Resource to Risk:
Disaster in Data Driven
Governance & Leadership
According to Warwickshire Fire and Rescue Service
A £2 million ‘investment’ from WCC reserves will fund:
An emergency coverage ‘Postcode Lottery‘, where somehow the most deprived, rural AND densely populated communities ALL receive a reduction in current levels of emergency cover.
120 On-call firefighters facing redundancy
At least 40 Wholetime firefighters facing pay cuts
3 Fire stations being turned into uncrewed garages with a 2-hour emergency response time
Foreword from the FRSA Warwicksire Team
Resource to Risk (R2R) has been a contentious and difficult period for many across WFRS as far as our FRSA Warwickshire team have witnessed alongside our members and peers.
Those whose livelihood and passion faced the axe – as the result of an under-evidenced and under-detailed set of proposals – have been delivered this information without significant consideration for the uncertainty and discomfort that this injected into their lives. Oftentimes, this lack of information being excused by process and procedure. As a team we are putting forward the case that the rationale for these proposals is not founded in a basis of data or the experience that other firefighters (and more broadly all staff) have in emergency response.
Here we will evidence in fact, Resource to Risk is a ‘data-driven approach’ that is being used as political cover for an insufficiently evidenced approach.
Whilst there is merit to be had in data-driven resource alignment – within sensible limitations (like any tool) – data can be used to compress and represent large amounts of information to make the correct decisions. It cannot be stressed enough however, that when applying a data-based approach to decisions that affect the health and wellbeing of the Warwickshire (or any) community, it is essential that the methodology is well-founded and undermined by neither political currents nor ambition.
In other sectors (food, medicine, healthcare), where the use of information and data can affect the health of the community and the individual, appropriate standards are in place. Independent oversight and scrutiny is non-negotiable, to govern safe data-informed decisions.
The fire service has never had sufficient data scrutiny to the standard that Resource to Risk demands, and it has never been expected to.
As far as we are aware, there has never been an attempt to drive a fire and rescue service’s future fire cover with data analysis and modelling as the core foundation; the innovation we have seen ‘Resource to Risk’ sold as.
Therefore, it is important that when we take these decisions, we enable and encourage the same level of scrutiny that we apply to all sectors that effect our health and wellbeing as a community. The ‘Resource to Risk’ consultation does not demonstrate a suitable level of scrutiny, given how dependent the distribution of life-saving resources is on data-driven decision-making.
Not only is the community of Warwickshire at risk, but we are risking permanently losing significant numbers of firefighters whom have continued to serve Warwickshire, and their local communities, despite facing redundancy, pay cuts and poor management.
The information you see below will attest to the narrative of Resource to Risk, from our as both staff members and representatives.
As on-call firefighters, we lent our primary employment expertise to this process, with the cumulative experience of our team in science, economics, business and fire, informing our holistic criticism of the Resource to Risk proposals.
There is a lot of information below; we have tried to keep it punchy and to the point. We aimed to deal with this issue internally through professional channels, when we tried earnestly to consult the different decision-makers and policyholders throughout WFRS and WCC.
Sadly, we believe they did not heed our warning seriously.
So now we will present our case to see what others think of the Resource to Risk process, and how it presents a dangerous precedent for the use of data and statistics in life-saving operations in our community. We are placing the burden of evidence on the fire service and fire authority,
in the open, with the aim to action change and provide an honest accounting of the impact these proposals will have on our community and our firefighters.
We will provide:
1. A timeline of the Resource to Risk public consultation.
2. Address the accuracy and completeness of the information delivered to the public, and firefighters by representatives of WFRS and distributed by WCC.
3. Address the flawed rationale behind the development of both Model 2A and Model A and their impact on Warwickshire Communities.
4. Address the governance of the Resource to Risk public consultation and how scrutiny of the proposals was quashed during WCC overview and scrutiny committee meetings and public input to Model A changes were prevented.
5. Evidence the need for a further public consultation.
This web page will therefore serve as a platform for R2R information. As we continue to peel away the layers on the undemocratically consulted Resource to Risk Model A, we will publish our findings here.
Take care and stay safe,
FRSA Warwickshire Team.
Resource to Risk : The Timeline Summary
Continue scrolling to get further information on the public consultation process.
September 2023:
WFRS - A service in dire straits.
WFRS had an inspection by His Majesty’s Inspectorate of Constabulary and Fire & Rescue Services (HMICFRS).
The results were concerning.
From the BBC news article report:
“She (Inspector Wendy Williams) added in her report she was ‘disappointed’ some high-risk residential buildings had not been inspected, despite being highlighted as a cause for concern in a previous inspection during 2020/21. The service still needs to put in place a suitable risk management system, she added.”
From the corresponding report:
“We found evidence that it hadn’t inspected all its high-risk residential buildings, including some high-rise buildings.”
In response to the report, CFO Ben Brook had this to say:
“We have a robust plan of action in place which is already making significant inroads into improving the services we provide.”
Portfolio Holder Andrew Crump said he was confident that the service has
“The right plans in place going forwards”
Continue to see how these plans developed.
The summary judgement of the inspector of WFRS’ performance in 2023
October 2023:
WFRS Delivers Model 2A to staff
WFRS officers (middle-management) are sent to stations across the county to deliver a presentation to staff members on the ‘Resource to Risk’ concept and resulting changes to service delivery and working patterns.
Heavy stress is placed on the innovative nature of the work and the fact that an external contractor employed data modelling and statistical methods to create the service delivery proposal Model 2A.
The Officers’ delivery is received poorly, in part because they appear to lack the sufficient detail of how, practically, Model 2A could possibly be implemented, as it failed to hold up
common-sense scrutiny from experienced operational firefighters.
Further frustration is found in the evidencing for the wholesale removal of the on-call stations. With the evidence provided by the fire service being significantly lacking or incomplete, it does not match up to the on the ground experience of on-call staff, whom have largely felt neglected by a service run by those who do not understand them.
We have developed a large base of evidence for Model 2A’s lacking rationale and fundamentally flawed risk analysis.
November 2023:
We Accessed WFRS' Data and Methodology for Model 2A
With service representatives scant on information about how we could predict future response times based on risk, the foundational assumption of Model 2A, we sought the information ourselves.
We issued freedom of information (FOI) requests to Warwickshire County Council to gain access to the data used to generate Model 2A; how the work was conducted between WFRS and the external contractor and how the Community Risk Management Plan (CRMP) created by WFRS factored into Model 2A.
The FOI Request may be found here
December 2023:
Public Consultation Announced, with immediate misinformation.
Warwickshire County Council announce the ‘Resource to Risk’ public consultation and released Model 2A for public scrutiny.
Public consultations would start in earnest in January.
Originally, WFRS and portfolio holder opted for a public consultation that was only an online form, if not for intervention of the WCC Overview and Scrutiny Committee.
A ‘Let’s talk Warwickshire’ WCC podcast was released, with CFO Ben Brook and ACFO Sally Waldron discussing the new service proposal ‘Model 2A’.
The ACFO discusses the innovative nature of the work, that in fact it was not the service at all generating the delivery models, but an external modelling company.
“We’ve worked with a (data modelling) company called ORH who have done a lot of modelling for emergency services…what they do is look at historical demand levels, demographics and other foreseeable future risks and they use an algorithm to produce a set of delivery models. That can better match resource to where the demand is. So it’s not us sitting here like ‘right we’ve got a blank canvas and this appliance goes here and that appliance goes there’, it’s not us at all…”
– ACFO Sally Waldron, Let’s talk Warwickshire Podcast, December 2023.
Listen to the above in full:
In fact, contrary to the above, choosing where the fire cover was placed was EXACTLY what WFRS did.
As an excerpt from FOI 9653920 proves:
“The different delivery option models were generated internally using Service expertise and experience using the data provided. These models were then provided along with the data to a third-party data modelling consultancy, ORH, who analysed the performance of the models against the data.”
– WCC FOI Request 9653920, December 2023
Meanwhile, in a BBC News article CFO Ben Brook mistakenly reports on-call day time availability at 17%.
(Warwickshire fire service’s staffing shake-up necessary – fire chief – BBC News)
That month (December 2023), by their own metric, on-call availability was at 46.7%, with a 29.5% availability to emergency P1 – P2 incidents.
In fact HMICFRS reported on-call availability was reported as 67%.
We will provide evidence for why the on-call availability may be low and why the rationale provided by WFRS is totally inadequate, in our breakdown of Model 2A.
We have found NO evidence internally or externally for a 17% availability figure being representative for WFRS’ on-call availability, we would place the burden of evidence on WFRS to back this figure and how it is representative of on-call availability.
Meanwhile at WCC Fire Overview and Scrutiny Committee...
CFO Ben Brook correctly highlights that when using the WFRS’ own incident data it is important to align resources to incident frequency amongst other variables.
“CFO Ben Brook said the rationale behind the proposed reconfiguration was so greater resources could be in place where they were more likely to be used. He said there were, on average, 85 incidents in Nuneaton each year where life and/or property was in danger, compared to areas like Bidford or Fenny Compton where the number of incidents was in single figures.”
In June 2024, WFRS gained £2 million extra budget AND still
HALVED the emergency cover in Nuneaton, a town of 90,000 to ONLY ONE fire appliance.
REMOVED emergency cover from Fenny Compton.
HALVED Bidford emergency cover.
No post-consultation Fire OSC session was planned beforehand.
“There is no new money, I think that’s really important to say.”
– CFO Ben Brook, December 2023
Snippet from WCC Fire Overview and Scrutiny committee, December 2023: https://warwickshire.public-i.tv/core/portal/webcast_interactive/818133
January 2024:
Serious Investigation Begins
Finding contradictory information between our freedom of information request*, and what we had been informed as staff members, we decided to investigate further.
In fact, the external contractor, ORH was not involved in creating the service delivery model, just in benchmarking it, as discovered in December. This means that ORH tested, but did not create the models that WFRS delivered.
The delivery option models were generated by WFRS and no algorithm was involved.
This tracks with the cost of the work funded to the external contractor, which was just under £25,000, which is WCC’s threshold for competitive procurement. This type of budget could not feasibly be used to totally regenerate the response models from scratch, with Warwickshire-specified risk algorithms and data collection.
Instead, risk information came from the Community Risk Management Plan (CRMP), an integrative risk map created by WFRS.
We found WFRS CRMP risk map was completely lacking the robust data practice to warrant radical change in emergency response delivery. This is UNPRECEDENTED.
IF THIS APPROACH TO DATA-INFORMED RESPONSE CONFIGURATIONS SETS PRECEDENT LIVES WILL BE LOST.
It is quite the undertaking scientifically to predict the future risk, to a level where we can optimise for lean, efficient risk-based fire cover. The underlying risk profile must be extremely well founded in complete and representative data, which the CRMP certainly was not.
We sought further expert help, through our professional networks, we gained the assistance of two professors in Applied Statistics and Risk, from 2 top-10 UK universities, whom lent their advise and expertise to our investigation.
At this time, we were unsure as to whether WFRS/WCC representatives were aware of these fundamental flaws assigning resources to risk so we begun preparing a presentation for the upcoming ‘Dragon’s Den’ session, arranged for staff to feedback on Model 2A to WFRS Fire Leadership Team.
*The FOI Request may be found here.
Febuary 2024:
The Dragon's Den Sessions
To gather feedback on proposed Model 2A from staff, WFRS’ Fire Leadership Team (FLT) organised staff input sessions which they titled ‘Dragon’s Den sessions’.
With our slot booked we attended and delivered the presentation collated in January.
We delivered our frank and professional criticism of the Model 2A based on the research into it. We broke the presentation into 3 sections:
- Background on why we use Data modelling
We introduce, from our own and from our advisory team’s experience, how risk modelling is usually undertaken and how the approach to working with data in Resource to Risk did not conform to any known standard.
- Concerns with Current Strategy
Where we evidence the danger and cost-inefficiency of Model 2A, the lack of proper support of on-call recruitment and retention. Importantly we warn the potential of a emergency response system collapse through lack of resilient fire cover.
- Route forward
In the hope of informing the correct approach, we advised the FLT on approaches to fund and deliver risk analysis and future risk prediction to actually achieve ‘Resource to Risk’ through different grant-funded (not funded from WFRS’ budget) avenues from our own professional experience of acquiring funding for, and delivering scientific innovation.
A follow up meeting with the WFRS’ CRMP manager was arranged by FLT,
there was no follow up action afterwards.
Other staff members notably came forward with their own proposals based on years of experience and their own professional skills from within WFRS and other primary employment.
Click above to view our technical scrutiny of the data-based ‘Resource to Risk’ proposal.
Delivered to WFRS FLT, WCC Fire Portfolio Holder and WCC Executive.
March 2024:
Public Consultation Ends.
Alternatives Sought.
Misinformation Continues.
Overall the public consultation and the proposal received overwhelmingly negative responses (see here)
The reception of the consultation from our observed meetings was contentious, very little positive feedback was to be found from either public or staff and information on the validity of the proposals could not be held up to proper scrutiny.
Without a clear mandate from public or policymakers WFRS seek alternative proposals to Model 2A.
In the meantime, we delivered the same warning of the danger these proposals to the Fire Portfolio Holder as we did to WFRS’ FLT a month prior.
We believe the consultation on Model 2A breached Gunning Principles 2 and 4, the legal principles for public consultation.
April 2024: Alternatives Considered.
With just under 2 weeks notice the FRSA and Fire Officer’s Association (FOA) came together to deliver a joint alternative proposal, based on an incentivised on-call system that drives for improved availability to P1-P2 incidents based on an incentivised retaining fee structure.
WFRS accepted numerous proposals from across different communities in Warwickshire.
Click here to see the FRSA / FOA incentivised on-call model
This month, an 8-pump fire occured in Stratford overnight
Under Model 2A the working proposal at the time,
this would leave 0 fire appliances available to rest of Warwickshire overnight.
This is evidence of an incident, that under Model 2A, would have been an emergency response system collapse we warned WFRS FLT about in February.
(see technical scrutiny presentation – slide 14)
May 2024:
Council Chief Executive Warned.
After informing WFRS’ Leadership team we also presented our warning presentation to the WCC Chief Executive.
June 2024:
Model A Delivered. North Warwickshire Pays.
After Model 2A’s negative reception and feedback,WFRS were made to change their proposal.
Model A was delivered. After warning WFRS, WCC Portfolio Holder & Chief Executive of the danger of the foundational risk analysis driving Resource to Risk; The risk analysis was removed from Model A.
The result is a kneejerk reaction to the responses we witnessed in South Warwickshire, where responses were very negative.
We see a total reduction in Northern Warwickshire fire cover to replace concerns highlighted in the South of Warwickshire.
For example, focusing on the North of Warwickshire, where the resilience and appliance availability is reduced we see:
A reduction from what we have currently 5.2 appliances (with a best-case scenario of 7 appliances) to a best-case scenario of just 4 life-saving fire appliances across Northern Warwickshire.
Table 1. Displaying the reduction in available fire appliances and maximum potential fire cover between Model A and Current appliance arrangements. Potential cover assumes 100% availability in on-call and whole time fire appliances. Actual cover takes the on-call availability to P1-P2 incidents as reported in May 2024.
WFRS is introducing a 'Postcode Lottery' for Fire Cover
The argument that the moving of one wholetime appliance to Bedworth from Nuneaton does not stand up to basic scrutiny.
Using WFRS' own data, we have looked at just how the interdependence between Nuneaton, Bedworth and Atherstone might work, considering Nuneaton now has one fire appliance.
We ran a simultaneous activity analysis using WFRS’ own R2R modelling data of 2018 – 2022. We took all of the incidents that occurred at the same time between Nuneaton, Bedworth and Atherstone stations, at different time intervals.
It is essential that policymakers understand that emergency fire cover MUST be considered from a regional perspective as opposed to along local geopolitical boundaries, as we will evidence.
We found 156 occasions (Table A) where Atherstone night-time on-call, that is now being removed under Model A, attended incidents within a 3-hour period of Nuneaton. We found 771 occasions where Nuneaton was called out at the same time as Bedworth (Table B).
Under Model A this would be 771 emergency incidents where the lack of fire cover in Nuneaton and Bedworth would mean a reduction in fire cover from surrounding towns.
Atherstone are also facing a critical reduction in fire cover, with one of the best performing on-call sections regularly achieving 80~90% availability, they are a shining example of the on-call system working in Warwickshire and the community is anticipated to lose their night-time fire cover. In Table A, there have been 39 occasions under model A where Atherstone night time on-call responded to an incident within 30 minutes or less of Nuneaton.
That’s at least 39 emergency incidents, under Model A, where Atherstone is relying on Polesworth, Bedworth or even Rugby for emergency response.
Bedworth have been left without a Watch Manager for 4 years, and therefore have a lower current availability of 40%, Polesworth perform well at 60% availability. If managed well, the potential for the on-call system for cost-effective, resilient fire cover is huge.
Meanwhile at WCC Fire Overview and Scrutiny Committee...
The fire portfolio holder presented Model A as radically different from 2A, which was consulted on.
“As you can see, this proposal (Model A) is radically different”
– Fire Portfolio Holder, Cllr Andy Crump, June 2024.
“there was no need for the revised model to go back out for consultation”
– CFO Ben Brook, June 2024.
Overview and Scrutiny Committee (OSC) Cllrs then proceed to query
the CFO for the following information:
– How the average response times are calculated?
See below as to why this is important.
– Where are 2nd fire appliance response times?
2nd fire appliance response are essential considerations for firefighter safety.
– Why the heat map (i.e. the RISK analysis) is entirely missing?
The entire foundation for the proposal is a resource to risk process.
In response to these questions by Overview & Scrutiny Committee Councillors, Fire Portfolio holder, Cllr Andy Crump had this to say:
“The Chief … Fire Officer even, has asked for a certain level of resources and we are providing him with those. It is up to him how he employs them. It is not for councillors to impinge on his operational independence and if we percieved to be doing this, I’m sure HMICFRS and the Home office would have something serious to say about this. Ben is the expert.”
– Fire Portfolio Holder, Cllr Andy Crump, June 2024.
When reminded of the importance of the committee for scrutiny, the Portfolio Holder had this to say:
“As long as their evidence based, rather than anecdotal”
Snippet from WCC Fire Overview and Scrutiny committee, December 2023: https://warwickshire.public-i.tv/core/portal/webcast_interactive/818133
Why Data Matters
The fire service prioritises incident types into 5 categories of priority (P1-5).
The highest of which (P1-P2) correspond to emergency threat to life and property. These make up just 11% of Warwickshire’s total incidents (ALL of WFRS modelling is based on just these 11% of incidents).
Both models report the attendance times to incidents using an average.
The problem is, when relying on data for the basis of an unprecedented proposal,
an average simply does not cut it.
Resilience is how well a system copes with extremes. The average is a poor measure of resilience, especially for highest priority incident response times, as both resilient and non-resilient systems can still have the same average.
As you can see in the example above, whilst the average response time is 10 minutes for both, a non-resilient system (left) would have a wide spread of response times, at best 5 minutes and at worst frequently approaching 20 – 35 minutes. However, a resilient system (right) would have a narrow spread of response times, at best 5 minutes but at worst 15.
This matters!
Gaps in fire appliance coverage create a non-resilient system, as response times across a sparsely served county could vary significantly based on the strain placed on the service at any given time and location(s).
Both Model A and Model 2A proposals would likely fall under a non-resilient system, despite the confidence in the use of an average by WFRS. No evidence has been provided to show otherwise.
This means that as demand is placed on the service, P1-P2 response capacity across the county could collapse under the disparate cover provided by model 2A or A, with a 2-hour wait until resilience crews were available.
” It is not for councillors to impinge on his operational independence and if we percieved to be doing this, I’m sure HMICFRS and the Home office would have something serious to say about this.”
– Fire Portfolio Holder Cllr Andy Crump, in response to Overview and Scrutiny Councillors requesting the kind of detail we provide on the left.