To make UK elections more representative, the main issue to fix is single-member winner-takes-all constituencies. Above the discussion about the exact mathematics of different voting systems, it is crucial to understand that the fragmentation of the electorate is the root cause of overall disproportionality.
As such, the establishment of multiple-member constituencies is the key to electoral reform. But, given that we are in the UK so used to having around 650 small constituencies, it is quite hard to imagine what that would look like.
The Electoral Reform Society advocates for electoral reform, and in particular, the single transferable vote, which uses multiple-member constituencies. To aid understanding of how STV could look, they have drawn an example constituency map of London:
This image, from the Electoral Reform Society shows Greater London divided up into 23 constituencies, each of which is made up of 3 or 4 existing constituencies, and as such elects 3 or 4 MPs.
It is possible to analyse the kind of result we would expect from this, but there are a number of caveats to this. By using actual, first-past-the-post election data, we cannot easily simulate how voters would vote on an STV ballot with several preferences.
However, if we assume that STV provides as proportional a result as it can within each constituency - mapping percentage of vote to percentage of seats, allowing for some rounding - then we can compare how these new constituencies might change the result.
We see that first-past-the-post does not produce a particularly proportional result here, and the larger parties (in particular Labour) are over-represented at the cost of smaller ones (especially the Lib Dems). The multiple-member constituencies are quite close to being proportional, with only a slight over-representation for Labour and the Conservatives.
The result, compared to FPTP, is much more proportional but note that the very small parties (Green and the Brexit Party) are unrepresented. This is because the electorate is still fragmented to some extent, just less so than with single-member constituencies: the more members per constituency, the closer to exact proportionality the results will be.
Depending on who you ask, this could be a good thing. One argument frequently used in favour of FPTP is that it keeps minority 'extreme' positions out of Parliament. Whilst structuring a system with this aim in mind would be questionable regarding fairness, requiring a party to get significant support across a wide area could be a benefit.
Regardless of this, it is undeniable that the multi-member constituency example is a huge improvement, and this reflects quite how disproportionate a result FPTP and its small constituencies create.
But the London example cannot speak for the whole UK. For one, almost the entire area is urban, built-up, and densely populated. The large multi-member constituencies are still geographically smaller than many current single-member ones across the rest of the country, and the demographic is very different.
Inspired by the London drawing, I have designed 17 constituencies across the South West, each comprising 3 or 4 of the current 55 seats. The region is mostly rural except for a few urban areas, which can be quite easily seen with the original constituencies much smaller in area.
The South West's rural-urban mix is not too dissimilar to the rest of the country (although it does have the lowest population density), and brings up a number of issues. Choosing which constituencies to combine becomes, in some areas, a very difficult and contentious matter (and I have not necessarily designed these constituencies with much thought for the local geography). Where a city has only 1 or 2 current constituencies, it generally has to be combined with surrounding rural areas.
It is important to note of course, that this task of drawing up constituencies using current boundaries, while necessary to allow easy calculation, is not representative of the actual operation. Given a target, for example, of 100,000 electors per member, it would be much easier to draw from scratch boundaries for a 3 or 4 member constituency, compared to amalgamating current ones.
Nevertheless, the task provides insight into how multi-member constituencies would still improve proportionality greatly, even if not perfect.
As poorly-designed as my South West constituencies may be, they provide a result that is far more proportional than FPTP in the same area. In this case, the Conservatives benefit hugely for being the most popular party in the South West, winning 48 of 55 seats. The proportional multi-member design brings their representation much closer to their vote share, allowing other parties to be better represented.
The extent to which this crude example improves proportionality is testament to how important and effective the implementation of multiple-member constituencies is. Even with just 3 or 4 existing Westminster seats hastily bundled together, the system is far more proportional than before.
This shows how much first-past-the-post harms fair representation, especially in areas like the South West, where a party with 55% of the vote takes 85% of the seats. The only way to rectify this is to stop fragmenting the electorate.
Comments