In a first-past-the-post system, redrawing constituencies is essential. Had the UK, in the 2019 election, held on to the same boundaries used in 1945, the voters of Milton Keynes and the northernmost half of Buckinghamshire, for example, would have found themselves represented by a solitary MP, as opposed to the three standing today.
Redrawing boundaries to better reflect the growing conurbations and concentrations of the nation’s population can theoretically lead to a more representative parliament at Westminster. But it doesn’t always work that way.
As the current review into constituency boundaries gets ready to publish its proposals this summer, with some concerns that the changes will notionally favour the Conservatives, it is worth examining how the UK system works – and what could be done to improve it.
The current rules stipulate that constituencies must “have no less than 69,724 parliamentary electors and no more than 77,062”, a range of just over 7,000 – or 5 per cent either way.
The UK constructs its constituencies by combining wards – electoral boundaries for local councillors. These wards often differ in size, and undergo their own regular redrawing. In practice, they represent a single community, or half of one, or the edge of one plus the nearest next village, or – well, you get the picture.
In short, wards are not ideal building blocks, and merging disparate communities together can seem nonsensical. The size difference of these building blocks also complicates matters. In England, you can come across wards with as many as 26,000 electors, and others with as few as 1,000. Wards with smaller populations are, obviously, more useful in terms of carving out a decently drawn parliamentary seat so as to comply with the 69,724-77,062 rule, but they tend to be found in countryside constituencies. Wards with larger populations, meanwhile, are often those in cities or towns, and make for increasingly unwieldy building blocks.
Andrew Teale, editor of the Local Election Archive Project and previewer for Britain Elects, observes of the problems presented by large wards:
“In the metropolitan borough of Wigan, the current Makerfield (74,400) and Wigan (75,607) seats are within the required tolerance, but the Leigh constituency (77,416) is 354 voters over the upper limit. This can be fixed, but the least disruptive fix involves transferring the two wards covering Golborne and Lowton (19,091 electors) from the Leigh seat into the Makerfield seat while Hindley and Hindley Green (18,422 electors) go the other way; that’s a minimum of 37,513 electors moved between constituencies to fix one seat being 354 voters too large.”
When you are a local area represented by large wards with more than, say, 18,000 electors, the room for geographical manoeuvre is limited, if it exists at all. When the acceptable range is a touch over 7,000 electors (the 5 per cent rule), a constituency that resembles a community may find itself sacrificed or drastically altered to create a seat that sticks to the rules.
The seats of Chester and Ellesmere Port and Neston are, in their current form, well within the required range. On paper, their MPs should feel comfortable knowing that their seats do not need altering. But, in reality, they may suffer quite significant redrawing, because neighbouring seats on the Wirral peninsula need to be altered to fulfil the necessary quota.
In practice, what this may lead to (and has led to in the past) is the creation of some bonkers (a technical term) constituencies.
This was the case in the boundary proposals of 2011, in which there was one particularly problematic example: Mersey Banks.
Mersey Banks was a constituency that would have covered Bromborough on the Wirral, half of Ellesmere Port in Cheshire, the suburbs of Northwich and, across the river Mersey, the outskirts of Widnes and the village of Hale. This was despite the fact there was no river crossing in the seat.
The constituency, as Teale notes, was “the worst proposal in a car-crash map for north-west England”. He described it as a “Frankenstein’s monster of a constituency”, one that was, in the end, thrown out after a public consultation.
Mersey Banks is by no means the only example of a monstrous constituency proposed in order to satisfy the 5 per cent rule. In 2006 an attempt to create a seat covering Wallasey on the Wirral and a few neighbourhoods in northern Liverpool (again, across the Mersey) was also scrapped following public outcry.
These examples expose the highly restrictive rules that the independent boundary commissions for England, Scotland, Wales and Northern Ireland must prioritise, sometimes to disastrous effect: the need for constituencies to be equalised in terms of population often turns out to be incompatible with the need for constituencies to fit with physical and social geography, the priorities of the people who actually live there. A seat covering two sides of the River Mersey, with nothing in common except the river itself, is not a natural fit – but it did fill the quota, so it was drawn up as a viable proposal.
Put simply, the quota system we have for the current review is far too narrow to achieve decently drawn seats. It’s like trying to recreate Big Ben using a limited number of Lego Duplo bricks.
As Teale says: “You don’t have to be a mathematician to realise that this is going to run into trouble.”
A better practice for the boundary review might be to employ the mathematical requirements as a guideline, rather than a cast-iron rule. Some numerical inequalities would remain (as they already do and will do for the Western Isles, the Isle of Wight, the Orkney and Shetland isles, and Anglesey), but there would be less risk of communities being split apart, or those with nothing in common herded together. In turn, this may prevent MPs having to struggle to work out who, and where, they are supposed to represent.
It may also be worth considering an appropriate alternative to wards as the building blocks for seats. Statistical bodies may need to think about more local structures and boundaries that define our communities, so that constituencies such as Blackburn do, in the end, actually cover the whole of Blackburn, not just 90-odd per cent of it.
Do the boundaries of a constituency really make much difference to the democratic process? The answer, it seems, is yes. A 2016 study of the Australian legislative system found that unwieldy and diverse Senate seats often meant senators were less willing to engage with and act on the behalf of their constituents, since diffuse constituencies tend to contain communities with opposing interests.
The debate over boundary changes is often seen as a classic “political anorak” issue – all maps and spreadsheets and obsessions over names that no longer have much resonance (looking at you, Elmet). But it is, in fact, about the right of voters not just to fair representation, but to an active MP who stands for somewhere they broadly recognise as their community. That isn’t an anorak issue; it is fundamental to ensuring the people we need to be engaged with the political system in fact are.
For anyone tempted to try redrawing the UK’s parliamentary constituencies themselves, I recommend Kevin Larkin’s excellent tool, Boundary Assistant.