The writer is professor of politics at Queen Mary University of London
One of the six demands made by the Chartists in 1838 was that parliamentary constituencies should contain roughly equal numbers of people. It is a principle that nowadays seems difficult to disagree with. It is why regular constituency reviews — England’s Boundary Commission published its initial proposals this week, others for the rest of the UK are due later in the year — are an essential part of political life. They ensure that electoral boundaries reflect the reality of where voters live now, rather than where they used to live decades ago.
The current reviews are long overdue. An earlier attempt to update the boundaries began in 2011, before being derailed under the Conservative-Liberal Democrat coalition and then junked entirely in 2018. As a result, the 2019 election was contested on boundaries drawn up by a process that began in 2000. The next election should not be fought with seats based on population data that will by then be almost a quarter of a century old.
That is probably where the agreement ends, though. If it sounds relatively straightforward to draw up constituencies of equal size, it is not. For one thing, at the same time as we want seats to be roughly numerically equal, we also want them to represent real places. This tension, between what one book on the subject calls the organic and the mathematic, has long been there, although more recently the mathematic has been winning out. This is especially true this time because of the new requirement that seats can vary by no more than 5 per cent from the national average, with only a handful of exceptions allowed.
You see the consequences of this in the ever longer and uglier names of constituencies, as the Boundary Commissions struggle to patch together otherwise unrelated geographical areas. We no longer have constituencies called Rye or Eye, but we do have Cumbernauld, Kilsyth and Kirkintilloch East. If enacted, the proposed English constituencies would have the most cumbersome titles since the 1950 general election, the first contest when all members of the House of Commons were elected in single-member constituencies. The Scottish constituencies, still to come, are usually longer still.
But even if you have seats that are mathematically equal, it still matters where you draw the constituency boundaries. That’s why in Politics 101 at university you learn about Governor Elbridge Gerry and the highly partisan, salamander-shaped district he proposed in Boston in 1812 — the episode that gave rise to the phrase “gerrymandering” — and why by the third year you’re reading work on the decomposition of electoral bias (which shows that it can exist, even when there is no deliberate gerrymandering).
It’s one reason why most MPs hate the idea of boundary changes. Some find their constituencies have gone completely, leaving them scrabbling for new homes, not always successfully. Others lose areas of party strength or get lumbered with a ward full of their opponents, turning a previously safe seat worryingly marginal. Even those who nominally gain from the process often don’t like the uncertainty and the need to build up relationships with people who were previously not their constituents.
You will read countless claims that tell you confidently that party X would have won Y seats more or less, had the last election been fought under the new boundaries. But the last election wasn’t fought under those boundaries; it might have been fought differently if it had been. Plus, these are proposals — and past experience is that any initial plans often change, once they go out to consultation. And with the electoral tectonic plates shifting beneath us, it is even harder than normal to make estimates about what changes might mean. In other words, treat any claims with the caution you would normally apply to estimates of provisional proposals based on suboptimal data.
Yet overall there is little doubt that this set of boundary changes will work to the benefit of the Conservatives, probably by around 5 to 10 seats net. Or to reframe that sentence, the current outdated boundaries unfairly cost the Conservatives around 5 to 10 seats at the last election. Regardless of the precise detail, the broader changes — Scotland and Wales losing seats, England, especially the south, gaining them — have been set from the beginning of the review and have fairly predictable consequences. This isn’t gerrymandering; it’s just a reflection of where people live now. As a proportion of the UK’s electorate, the south-east has grown since 2000 and Scotland has declined.
In the grand scheme of things, even 10 seats is fairly small change. Relative to some other factors that determine election outcomes, the disparities in constituency size are, and always have been, relatively minor. But still, as the advert says, every little helps. Or not, if you’re the Labour party — whose path back to power just became that little bit steeper.