 3

59. Option 3 will incur the same familiarisation costs as option 1 and 2, estimated to be between £0.5 million and £2.2 million, with a central estimate of £1.1 million.

60. It will also incur additional set-up costs associated with FRS senior leadership and trade union officials contributing to the setting of MSL at appropriate levels. There may also be some cost to other individuals in local authorities, however this has not been costed at this stage because they are currently not expected to be directly involved in the setting of MSL and because of uncertainties regarding the level of their involvement. This will be further assessed following the consultation.

Cost of setting MSL – FRS senior leadership teams (brigade and area managers)

61. It is expected that it will take between half a week (21 hours) and a week and a half (63 hours), with a central estimate of 1 week (42 hours), for FRS leadership teams to give their input into that the most appropriate MSL is. It is assumed that all brigade managers and all area managers will be involved in this senior input, and that they will all spend the same amount of time on this task. It is possible that some other ranks in FRS, such as group and station managers, could also be involved in this process, however this cost remains uncertain so is not monetised in this IA. The total costs of this are estimated to be between £0.4 million and £1.2 million, central estimate £0.8 million.

Cost of setting MSL – trade unions

62. It is expected that trade union officials will be in negotiations with FRS senior leadership when they are providing their input into setting the MSL. It is assumed that, for each trade union, 1 union general secretary and 4 other senior directors will contribute to these negotiations. Each of these trade union officials are assumed to take the same amount of time as FRS senior leaders (see paragraph 61) in these negotiations. The total costs of this are estimated to be between £15,000 and £46,000, central estimate £31,000.

Cost of setting MSL – FRS staff

63. Once the MSL is set, FRS staff (below brigade and area manager) will have to further familiarise themselves with the MSL and any guidance that accompanies this MSL. This familiarisation cost is in addition to the familiarisation costs detailed in option 2. It is assumed that each FTE firefighter (on-call and wholetime) and member of control staff will take between 30 minutes and 2 hours, with a central estimate of 1 hour, to familiarise themselves with the MSL that is set. The total costs of this are estimated to be between £0.5 and £1.9 million, central estimate £0.9 million.

64. The total set-up costs of option 3 are presented in table 4. They are estimated to be between £1.4 million and £5.2 million, with a central estimate of £2.8 million.

Table 4: Option 3 set-up costs, to the nearest million (2022 to 2023 prices)

Low Central High
Total familiarisation costs (from options 1 and 2) £0.5 £1.1 £2.2
Option 3 cost of setting MSL – FRS senior leadership £0.4 £0.8 £1.2
Option 3 cost of setting MSL – trade unions £0.0 £0.0 £0.0
Option 3 cost of setting MSL – FRS staff £0.5 £0.9 £1.9
Total set-up costs £1.4 £2.8 £5.2

Option 4: Secretary of State/Home Office set a national MSL and chief fire officers and their organisation decide specifics for local area

65. The set-up costs for option 4 are calculated in the same way as those for option 3 as the level of familiarisation and set-up cost is assumed to be the same.

Option 5: Set an MSL so that staffing levels ensure that cover is maintained on high-risk days or at high-risk times

66. The set-up costs for option 5 are the same as options 1 and 2 as the level of familiarisation is assumed to be the same, and there will not be the additional involvement of FRS senior leadership or trade unions as in option 3 and 4.

Total costs

Table 5: Costs for options 0 to 5, to the nearest million (2022 to 2023 prices)

Policy option Familiarisation – trade unions Familiarisation – senior FRS staff Familiarisation – FRS staff Cost of setting MSL – trade unions Cost of setting MSL – FRS senior management Cost of setting MSL – FRS staff Total discounted cost (10 years)
Option 0              
All £0.0 £0.0 £0.0 £0.0 £0.0 £0.0 £0.0
Options 1, 2 and 5              
Low £0.0 £0.1 £0.5 £0.0 £0.0 £0.0 £0.5
Central £0.0 £0.1 £0.9 £0.0 £0.0 £0.0 £1.1
High £0.0 £0.3 £1.9 £0.0 £0.0 £0.0 £2.2
Options 3 and 4              
Low £0.0 £0.1 £0.5 £0.0 £0.0 £0.5 £1.4
Central £0.0 £0.1 £0.9 £0.0 £0.8 £0.9 £0.0
High £0.0 £0.3 £1.9 £0.0 £1.2 £1.9 £5.2

Non-monetised costs

67. There are a number of additional costs, that apply to all these options to various degrees, that should be considered. These will be further assessed ahead of a final stage IA, following the results of the consultations.

  • Enforcement related costs: There may be costs to FRA of enforcing work notices given by employers. This could include administration and litigation costs. There could also be linked costs to trade unions and employees.

  • Trade union membership: It is possible that government setting MSL in FRS could have an adverse impact on union membership by either raising the barrier to industrial action or increasing the strength of mitigating actions. It is also possible that some individuals may currently be reluctant to join a union due to concerns around impact of disproportionate industrial action on the public in the absence of statutory MSL. This legislation may therefore in theory, mean some individuals feel more empowered to join a union as this concern will no longer apply.

  • Increase in strike action in the short term: There is a potential increase in strike action prior to MSL being introduced, as unions may seek to cause disruption which is not mitigated by an MSL before they are implemented, in order to maximise their leverage. This may be mitigated by the costs to unions and their members, principally loss of pay, of taking industrial action.

  • Changing nature of strike action: According to unions, a further consequence of this policy could be the increase in staff taking action short of striking which is not prohibited by this legislation[footnote 29][footnote 30] – if more strike action occurred because of these policies, there would be a societal cost in terms of lost firefighter hours. This is due to the incentive unions have to cause disruption in order to encourage employers to reach a favourable settlement in response to a dispute. Where services are reliant on staff working additional hours, this could have a significant negative impact on the level of FRS provided and it is important to note that such action could continue even when MSL are in place, (so it could be that instead of taking strike action, action short of strike becomes a more prevalent form of lawful protest). Although hard to quantify, this is likely to be less disruptive than industrial action without MSL in place. It may nevertheless lead to a prolongation of the dispute. It is unclear what the net impact of a move from strike action to action short of a strike would be, but it is likely, on balance, to be lower than strike action without any form of MSL. This is because the impact of action short of strike can be mitigated through the changing of some working practices, such as not relying on overtime or ensuring that overtime working is incorporated into the employment contracts of workers. However, as staff would be paid their normal wage and carry out their contracted functions during action short of a strike, it is possible – although hypothetical - for such actions to be more protracted, which could build in a lower level of service or performance (if employers continued to rely on workers’ goodwill).

  • Operational costs of MSL to employers and trade unions: The legislation would include a requirement for employers to inform workers and unions of those workers required to work to provide the minimum levels of service, and to consult unions while selecting the workers required. This would need to be done at least 7 days prior to the strikes starting. This process could lead to costs to both unions, who would be required to take steps to ensure that workers specified to work, as part of a work notice, do not participate in strike action, and costs to employers who would have to inform trade unions of those who need to work. Further details on this potential cost will be provided in the final stage IA.

  • Reduced benefits of being in a union: There are a number of benefits of being part of a union. One of these benefits is that unions help counterbalance the bargaining power that employers have over their staff. Strike action may in some cases lead to improved terms and conditions, including increased pay deals, which can have impacts of staff morale and motivation. If any of the proposed options were to change the balance between unions and employers, this may reduce the value that workers receive by being part of a union. If any of the options reduce the impacts of strikes, this could lead to potential reductions in future pay or working conditions for FRS staff compared with option 0. This potential reduction in terms and conditions for workers in unionised sectors over time (if bargaining power is substantially weakened) could have a downward effect on terms and conditions more generally in the labour market.

There will be fewer instances of pay being withdrawn on the basis of striking when comparing an individual strike in the counterfactual scenario to introducing an MSL. The net effect of such over a certain period of time is uncertain, as this is dependent on the extent to which strike action occurs and MSLs are applied and how they vary from any BCP which would have already been in place.

Workers who strike are not paid by employers for the period they are taking industrial action. If the MSL results in fewer individuals involved in strike action, employers would have reduced instances of withheld pay. Individuals who wanted to strike, but were unable to due to an MSL, would retain their pay for that strike period, but may also incur a cost (given that the counterfactual is that the worker may have preferred and wanted to strike). There is no guarantee that strike action leads to more favourable terms and conditions for workers and the proposed options for MSLs protects the ability for workforces to strike. If fewer strikes were successful in achieving improved terms and conditions as a result of MSLs, that would represent a cost to the worker. If MSLs are set at a level similar to BCP which would already have been in place, there could be limited material difference to individuals.

Benefits

Set-up benefits (all options)

68. There are no set-up benefits associated with any of these options. All benefits are assumed to be ongoing.

Ongoing and total benefits

Overall method for all options 1 to 5

70. There will be ongoing benefits across the 10 year appraisal. These are benefits from an increase in hours worked by firefighters who would have otherwise undertaken strike action. These additional hours provide economic value to society, if it is assumed that the work that firefighters and control staff do has a benefit to society through the improvement of public safety, above their wage. The wage of firefighters is excluded from this analysis as it represents an economic transfer because it is a cost to the government, and a benefit to individuals. For simplicity, it is not put on both sides of the analysis, and instead net benefits are used in this section. The savings to FRA during strike action of firefighters not working could be classed as a benefit in option 0 (do nothing), with full values used in this section, however the total NPV would remain unchanged with this method.

71. As long as the MSL is greater than the percentage of the workforce that would have attended work during an industrial dispute (not undertaking strike action), then the MSL will prevent some firefighters and control staff from striking to the full extent that they would have done had the MSL not been in place. The net value of the hours of firefighter and control staff work that are gained because of the MSL represent a benefit to society and are monetised in this section.

72. The broad calculation used is number of hours saved multiplied by the net benefit of firefighter/control staff. The net benefit is calculated using two different methods. The value of an HFSC is the low estimate, and the value of a response to a non-fire incident is the high estimate. These are used because it is assumed that during strike action, BCPs (option 0) will still allow FRS to respond to fire incidents, albeit potentially with slower response times and with weakened response capability (discussed in the non-monetised benefit section of this IA). However, they will likely de-prioritise and not undertake certain prevention activity and/or attend non-fire incidents. It is not currently expected that prevention and protection activities undertaken by FRS, such as HFSC, will be subject to MSL. They are just used a proxy for the value of firefighter hours in this context. The consultation asks respondents to consider if any further activities should be brought into scope of MSL so this will be further considered following the consultation.

73. These estimates remain broad proxies for the level of value that could be lost during strike action, with large uncertainty, as there is very limited data available to monetise this. These will be further refined in the final IA. The mid-point of the 2 values gives the central estimate, and the methodology to calculate them is explain in full below.

Monetising net benefit per individual firefighter hour

High estimate for net benefit per individual firefighter hour – non-fire incident response

74. The high estimate is estimated assuming that there are certain non-fire incidents that would not be responded to by FRS during strike action but would be responded to with additional staff (due to a MSL being in place). This is a high estimate because the level of MSL set will determine whether FRS have capability to respond to more non-fire incidents (for example an MSL of 5 per cent more staff than what they would have had during strike action may not mean they can respond to the same non-fire incidents as ‘normal’ times). This is the best possible proxy for what additional response could incur in terms of benefits. This could also be considered as a broad proxy for improved fire response; however, this needs to be further refined following the consultation.

75. To estimate this value, data from the NFCC report on the value of UK FRS[[footnote 31] is used which estimated the value from non-fire incidents between 2016 to 2017 and 2018 to 2019 to be approximately £9.6 billion.[footnote 32] This is estimated by multiplying the DfT value of a fatality (cited in the report as £2.1 million) by an estimation of the number of lives saved from non-fire incidents. To get the cost into 2022/23 prices, in line with the rest of this report, it is uprated by approximately 15 per cent.[footnote 33] This gives an equivalent value of £11.1 billion, assuming that the same number of lives would be saved during this period and during any potential strike action. If the frequency of non-fire incidents were to be higher than over this period, the value could be higher than this, because there would be a larger total impact from firefighter response.

76. To estimate an hourly value of non-fire incident response, the number of hours worked by firefighters undertaking strike action needs to be estimated. It has been assumed in this IA that strike action would be undertaken by wholetime firefighters and control staff of watch manager to non-managerial firefighter rank. To calculate this, the number of annual hours worked per firefighter (6,570[footnote 34] is multiplied by the average number of staff in England (as this is the scope of the NFCC report) at these ranks over the period (21,705[footnote 35] to give approximately 143 million hours worked over the period.

77. To estimate the value per hour worked, £11.1 billion is divided by 143 million hours to give £78 of value in total per hour. This assumes that non-fire incidents are equally likely to occur during any period of strike action, and the straight division with just these hours can be done because it is assumed without either these ranks of firefighter or control staff, no response would happen. Therefore, all this benefit is marginal based on these individuals being available and on shift.

78. This value is then reduced by the average labour cost of a firefighter/control staff member of watch manager and below (£20.51, assumption 1.12) to give the estimate of the value of one hour of individual time to be £57. This estimate has significant uncertainty in it, so it is tested in the sensitivity analysis, section G. The net value estimate is then multiplied by total staff hours gained from MSL to estimate the total high benefit estimate in each option.

Low estimate for net benefit per individual firefighter hour – home fire safety checks (HFSC)

79. The low estimate is calculated using estimated benefits from HFSC, a type of prevention activity that would likely not be undertaken during strike action. NFCC regression modelling found that on average, a targeted home visit contributes to a reduction in the incidence of accidental dwelling fires with a coefficient of -0.00422.[footnote 36] The published research report then monetised this coefficient using a unit cost of fire from the Greater Manchester Combined Authority (GMCA) unit cost database[[footnote 37], giving an estimate of the value of an HFSC to be approximately £226.[footnote 38] To be consistent with the modelling approach taken throughout this IA, this broad approach is followed, however, the available cost of a dwelling fire in the GMCA unit cost database in 2019/20 prices is uprated to 2022 to 2023 prices using the GDP deflator[footnote 39] to give a figure of £49,250. This is multiplied this by the coefficient, to get a value of an HFSC as £208.

80. This estimation is highly indicative and relies on the NFCC coefficient analysis and a single estimation for the unit cost of dwelling fires. The coefficient of HFSC and their impact on fires is likely to vary significantly depending on how, and who undertakes the HFSC, and the cost of fire estimation used here does not fully account for the changing nature and number of fires in recent years. This is therefore a highly indicative method, that will be improved in the final IA. This figure is tested in the sensitivity analysis (section G).

81. As completed in the high estimate, the wage cost of an HFSC is taken from this value so it only provides a net benefit estimate. It is assumed that it will take two firefighters, two hours to complete a HFSC[[footnote 40], equivalent to 4 firefighter hours to complete an HFSC. The average hourly labour cost of a firefighter/control staff member in-scope is assumed to be £20.51 (assumption 1.12) so the cost of an HFSC is estimated to be £82.

82. The net benefit of an HFSC is estimated to be £126 (£208 - £82). This can be converted into benefit per hour by dividing this by 4 (as there are 4 firefighter hours spent on each HFSC). This gives a net benefit per hour of £31.

Central estimate for net benefit per individual firefighter hour

83. The central estimate for the benefit of an hour of firefighter time that is saved from strike action is the mid-point of the low and high estimate, so is £44.

These estimates all remain highly indicative and should not be interpreted as the actual activity that will be completed with MSL in all FRS, at all times. They are broad measures of the value firefighters provide to society using two different metrics and could be considered proxy for the activity they do, or the resilience they provide by being available. Although control staff are not expected to undertake prevention activity, this can be used as a broad proxy for some of their activity. They are included in the calculations of non-fire benefit because they are required to facilitate an effective fire response.

Individual hours saved

85. It is assumed that each national strike action occurrence consists of 295.5 hours where there is a national firefighter strike taking place (assumption 1.6). This estimate is derived from analysis of the 2013 to 2015 firefighter strike action. The number of individual hours of firefighter/control staff time saved by an MSL is determined by the MSL level, the strike turnout in the do-nothing option, and the assumed number of hours for which there is a firefighter strike taking place. As a worked example:

  • In the central estimate scenario for option 1, it is assumed that the MSL is 37.5%, potential strike turnout is 76.1%, meaning that 13.6% of the in-scope workforce cannot strike because of the MSL.

  • It is also assumed that 3 strike incidents occur over the 10-year appraisal period, so a total of 886.5 hours in which national strike action is taking place over (88.7 a year).

  • Given that there are approximately 6,616 wholetime firefighters and control staff working at any one time[footnote 41] (assuming all individuals work a 2-2-4 shift system, so each individual is on shift 25% of the time), this leads to a total of 79,881 individual firefighter/control staff hours being saved per year (6,616 * 88.7* 13.6%).

  • It is assumed that strike action is distributed equally across the 10-year appraisal period, and the same number of firefighters and control staff are in-post each year, so 79,881 hours are assumed to be saved per year.

Total benefits

86. Table 6 shows the total benefits for each of the policy options 0 to 5. In the low scenario, the MSL for each option are not high enough compared to the intended strike turnout for the policy to prevent any hours of strike action, meaning there are no monetised benefits from the policies in the low estimate. In the central and high scenarios, the number of hours saved is multiplied by the estimates for the value of an hour of individual strike action prevented to give a total benefit estimation. For example, the central estimate of option 1 is calculated as 79,881 * £44[footnote 42] = £3.5 million per year, which equates to £30.4 million in present value terms (discounted) over 10 years.

Table 6: Benefits for options 0 to 5 (2022 to 2023 prices)

Policy option Assumed MSL level Intended strike turnout Percentage of workforce prevented from striking Hours saved per year Total benefit per year Total discounted benefit (10 years)
Option 0            
Low 0% 64.2% 0% 0 £0 £0
Central 0% 76.1% 0% 0 £0 £0
High 0% 88.0% 0% 0 £0 £0
Option 1            
Low 25% 64.2% 0% 0 £0 £0
Central 37.5% 76.1% 13.6% 79,881 £3.5m £30.4m
High 50% 88.0% 88.0% 371,448 £21.2m £182.5m
Option 2            
Low 47% 64.2% 11.3% 22,059 £0.7m £6.0m
Central 77% 76.1% 53.1% 311,178 £13.8m £118.6m
High 90% 90% 78.3% 765,541 £43.7m £376.1m
Option 3            
All - - - - - -
Option 4            
Low 30% 64.2% 0% 0 £0 £0
Central 43% 43% 18.6% 109,206 £4.8m £41.6m
High 55% 88.0% 43.0% 420,322 £24.0m £206.5m
Option 5            
Low 26% 64.2% 0% 0 £0 £0
Central 41% 76.1% 17.0% 99,966 £4.4m £38.1m
High 63% 88.0% 50.5% 493,634 £28.2m £242.5m

Summary of ongoing and total benefits for each option

Option 1: Set a MSL such that staffing numbers must never go below a certain level of attendance

87. The total hours saved per year from option 1 is calculated to be between 0 and 371,448, with a central estimate of 79,881. This equates to a total discounted monetary benefit (over 10 years) of between £0 and £182.5 million with a central estimate of £30.4 million.

Option 2: Set an MSL such that staffing levels must be geared to respond to specific risks and major incidents

88. The benefits of this option are calculated using the same approach as option 1 but with the assumption that MSL levels are set between 47% and 90 per cent, central estimate 77%. The total hours saved per year from option 2 is calculated to be between 22,059 and 765,541, with a central estimate of 311,178. This equates to a total discounted monetary benefit (over 10 years) of between £6.0 million and £376.1 million with a central estimate of £118.6 million.

Option 3: Set individual MSL following local leader/organisational input in collaboration with Home Office

89. No benefits have been monetised for this option due to uncertainty on MSL that would be set.

Option 4: Secretary of State/Home Office set a national MSL and chief fire officers and their organisation decide specifics for local area

90. The benefits of this option are calculated using the same approach as option 1 but with the assumption that MSL levels are set between 30% and 55%, central estimate 43%. The total hours saved per year from option 4 is calculated to be between 0 and 420,322, with a central estimate of 109,206. This equates to a total discounted monetary benefit (over 10 years) of between £0 and £206.5 million with a central estimate of £41.6 million.

Option 5: Set an MSL so that staffing levels ensure that cover is maintained on high-risk days or at high-risk times

91. The benefits of this option are calculated using the same approach as option 1 but with the assumption that MSL levels are set between 26% and 63%, central estimate 41%. The total hours saved per year from option 5 is calculated to be between 0 and 493,634, with a central estimate of 99,966. This equates to a total discounted monetary benefit (over 10 years) of between £0 and £242.5 million with a central estimate of £38.1 million.

Non-monetised benefits

Reduction in costs the FRS

92. MSL will reduce the public cost of FRA arranging contracts with private contractors or third parties order to bolster or secure their BCP. They may also incur costs from employing resilience fire cover. These can often come at a significant cost, which varies significantly across FRS. MSL will likely reduce these costs, however these costs are highly uncertain, and commercially sensitive for FRS, so have not been monetised in this IA.

93. MSL will also lead to benefits in-terms of removing the need for military assistance during strike action, which can be provided to support FRS and prevent any FRS being overwhelmed during strike action. This assistance has a financial and economic cost on society (as it takes military personnel away from their own duties), and so the reduction in this assistance has benefits. This cost is highly uncertain as it will vary significantly depending on the type of strike action that occurs, so it has not been monetised in this IA. Further work on monetising this cost will be completed following the consultation.

94. MSL may also provide additional certainty to areas and FRS, allowing local leaders to more easily plan for periods of strike action by giving them a clear indication of how many staff they will have available on strike days.

Improved public safety

95. The overarching aim of the options considered in this IA is to improve public safety, so this is one of the most significant potential benefits of each of these options. Presuming that the introduction of MSL either reduces the amount of strike action or ensures that a higher level of service is in place in FRS on days where strike action is occurring compared to what would have happened without MSL, there will be public safety benefits from these options. These could include:

  • Improved FRS resilience and ability to respond to fire incidents. If it is assumed that FRS operate at an efficient level in which they are resourced (in terms of firefighters) to sufficiently mitigate fire and rescue risk, any reduction in the number of staff available below ‘normal’ levels will be accompanied by an increased risk of FRS inadequately responding to fire and rescue incidents. However, current business continuity plan arrangements under option 0, the counterfactual, do ensure that during current strike action, each FRS will maintain a level of fire incident response capability. This can be through non-striking staff, contractors, military personnel and/or improved call triaging. Therefore, the public safety benefits are more about improved response quality, as opposed to more fire incidents being responded to. With MSL, there would likely be more available firefighters. This will have response benefits in terms of reduced response times to fire incidents (due to wider FRS cover, and potentially because there will be wholetime cover in some FRS instead of on-call), and potential improved quality and weight (number of appliances) of response.

  • An increase in the number of smaller fire incidents and non-fire incidents attended during strike action and/or the amount of fire safety activity completed will increase public safety. Having more firefighters available during strike action will have benefits because they will be able to attend more incidents and do more activity. As discussed above, any reduction in the number of available firefighters will increase public safety risk. However, it is likely this impact is non-linear (an FRS operating with one less member of staff will have less of an impact than if multiple staff were unavailable, which could mean whole appliance had insufficient crew members to be available for use). So, the public safety impacts may be delayed or non-fire related because FRS may be able to operate at lower staff levels at the expense of attending non-fire incidents or doing training/prevention and protection activity. This may have longer term negative impacts, and will depend on how many fires occur on a given strike action day. In FRS, there is a difference between potential and actual risk on each day – for example some days there may be few fires, and so FRS do not respond to many incidents and not all staff are needed, so there is no impact on public safety from strike action, and no ‘actual risk’ because the number of firefighters simultaneously mobilised never exceeds their business continuity plans. However, there is always a chance that a major incident could happen, or a fire could happen, so the ‘potential’ risk on these days will be larger with fewer staff.

96. For these reasons, it is not possible to cost the public safety benefits of each option, because these will depend on all the complex factors above. However, there will certainly be a level of public safety benefit as long as any firefighter hours are saved.

97. The level of benefit in options 1 to 5 is completely dependent on what occurs during the counterfactual scenario assumed in option 0, do nothing. If in the counterfactual, no strikes would have happened (as has been the case since 2015), then no public safety benefits or hours saved benefits are achieved from this legislation. In addition, the level of FRS business continuity plans (BCP) under option 0 also determines the public safety benefits of the other options. If BCP levels are the same level as MSL, then in some FRS, there will be no benefits of this legislation apart from formalising current agreements. In addition, if all fires can still be attended by staff not undertaking strike action, or by contractor/army officials, the public safety benefits are lower than if this isn’t the case. In addition, if there are robust return to work agreements during strike action, then the benefits of the other options are limited. However, the public safety benefits would still be above zero in these scenarios because although return to work agreements allow a firefighter response, they may lead to significantly higher response times than would happen with an MSL and employed FRS firefighters are likely to be more experienced and have certain offensive firefighting skills that alternative staff would not have, meaning the quality of response would be better in this scenario. There are also risks posed by non-fire incidents and a loss of other activity (as outlined in paragraph 95), which would be somewhat mitigated by an MSL.

98. Although this IA cannot put exact public safety benefits on each of the options, they can still be ranked in order of public safety achieved (assuming all other assumptions remain the same). This will be in the same order as the benefits from hours saved of the policy because it is assumed that there is a direct relationship between the 2. Option 1, which incurs the lowest level of working hours saved, will have the lowest level of public safety benefit followed by option 5. Options 2 and 4 will incur the highest level of public safety benefit, however, this will somewhat depend on whether the low, central, or high scenarios are incurred as this impacts the level of benefit. MSL in option 3 remain uncertain, so it is not assessed in the IA. Option 2 will likely incur more benefits than option 4 as it covers all major incidents unless option 4 significantly reduces strike hours (under the high scenario) and/or no major incidents occur during days of strike action.

99. There are also a number of other wider non-monetised benefits of these options:

  • Improved economic output: If response to fires by FRS is better and quicker because of more firefighters being available, then there will likely be less property damage and business impacts from fires, therefore improving economic output.

  • A reduction in the fear of fire: Any firefighter strike action is likely to get significant media coverage, and so individuals will likely know that fire response cover will be reduced on these days. In 2020 to 2021, DLUHC’s English Housing Survey found that many individuals, especially in high-rise buildings, are fearful of fire.[footnote 43] It is possible therefore that firefighter strike action could increase this fear, which would impact individuals’ wellbeing. This has not been monetised at this stage.

  • Improved firefighter safety during strike action: If there are more staff members available during strike action because of MSL, then there may be a benefit from them in terms of ensuring responsible crewing levels to maintain firefighter safety during strike action.

NPSV, BNPV, EANDCB

Net present social value (NPSV)

100. The costs, benefits, and NPSV of these policies (over 10 years) are presented in table 7. The range in the NPSVs is calculated by comparing the low cost estimates, to the high benefit estimates, and the high cost estimates to the low benefit estimates, to give a full potential range.

Table 7: NPSV for each policy option, millions (PV terms over 10 years)

Costs

Low Central High
Option 0 £0 £0 £0
Option 1 £0.5m £1.1m £2.2m
Option 2 £0.5m £1.1m £2.2m
Option 3 £1.4m £2.8m £5.2m
Option 4 £1.4m £2.8m £5.2m
Option 5 £0.5m £1.1m £2.2m

Benefits

Low Central High
Option 0 £0 £0 £0
Option 1 £0 £30.4m £182.5m
Option 2 £6.0m £118.6m £376.1m
Option 3 - - -
Option 4 £0 £41.6m £206.5m
Option 5 £0 £38.1m £242.5m

NPSV

Low Central High
Option 0 £0 £0 £0
Option 1 -£2.2m £29.3m £182.0m
Option 2 £3.8m £117.5m £375.6m
Option 3 -£5.2m -£2.8m -£1.4m
Option 4 -£5.2m £38.8m £205.1m
Option 5 -£2.2m £37.0m £242.0m

Equivalent annual net direct cost to business (EANDCB) and business net present value (BNPV)

101. The business net present value (BNPV) and equivalent annual net direct cost to business (defined as EANDCB) of these options are calculated based on trade union familiarisation costs. For options 1, 2 and 5, the cost to business is estimated to be between £3,000 to £12,000 and so the BNPV is between -£3,000 to -£12,000. All these costs occur in year 1, and so the EANDCB is considered to be minimal. For options 3 and 4, the BNPV is marginally lower, at -£18,000 to -£57,000. Again, all these costs occur in year 1, and so the EANDCB is considered to be minimal.

Value for money (VfM)

102. For a policy to be considered value for money (VfM), it must meet its strategic and policy objectives. In the central and high scenarios, Options 1, 2, 4 and 5 meet the strategic and policy objectives of MSL, however, in the low benefits scenarios only option 2 achieves VfM according the NPSV calculations in Table 7 (the NPSV of all the other options is negative). This is because there are no monetised benefits associated with these options in this scenario, because the MSL level is assumed to be set lower than the number of available staff during strike action, on average over the year,[footnote 44] and so no staff hours are assumed to be saved from MSL. As this is the only monetised benefit in this IA, this option does not have monetised benefits that outweigh its costs. However, as there are a number of non-monetised benefits and costs, it is not possible to fully assess the VfM of these options. No benefits are monetised for option 3, so it is not possible to assess the VfM of this option.

103. The monetised VfM assessment in this IA is mainly determined by how many additional hours of firefighter time worked MSL leads to. If MSL are lower than what would have happened without it (in terms of strike turnout), the legislation is not likely to be VfM using the monetised benefit approach in this IA, because it doesn’t lead to additional hours worked over the counterfactual. It will only incur monetised costs, and no monetised benefits. However, if the MSL level set is higher than the number of firefighters or control staff who would be available on days on strike action, then the policy would likely be VfM, as set-up costs are low, and it only requires a few additional hours of work for these costs to be outweighed by the benefits (see risks section G).

104. The bigger the difference available firefighters or control staff on strike days, and MSL, (assuming the MSL is higher than available staff on strike days), the more an option will be VfM. This is shown in Table 7. In the low scenario, the MSL levels are all below expected availability on strike days and so no option is VfM. In the central and high option, all MSL are higher than availability, so they are all VfM. So, from the monetised analysis in this IA, the option with the highest MSL will be viewed as the most VfM, because it only accounts for hours saved. Of the options with monetised benefits, option 2 is the most VfM, whereas option 4 (which has additional set-up costs) is the least.

105. However, there are a large number of costs and benefits that are not monetised in this IA, and so the VfM calculation is more complex in reality than what is presented in this IA. On the costs side, the reduced benefits of being in a union, and the changing nature of strike action are not monetised, and on the benefits side, the impact on public safety, and cost savings from BCP no longer being required, are not monetised. These additional considerations could have a significant impact on which option is actually the most VfM. It would be logical to assume that some of these costs (especially the reduction in benefits of being in a union and reduced collective bargaining power) would increase as MSL increase, being the options with the highest monetised benefits would have the highest amount of these costs, however the linearity of this relationship is not assessed, and it is likely that these options with high MSL would also incur more non-monetised benefits.

106. This lack of monetisation makes any full VfM assessment of the options difficult at this stage. For the reasons outlined above, depending on the relationship between MSL and union benefits and cost saving benefits, it could be the case that any of option 0 to 5 is ultimately the most VfM.

107. This IA is not able to identify, at this stage, what the most VfM option is. Further work needs to be completed to assess the impact of each option on FRS staff, and it is hoped that the consultation will be able to assist with this. The consultation process will allow individuals to voice their opinions on different options, so a fuller assessment of VfM can be completed in the final-stage IA. The purpose of this IA is to provide an indicative estimation of costs and benefits which can be monetised, so that individuals can make informed responses to the consultation The IA identifies the potential costs and benefits of each option, so of which cannot be monetised, and allows individuals, depending on their views on the four crucial assumptions in this IA (MSL level, strike turnout in the counterfactual, reduction in benefits from being a union, and cost savings from BCP), come to a view on which option is best, and provide this view in the consultation. As costs and benefits are not fully monetised in this IA, it is not possible to accurately determine which option offers the highest benefit cost ratio at this stage.

Place-based analysis

108. This policy does not have any specific spatial objectives, however there are a number of place-based impacts that are worth considering. Currently the analysis in this IA looks at FRS in Great Britain as a single entity, however there are 48 FRS across Great Britain, which may be impacted differently by these options. A more complete place-based analysis will be completed in the final-stage IA, accompanying any secondary legislation, however for now, initial areas for consideration are presented:

  • Impact on public safety of option 1: This option proposes a set percentage level being applied nationally across all FRS. Although simple in principle, this may have divergent benefits and impacts on public safety, depending on what the level is set at. Each FRS has different levels of ‘normal’ appliance utilisation, with some more frequently operating with more appliances mobilised than others. This means that a single set percentage that may mitigate most risk in one FRS, may not mitigate risk fully in another. Alternatively, it may lead to the MSL being set at a level to mitigate risk in the FRS that needs the highest percentage, which may lead to inefficiencies.

  • Local differences in trade union membership: This IA has also assumed that any action will be distributed equally across the country and be co-ordinated by all union members. However, in reality, trade union membership as a proportion of total firefighter workforce will vary across the country. This could be impacted by wholetime/on-call split of workforce. The impact of this is that some areas may currently (in option 0) have more limited local impacts of strike action, because they have fewer staff who will go on strike. In these areas, the benefits of these options will be lower. Equally, there are some areas which are currently more impacted by strike action because of the likelihood of their local service, or station, undertaking strike action. These areas may experience more benefits from the legislation, as they will have better, and more secured business continuity plans in the event of strike action which will provide them with public safety benefits. However, they may also experience more ongoing local strike action in future. This could also lead to differing local employment conditions, depending on local collective bargaining power.

  • Impact on public safety of option 4 and 5: Although options 4 and 5 may lead to less local discrepancies than option 1, for the reasons outlined above, they may also lead to place-based impacts because each CFO, local leader, and/or their organisations will set MSL, or decide specifics. This means that in practice, MSL could differ significantly across the country, depending on local risk thresholds, local politics, and local union relationships. This could lead to differing levels of public safety during strike action across the country, and also lead to different levels of local collective bargaining power. The latter could have unintended consequences in terms of recruitment and retention, as staff may consider MSL levels and strike ability when applying for roles.

Impact on small and micro-businesses

109. The legislation will impact four trade unions – the Fire Brigade Union (FBU), the Fire and Rescue Services Association (FRSA), the Fire Officers Association (FOA) and the Fire Leaders Association – each will incur familiarisation costs from the proposals. These are the only businesses in-scope of the legislation. Analysis of the annual returns of the four in-scope trade unions suggests that these are all small or micro businesses, as although they have large membership numbers, they have few staff.[footnote 45]

110. As trade unions are crucial to this legislation, and all are small businesses, it is not possible to exclude small businesses from this legislation whilst still meeting the policy’s objectives and without undermining the policy. The impact of the policy is not expected to be disproportionate on small businesses because the cost to trade unions is expected to be relatively low (no more than £60,000, see table 12).

111. The consultation will be used to seek further information from trade unions on whether they face any significant disproportionate impacts.

F. Proportionality

112. The level of analysis in this IA is considered proportionate at this stage, ahead of receiving consultation responses. There are a number of significant uncertainties in this analysis, and a number of assumptions have been made based on limited evidence. The best available data is used in the analysis, informed by experience and expertise. The consultation will seek input from stakeholders and the public. It is hoped that this will allow refinement and a strengthened assessment of the potential impacts of these policy changes in the final IA, ahead of any legislation.

G. Risks

113. The monetised benefits of this impact assessment assumes that strike hours will be prevented as a result of this policy. Any displacement of strike hours (e.g., through action short of strike, or an increase in the volume of strikes) will reduce the NPSV of this policy, and have not been monetised.

114. This impact assessment makes no assumption about the distribution of costs and benefits on individuals. Future strike action under MSL may only be carried out by certain individuals outside of MSL who are not served work notices. These individuals would bear the cost of striking (in terms of foregone wages), however all unionised individuals employed by the FRS, even those who did not strike, would gain from the results of strike action.

115. As responding to major incidents will likely require certain experienced FRS personnel, and potentially specialist equipment like aerial ladder platforms, option 2 could mean that some staff (those which are trained and required for major incident response) are more likely to receive work notices, and so are less likely to be able to undertake strike action, when compared with other staff. This IA has not assessed the impact of this option on staff willingness to undertake the training required for MSL critical roles.

Sensitivity analysis – strike turnout

116. The appraisal contains assumptions on the proportion of the in-scope workforce that will withhold their labour during a strike incident (referred to as ‘strike turnout’), and assumptions on the number of hours that FRS will go on strike for in the event of a strike incident (see assumptions 1.5 and 1.6). To demonstrate the impact that altering these assumptions has on the NPV, sensitivity analyses has been produced.

117. Table 8 shows what happens when, holding all else constant, the assumption on strike turnout is changed. For the example in table 8, it is assumed that the MSL is set at 60% for purely illustrative purposes. All other assumptions are held constant in line with the central estimates in this IA (so there are 88.7 national hours of strike action per year, the central estimates for set up costs are used, and the net benefit of an hour of firefighter labour saved is assumed to be the central estimate of £44).

118. Table 8 shows that when strike turnout increases, the total number of firefighter/control staff hours that are saved increases. This is because, with a higher a strike turnout, more firefighters and control staff will strike in the do-nothing scenario (option 0). Therefore, the same MSL will prevent more of the workforce from going on strike, so prevent more hours of firefighter/control staff labour being lost to strike action. As table 8 shows, a higher intended strike turnout leads to a larger NPV for the same policy option because benefits are calculated as a function of strike hours saved (see paragraph 85).

Table 8: Sensitivity analysis – proportion of individuals who will strike assuming a 60% MSL and central cost and benefit assumptions to the nearest thousand hours and £ million

Strike turnout Total individual hours saved (over 10-year appraisal period) NPV – with cost of options 1, 2 and 5 NPV – with cost of option 3 and 4
≤40.0%[footnote 46] 0 -£1.1 -£2.8
50.0% 586,500 £21.3 £19.5
60.0% 1,173,000 £43.6 £41.9
61.0% 1,231,600 £45.8 £44.1
63.0% 1,348,900 £50.3 £48.6
65.0% 1,466,200 £54.8 £53.0
67.5% 1,612,900 £60.4 £58.6
70.0% 1,759,500 £66.0 £64.2
75.0% 2,052,700 £77.1 £75.4
80.0% 2,346,000 £88.3 £86.6
85.0% 2,639,200 £99.5 £97.7
90.0% 2,932,500 £110.6 £108.9
95.0% 3,225,700 £121.8 £120.1
100.0% 3,519,000 £133.0 £131.3

Sensitivity analysis – number of strike hours per year

119. Table 9 shows what happens when, holding all else constant, the assumed number of hours for which national strike action takes place per year in the do-nothing scenario (option 0) is changed. The same assumptions are used that are used in the central estimate, except strike turnout is held constant at the central estimate of 76.1%, and the number of strike hours per year is varied. Table 9 gives the low (29.6 hours), central (88.7 hours) and high (147.8 hours) estimates for the number of hours for which national strike action occurs per year, used in the NPV calculations for each policy option, plus a range around these assumptions. In all scenarios, the number of strike hours per year is calculated by combining the number of strike incidents that would in the 10-year appraisal period and the number of hours of national strike action in a strike incident into one assumption on the number of hours of strike action per year. Therefore, table 9 shows what will happen to the NPV when the assumption on the number of strikes and/or the length of strike action is assumed to be higher or lower than that which is assumed in the assumptions table 1.

Table 9: Sensitivity analysis – number of strike hours per year, and updated NPV of the policy

Number of national strike hours per year Total individual hours saved (over 10-year appraisal period) NPV – with cost of options 1, 2 and 5 (£m) NPV – with cost of option 3 and 4 (£m)
0 0 -£1.1 -£2.8
10 239,000 £8.0 £6.3
25 597,400 £21.7 £19.9
29.6 - Low 706,100 £25.8 £24.1
50 1,194,800 £44.4 £42.7
75 1,792,200 £67.2 £65.5
88.7 - Central 2,118,400 £79.6 £77.9
125 2,987,100 £112.7 £111.0
147.8 - High 3,530,700 £133.4 £131.7
150 3,584,500 £135.5 £133.8
200 4,779,300 £181.0 £179.3
400 9,558,600 £363.1 £361.4
500 11,948,300 £454.2 £452.4

120. Table 9 shows that, when the number of strike hours per year in the do-nothing scenario (option 0) increases, the total number of hours of firefighter/control staff hours labour that is saved by the MSL increases. This is because, when the assumed number of strike hours increases, the policy will prevent more hours of strike action from going ahead, meaning fewer hours of firefighter/control staff work will be lost to strike action. This will increase the NPV. Even preventing 10 national strike hours per year, means that the NPV of all options is positive.

Sensitivity analysis – value of a firefighter hour

121. The net benefit of a firefighter/control staff hour saved is estimated using the proxies of a value of a HFSC and the value of responding to a non-fire incident, however the actual value of an hour of firefighter labour that is saved from preventing strike action is highly uncertain. As a result, a sensitivity analysis is given in table 10 that shows how the NPV for each option changes when the value per hour of firefighter/control staff time is changed. Table 10 also gives the low, central, and high estimates for the value of a firefighter/control staff hour that is used in the main NPV estimates.

122. For the sensitivity analysis in table 10, all the same central cost and benefit assumptions are applied as in the appraisal section of this IA. Therefore, the scenario with a £44.27 value of labour, aligns with the central estimates presented in the appraisal section.

Table 10: Sensitivity analysis – value per firefighter hour, options 1, 2, 4 and 5, using central assumptions for costs and benefits, £ million

Value per hour assumption NPV – Option 1 NPV – Option 2 NPV – Option 4 NPV – Option 5
£5 £2.3 £12.3 £1.9 £3.2
£10 £5.8 £25.7 £6.6 £7.5
£20 £12.7 £52.5 £16.0 £16.1
£31 – Low £20.5 £83.2 £26.7 £26.0
£40 £26.4 £106.1 £34.8 £33.3
£44 – Central £29.3 £117.5 £38.8 £37.0
£50 £33.3 £132.8 £44.2 £41.9
£57 – High £38.2 £151.8 £50.8 £48.0
£80 £53.9 £213.2 £72.4 £67.7
£100 £67.7 £266.8 £91.2 £85.0

Central value per hour NPV aligns with the main central NPV estimates, while low/high estimates do not align with their respective main NPV estimates because the central estimates for other variables are used for the purposes of this sensitivity analysis.

123. Table 10 shows that when the estimate for the value of a firefighter/control staff hour increases, the NPV for each of the policy options increases because the benefit to society of preventing an hour of a firefighter striking increases. This means that each policy option will deliver a higher level of benefits to society, all other things being equal, if the value to society of an hour of a firefighter’s work is assessed to be higher than what is monetised in this IA, and lower if it is deemed to be lower.

Sensitivity analysis – workforce in scope

124. When estimating the benefit from the firefighter/control staff hours saved, the analysis in this IA assumes that only hours from firefighters and control staff roles (of rank non-managerial, crew manager or watch manager) are saved as a result of MSL, as these are the individuals most likely to be subject to MSL as they have the most directly operational response roles, and undertake strike action (Assumption 1.2). This is a broad assumption, though, as any role could undertake strike action, and the legislation allows for individuals of any rank to receive work notices. To account for the uncertainty of this assumption, sensitivity analysis was completed that demonstrates the impact that changing the assumed in-scope workforce has on the NPV for each option (shown in table 11). For this analysis, the central assumptions for each option are made (on MSL, strike turnout (76.1%) and number of hours in which national strike action took place per year (98.6 hours)) are applied. The central cost estimate is used in this example, and it is assumed that costs remain constant as set-up costs are assumed to be unaffected by the size of the workforce that are in scope of the legislation. The benefit of a firefighter hour is assumed to be equal, regardless of the rank of the firefighter/control staff.

Table 11: Sensitivity analysis – FRS workforce in scope, options 1, 2, 4 and 5, using central assumptions for costs and benefits, £million

FRS workforce in scope NPV – Option 1 NPV – Option 2 NPV – Option 4 NPV – Option 5
Watch manager and below (central estimate) £29.3 £117.5 £38.8 £37.0
Station manager and below £31.2 £124.5 £41.3 £39.3
Group manager and below £31.8 £127.2 £42.2 £40.1
Area manager and below £32.1 £128.2 £42.6 £40.5

NPV for ‘watch manager and below’ for each option matches central estimate NPVs.

125. Table 11 shows that when the in-scope FRS workforce increases, the NPV for each policy option increases, because they save more hours of strike action, thereby delivering more benefits to society.

H. Direct costs and benefits to business calculations

126. All the costs to businesses fall to trade unions in year 1. The total cost to businesses is set out in table 12 below. There are no costs of option 0 to businesses. All the costs to businesses are assumed to be minimal.

Table 12: Costs to business (rounded to nearest hundred) for options 1 to 5

(10 yr. PV) Low Central High
Options 1, 2 and 5      
Total set-up cost to business £2,900 £5,800 £11,600
BNPV -£2,900 -£5,800 -£11,600
EANDCB £300 £600 £1,200
Option 3 and 4      
Total set-up cost to business £18,200 £36,300 £57,400
BNPV -£18,200 -£36,300 -£57,400
EANDCB £1,800 £3,600 £5,700

I. Wider impacts

127. An equality impact assessment has been completed as part of the consultation planning process, considering potential impacts of MSL both on FRS workforce and on the public. The main conclusions from the EIA are as follows:

  • Working on the assumption that MSL would most likely apply to firefighters and control room staff as they are most likely to work in roles that directly address scenarios that pose an immediate risk to public safety, there were some impacts on FRS staff identified because of the demographic make-up of these staff groups. The fire and rescue workforce and pensions statistics covering the period from April 2021 to March 2022[footnote 47](England only; published October 2022) shows that firefighters and control room staff are likely to be younger than the staff for whom MSL would be less likely to apply, and that staff in-scope of MSL are more likely to be male than female. This could put FRS staff who are male and/or in younger age brackets at a disadvantage compared to staff who are female and/or younger, as they would be less able to take strike action as a result of the roles they most commonly fulfil. However, any disproportionate impact on these staff groups as a result of MSL is assessed to be justifiable on the basis that the MSL are necessary to protect public safety. Whilst it is possible that these measures could impact good relations and equality of opportunity between FRS staff with protected characteristics who are most likely to be affected by the proposed MSL and those staff who do not share these characteristics, any such impact is assessed to be justified for the same reasons.

  • MSL in FRS could have the benefit of offering groups that are more likely to be seriously harmed in a fire greater reassurance that the FRS will be available if they need it. These groups include older people and people with disabilities. This could therefore help reduce any discrimination suffered by these groups, and also advance equality of opportunity (for example to be protected from fire and other risks) between those groups and others who do not share these protected characteristics. Similarly, MSL may foster good relations between firefighters (mostly male as set out above) and these groups. While FRS have BCP in place for periods of strike action and no causal link can be made between previous strike action and deaths or serious harm occurring, the introduction of MSL could also have the effect of meaning that individuals more likely to be involved in non-fire-related incidents (such as road traffic accidents) greater certainty of being able to draw on the fire and rescue service if the need arises.

  • The EIA has been completed in relation to MSL policy in general terms and will need to be revisited after the consultation responses have been reviewed and a specific policy approach for delivery of MSL identified.

128. This consultation is part of a wider bill by DBT that covers various different industries, including health services; education; transport services; decommissioning of nuclear installations and management of radioactive waste and spent fuel and border security.[footnote 48] The total societal impact of this legislation should therefore be considered alongside the impact of accompanying primary and secondary legislation in other government departments. This consultation is specifically related to FRS, but MSL are expected to apply in other sectors.

129. This IA has not fully considered the impact of MSL on co-responding incidents (where FRS support policy and health services, and vice versa), and the impact of cross-border support for fire and rescue response. This could be further considered in future analysis.

J. Trade impact

130. There are no expected trade impacts from this policy.

K. Monitoring and evaluation plan

131. It is expected that the measures set out in this IA will require secondary legislation. The consultation closes in May 2023, and any secondary legislation will follow this, subject to the parliamentary timetable.

132.Success will be measured against the policy and strategic objectives; however, any evaluation is likely to be complex due to difficulties in establishing a robust counterfactual. Monitoring of this policy will likely include considering the level of strike action in FRS and seeing if there is any impact on cost savings, additional hours available for, and collective bargaining power. Further consideration on monitoring and evaluation of any proposed legislation will be set out in the final IA, alongside information on enforcement.

133. It is likely that this policy will be evaluated in February 2028, however this will be confirmed in the final IA.

L. Annexes

Impact assessment checklist

Mandatory specific impact test - statutory equalities duties

Statutory equalities duties

Home Office officials are actively considering the impact of MSL and how it might or will affect people with protected characteristics. This is an ongoing process. Policy officials will regularly review the equality impact assessment (EIA) as the policy options develop.

An EIA has been completed as part of the consultation planning process. A summary of its findings is that the policy is likely to impact younger, and male, individuals disproportionately because these individuals are more likely to be firefighter or control staff, who are in-scope of MSL. However, any disproportionate impact on these staff groups as a result of MSL is assessed to be justifiable on the basis that the MSL are necessary to protect public safety. It also possible that MSL could have the benefit of offering groups that are more likely to be seriously harmed in a fire greater reassurance that the FRS will be available if they need it. These groups include older people and people with disabilities. This could therefore help reduce any discrimination suffered by these groups, and also advance equality of opportunity (for example to be protected from fire and other risks) between those groups and others who do not share these protected characteristics.

The SRO has agreed these findings.

Complete: yes.

Economic impact tests

Small and micro-business assessment (SaMBA)

A SaMBA has been carried out in section E: yes.

  1. Government sets out new plan to protect leaseholders and make industry pay for the cladding crisis

  2. Minimum service levels for fire and rescue services

  3. Home Office: ad-hoc fire and rescue statistics: Resourcing by incident type, England, 2021 to 2023, table A32. 

  4. Fire statistics data tables: FIRE1403, year ending March 2021. Figures above 100% assumed to be 100%. 

  5. Calculated as 22% + 25%, 52% + 25% and 65% + 25%. 

  6. (0.25 * (358/365)) + (1 * (7/365)) = 0.26 

  7. (0.375 * (345/365)) + (1 *(20/365)) = 0.41 

  8. See FIRE0801 at Fire statistics data tables

  9. (0.5 * (3/4)) + (1 * (1/4)) = 0.63 

  10. The Green Book (2022)

  11. Fire and rescue workforce and pensions statistics: England, April 2021 to March 2022. Headcount numbers by rank are not published for firefighters for England. Therefore, it is assumed that, for Brigade Manager and Area Manager, the FTE workforce numbers are equal to the headcount numbers, and for the ranks of Group Manager and below for on-call and wholetime firefighters, FTE staff numbers are uprated by an up-rate factor. The up-rate factor for whole-time firefighters is equal to total wholetime headcount divided by total FTE wholetime staff numbers (22,781/22,642 = 1.006). The up-rate factor for on-call firefighters is equal to total on-call headcount divided by total FTE on-call staff numbers (12,179/8,422 = 1.446). 

  12. Stats Wales, Fire and rescue service personnel. FTE data for Wales is not published, so it is assumed that headcount numbers are the same as FTE numbers for Wales. Headcount numbers are not broken down by rank. Therefore, the proportion of staff, by headcount, in each of the ranks is assumed to be the same as the proportion of the FTE staff in each of the ranks in England. 

  13. Fire Safety and Organisational Statistics 2021-22

  14. Fire and rescue workforce and pensions statistics: England, April 2021 to March 2022 

  15. (42 hours worked) / (168 hours in a week) = 25%. 

  16. Firefighters deliver decisive mandate for industrial action’. 

  17. 73% * 88% = 64.2%. 

  18. ((295.5*1)) / 10 = 29.6, ((295.5 * 5)) / 10 = 147.8, ((295.5 * 3)) / 10 = 88.7. 

  19. The Economic and Social Value of UK Fire and Rescue Services’. 

  20. Pay settlement 2021’. 

  21. Impact Assessment (accessible)

  22. Method aligns with DBT MSL IA: Strikes (Minimum Services Levels) Bill 2023 2

  23. Sourced from DBT MSL IA: Strikes (Minimum Services Levels) Bill 2023

  24. Sourced from DBT MSL IA: Strikes (Minimum Services Levels) Bill 2023

  25. In the Welsh data, only total headcount data is given for wholetime firefighters, on-call firefighters and control staff, and this headcount data is not broken down by rank. Therefore, it is assumed that the percentage of the headcount labour force in each of the ranks in Wales is the same as the percentage of the FTE labour force in each of the ranks in England. 

  26. For England, the headcount labour force data is not broken down by rank for wholetime and on-call firefighters. Therefore, the FTE numbers for each rank are up-rated by an uprate factor which is equal to the total headcount divided by total FTE workforce. For wholetime firefighters this uprate factor is 1.006. For on-call firefighters this uprate factor is 1.446. 

  27. Where headcount staff data does not exist for senior FRS management, it is assumed that FTE staff numbers are equal to staff headcount numbers. 

  28. TUC “this Bill will prolong disputes and poison industrial relations – leading to more strikes”: Strikes (Minimum Services Levels) Bill 2023

  29. RMT unions might have to resort to novel methods such as extensive overtime bans and work to rule. 

  30. The Economic and Social Value of UK Fire and Rescue Services’. This report only covers English FRS at this stage, so all the figures from it are England only. 

  31. Non-fire incidents include road traffic collisions which may still be attended during strike action. It is not possible to disentangle these from other non-fire incidents which would not be responded to. 

  32. The 2022 figure DfT figure for a value of a fatality is £2.5 million, and so an uprate factor of 15% is found by doing 2.5/2.1 = 1.15, TAG data book, A4.1.1. 

  33. Assuming the 2-2-4 (2 day shift, 2 night shift, 4 days off) shift system, each firefighter works 42 hours a week, so over 3 years this would be 6,570 (when multiplied by the 1095 days that are in the 3 years of 2016 to 2017 to 2018 to 2019). 

  34. FIRE1102, Fire statistics data tables

  35. The Economic and Social Value of UK Fire and Rescue Services (PDF, 1.41MB): page 44. 

  36. Research: Cost Benefit Analysis - Greater Manchester Combined Authority

  37. The Economic and Social Value of UK Fire and Rescue Services (PDF, 1.41MB): page 44. 

  38. 10.6% uplift GDP deflators at market prices, and money GDP

  39. Impact Assessment: Assumption 1.8, page 133. 

  40. See assumption 1.4. 

  41. Exact figure £44.27. 

  42. English Housing Survey, 2020 to 2021: feeling safe from fire’. 

  43. Calculations completed based on an average yearly MSL, however, in reality, option 5, may still incur some monetised benefits if strike action was prevented on certain days where there is an MSL above expected strike turnout. This detail will be further assessed in the final IA. 

  44. Trade unions: the current list and schedule’. 

  45. As the strike turnout approaches 1 minus MSL, benefits will reduce because there will be fewer, and then no, hours saved compared to the counterfactual scenario (option 0). This is because, say an MSL was set at 60%, this means that 40% of the workforce would need to be available during strike action. If the turnout of strike action was only 50%, then 50% of the workforce would be available during strike action. In this scenario, there are no saved hours from MSL, because all individuals who are subject to MSL, would not have been striking anyway. 

  46. Fire and rescue workforce and pensions statistics: England, April 2021 to March 2022

  47. Sourced from DBT MSL IA: Strikes (Minimum Services Levels) Bill 2023