04/09/2026
An Insight into local government reorganisation in Warwickshire from Cllr Dr Andrew Glowacki - Cawston & Admirals, Rugby
68% OF WARWICKSHIRE SUPPORTS TWO UNITARY COUNCILS!
Well...
Not quite.
In my profession I'm exposed to a vast number of studies, publications, statistics, analyses and all sorts of weird and wonderful interpretations of them. Some are perfectly legitimate. Some are... creative.
Reading statistics properly is surprisingly difficult. It's something you learn over years.
My favourite illustration is this:
A man takes his dog for a walk. Statistically, between them they have an average of three legs each.
Mathematically correct.
Utterly useless.
Welcome to statistics.
Statistics, however, is enormously valuable. In medicine it helps us understand whether treatments works, which drugs are appropriate and what actually improves outcomes.
And, perhaps surprisingly, statistics matter enormously in local government too.
Which brings me neatly to Local Government Reorganisation — LGR — and the proposal to replace Warwickshire's existing councils with two new unitary authorities, North and South.
Among the papers surrounding our Extraordinary Council Meeting (the one that's been cancelled twice now) appeared a rather impressive figure:
68% positive for support of LGR.
That sounds rather like Warwickshire has spoken.
So I went digging.
And, regrettably for my remaining free time, I kept digging.
FIRST PLOT TWIST: THE GOVERNMENT ITSELF TELLS US WHAT THIS EXERCISE WAS
The official Government consultation page says:
“Although this will be a technical exercise, focused on the criteria, rather than a public consultation...”
It then says residents, parish councils, businesses and community organisations were nevertheless welcome to participate.
So this wasn't designed as a referendum or representative opinion poll asking Warwickshire:
“Do you actually want this?”
It was principally a technical assessment of two proposals against Government criteria.
Fair enough.
But that's rather important when interpreting what came next.
SECOND PLOT TWIST: 14,099 RESPONSES!
Sounds enormous.
Except there weren't 14,099 people.
There were 2,120 respondents in total, of whom 1,937 were Warwickshire residents.
So where did 14,099 come from?
This is where the statistical sausage comes crashing into the room.
Respondents were asked seven scored questions about the two-unitary proposal. Roughly 2,000 answers were recorded for each question.
Add those seven sets of answers together and...
Magic! 14,099!
We haven't discovered another 12,000 Warwickshire residents hiding behind the sofa.
We've counted many of the same respondents seven times, once for each question they answered.
THIRD PLOT TWIST: WHAT DOES 68% ACTUALLY MEAN?
This is the really important bit.
The questions included things such as whether the proposed councils were based on sensible geography, could deliver their proposed outcomes, were financially resilient, could deliver sustainable services, reflected local needs, supported devolution and enabled community engagement.
What apparently wasn't asked was the wonderfully simple question many residents might reasonably expect:
“Which do you actually prefer: one Warwickshire council or two?”
Instead, answers to those criteria were scored.
“Strongly agree” and “somewhat agree” counted positively.
Those individual answers were then aggregated.
And that's where the 68% positive figure comes from.
There's a rather substantial difference between:
“68% of scored answers about the two-unitary proposal were positive”
AND
“68% of Warwickshire residents support splitting Warwickshire into two councils.”
Those sentences sound similar.
Statistically, they most certainly are not.
AND HOW REPRESENTATIVE WAS IT?
Warwickshire has more than 630,000 residents.
The consultation received responses from 1,937 Warwickshire residents.
That's roughly 0.3% of the population.
Now, before somebody throws a statistics textbook at me: a small sample can absolutely be representative. National opinion polls routinely use samples of around 1,000–2,000 people.
But that works because considerable effort goes into constructing and weighting samples so they resemble the population being measured.
Here, the published Government material describes an open consultation exercise.
I haven't found published methodology demonstrating that those 1,937 Warwickshire respondents constituted a demographically representative sample of Warwickshire's population.
So 0.3% isn't automatically the problem.
The problem is treating a self-selecting consultation as though it were a representative public opinion poll.
THEN WE HAVE THE ORGANISATIONS
There were also 72 responses from organisations, separately from 17 named consultees.
The published analysis we've examined doesn't give us a complete list identifying those 72 organisations.
Were they businesses? Charities? Community organisations? NGOs? Parish councils? Campaigning organisations?
Perhaps a mixture of everything.
We simply don't know from the published summary.
I'm perfectly happy to be corrected on that point — preferably by publication of the information.
WHAT ABOUT THE WRITTEN ANSWERS?
Another 40 responses arrived by email or post, including organisations, named consultees and individuals.
They weren't ignored - we're told.
Where somebody explicitly answered using the questionnaire's scoring system, their answers could enter the numerical analysis. Where somebody instead wrote an argument explaining their position, it was considered qualitatively rather than officials attempting to magically convert prose into “somewhat agree”.
That's actually sensible.
But it reinforces the point that the headline percentage doesn't tell the whole story.
And Question 8 itself was explicitly a free-text question, despite the final analysis apparently describing respondents as being invited to answer eight multiple-choice questions.
The Government's original questionnaire clearly shows Questions 1–7 were scored and Question 8 was written comments.
A small drafting error perhaps.
But when we're discussing methodology, details rather matter.
AND NOW MY FAVOURITE PART...
When Warwickshire County Council advertised the exercise to residents, it said:
“The Government is only consulting on whether residents would prefer one unitary council for Warwickshire, or separate north and south unitary councils.”
Beautifully simple.
Except the questionnaire itself apparently never asked respondents that straightforward binary preference question.
Instead, it asked them to score both proposals against seven criteria.
And somewhere along the journey from questionnaire → spreadsheet → presentation...
68% positive scored answers starts sounding remarkably like 68% public support.
And that is precisely why statistics deserve scrutiny.
SO, DOES WARWICKSHIRE SUPPORT TWO UNITARY COUNCILS?
Maybe.
Maybe not.
This consultation doesn't tell us that.
What it tells us is that 2,120 people and organisations responded to a Government technical exercise, including 1,937 Warwickshire residents, and approximately 68% of the scored answers concerning the two-unitary proposal were positive.
That's useful information.
It just isn't the same information as:
“68% of Warwickshire supports LGR.”
And if we're going to undertake one of the biggest changes to local government in Warwickshire for generations, I think residents deserve something terribly unfashionable:
the numbers, what they actually measure, and enough information to make up their own minds.
After all...
A man and his dog may statistically have three legs each.
But I wouldn't recommend buying either of them trousers on that basis
