Like with most of the country’s major infrastructure projects, electoral reform is one of those issues that British politicians often discuss, propose and even occasionally consult on but never either manage to either implement or cast off entirely. Although PR has a longer history in the UK than the “last bastion of FPTP in Europe” image of itself would suggest (Northern Ireland was one of the first polities in the world to run a PR election in 1921 and seats to Westminster’s university constituencies were filled by PR until 1950), attempts to enact it at a national level have failed, most notably in the landslide defeat suffered by the “Yes” side in the 2011 Alternative Vote referendum.
However, despite this long-standing debate, modelling of what the UK might look like under PR rarely go beyond either simply allocating seats in line with national vote share or estimating who would have won each single-member constituency under AV. In an effort to fill this gap, this two part series will attempt to gauge what the UK could look like under one of the most common PR systems in use: Single Transferable Vote in multi-member constituencies. This post (Part 1) will review the case for moving to STV, while Part 2 will look at how it could actually be applied in the UK and the results it would have brought in the 2015, 2017 and 2019 elections.
Some less spotted pros and cons of PR
The main arguments for and against PR are all pretty well rehearsed at this point. Proponents argue that PR ensures everyone’s vote counts equally, or at least more equally than under FPTP where the lopsided nature of many constituency elections mean that for many Britons, voting is more an exercise in civic duty than an act with a realistic chance of influencing a result. Moving to PR would remove this problem and ensure that Parliament better reflected the preferences of the voters, with greater representation of smaller parties and the end of parties winning large majorities on a minority of votes.
Opponents of PR warn that the Westminster system relies on strong, single-party majority government and that moving away from FPTP risks creating a never ending cycle of destabilising coalitions that so often plague countries in continental Europe (though even the strongest supporters of FPTP would likely concede that the reputation of large, single-party majorities to deliver stability has taken something of a beating in the past 5 years).
Rather than go into these often cited arguments in detail, I will instead mention a few of the other potential benefits and criticisms of PR (and STV in particular) that I think are often overlooked:
PR means that the geographical link to constituencies is lost. This argument is often made in a UK context, where the House of Commons operates on the basis of MPs representing specific geographic areas and that introducing a more national system via a “list” or other means would undermine this system. Politicians having to represent specific geographic areas certainly does have its advantages and the fact that the “Members of Parliament” section of certain constituency Wikipedia pages stretches back to the 14th century is not just very cool but also a testament to our long history of parliamentary democracy that would be a genuine shame to lose. However, the problem with this criticism is that moving to a PR system does not necessarily mean losing the constituency connection. That is one of the motivations for focusing this post on STV - it retains the “elections in constituencies” element that makes it (in my view) much better suited to the UK context than say a list system or MMP.
STV would however change the nature of those constituencies, and that’s good. Although STV would still retain constituencies, unless the British people are suddenly going to get onboard with a House of Commons of at least 2,000 MPs, then the current, single-member constituencies will need to be dragooned into larger, multi-member ones. Constituencies covering larger areas and more people would have two main effects. The first would be that constituencies would be more diverse and contain more different types of voter than the smaller, single-member seats we have currently. That, alongside the fact there are multiple seats available in each constituency drastically changes the incentives each MP or candidate faces (more on that below). The second is that larger constituencies would mean that issues pertaining to one, single area would be less important when it comes to the calculus required for an MP to get win election. For example, if there was a new housing development or wind farm that really annoyed 10,000 voters, it would be a lot easier for an MP to ignore them if they are in a multi-member constituency under STV with 500,000 voters than if they are in a single-member seat with 90,000. This in turn creates better incentives for MPs to vote in favour of net-beneficial policies that have concentrated harms but diffuse benefits. Moving to larger, multi-member constituencies could therefore help the UK overcome its chronic NIMBYism and related issues, at least at the margin1.
But what about the constituency work? Now, I can already hear howls from MPs and their staff at the prospect of multi-member seats, potentially covering populations of over half a million people and stretching across entire counties. I think there are several reasons however why these concerns, though valid, may not be as bad as people think. The first is that constituents will still tend to only contact one MP with their issues (potentially more if the first one fails to solve it), usually the one they voted for or support. In areas where one party is particularly strong, they will likely have multiple MPs in the same constituency, meaning they are able to divide the area up geographically. This is how constituency matters generally work for the Northern Ireland Assembly - DUP supporters will tend to go to the area’s DUP MLA and parties will often '“assign” different MLAs to different parts of the constituency, meaning MLAs are by and large not overwhelmed with surgery issues. Granted, this doesn’t mean that there won’t be any increase but this would simply be an argument for UK MPs to have larger staffing budgets (US members of congress represent areas that are much larger than what would be the case under STV in the UK, and are able to do so due to having many more staff). A move to STV may also be a helpful breakpoint to reorient the work of an MP away from the “glorified social worker” it has become over the past two decades, back to being more focused on legislating.
STV would result in many more different types of MP being elected. As already mentioned above, moving to STV would alter the incentives faced by MPs and this wouldn’t just be because of the larger population but also because of what they need to do to win (and hold) a seat. In a single-member constituency, you generally need to win at least 40% of the vote to stand a chance of winning. As a result, candidates are forced to appeal to as large a cross-section of the local population as possible. This has two downsides: the first is that it means that groups that are a minority in an area can often be ignored (this is particularly problematic from a representation stand-point if said group isn’t a majority anywhere) and the second is that since voter types tend to cluster, you end up with one party generally representing all of one type of seat. For instance, Tory MPs have become an almost extinct species in urban England since 1997 and Labour has always struggled to win seats in rural regions - despite both parties attracting significant numbers of votes in each of those areas. Moving to multi-member seats under STV radically alters this situation. Since a candidate no longer has to beat every other candidate to win a seat, they can afford to focus their appeal on a specific subset of the population, even if they are in a minority. This leads to more responsive and arguably effective representation, as well as a broader range of MPs and viewpoints within parties (imagine how different the Tory party would be if it had 40 of its MPs from the UK’s largest urban cores or if Labour had a caucus from rural and coastal areas).
And end one-party dominance of particular areas. Related to the point above, moving to STV largely does away with the concept of ‘safe-seats’. Even in constituencies where a single party is dominant, their need to win multiple seats means that any of those seats can be marginal. STV would also bring an end to single-party rule in particular areas, since there are very few examples where one party would be in line to win all seats in a multi-member constituency2. Overall, this would result in a significant increase in competitive pressure within our electoral system and just as competitive pressure improves outcomes for customers in markets, so it would improve outcomes for voters in elections. (Aside: this lack of competitiveness due to single-party dominance is much more of a problem at council level than in Westminster. The optimal strategy for implementing PR in the UK would be to start with councils - since it would be both less controversial and would familiarise the public into the system where the stakes are comparably lower).
Would that not just lead to more extremists being elected? One of the proudest boasts of British democracy is that while continental Europe fell to totalitarianism in the 20th century, the UK never elected a single fascist and only a tiny handful of communists to Parliament. First Past the Post is clearly a useful guard against fringe, extreme parties gaining a foothold in the legislature (and everything that brings with it) in that it sets a much higher bar for winning any seats than PR. However, that does not mean that it is a watertight seal, in fact when voters with more extreme views are concentrated in particular geographic areas, FPTP actually amplifies their representation compared to what would be achieved under PR. The example of UKIP also demonstrates how a party does not always require substantial Parliamentary representation to influence the political process under FPTP. Counterintuitively, fringe parties entering the legislature can also neutralise their appeal, often bringing greater scrutiny that exposes their weaknesses and limiting their ability to present themselves as the populist outsider, untainted by association with ‘the establishment’. Furthermore, the need to attract transfers from other parties to win seats under STV (more on this in Part 2) is a useful built-in mechanism to prevent parties drifting too far from the median voter.
Will we not just get endless “coalitions of chaos”? STV would mean that the odds of any single party winning a comfortable majority would be close to zero (even in 2019, the Conservatives likely would have come up just short). Although the argument that only FPTP can deliver strong, stable, single-party government has been undermined in recent years, the argument that coalition stitch ups between parties in smoke-filled rooms after the election is less democratic than FPTP retains considerable force. In fact, this was the argument that I used to find most convincing against PR. However, I think it is based on an unrealistic oversimplification of how political systems tend to work in practice. Even under FPTP, parties effectively operate as coalitions of different groups and interests under the same banner. Once in government, the policy process is often a negotiation between these various groups, not unlike what would happen under PR with separate parties. The key difference with PR however is that when these groups are in separate parties together in coalition, they are better able to extract credible commitments from each other in terms of agreed policy and concessions (since if the commitments are not held, the party will leave the coalition, ending the government - this mechanism is not as powerful when it is individual MPs within the same party). This in turn tends to increase the chances of parties and governments actually delivering on the priorities of the people who vote for them, which generally makes for a healthier democracy.
In short, the UK moving to STV isn’t just possible but may even result in several unexpected benefits beyond the usual arguments you hear in this debate. In Part 2 we will go through the set up of how UK elections could look under STV and review how that would have changed the outcomes of the 2015, 2017 and 2019 elections.
On a similar basis, one way to improve UK housing and infrastructure outcomes at the margin would be to move responsibility for planning from the district to county council level - the closer the population covered by decisionmakers in these decisions, the more likely they are to reflect total net benefits rather concentrated costs.
A possible exception to this would be the Labour party in Liverpool and certain parts of London, along with the Tories in some rural areas. However, even in these cases, the competition would be close for the final seat (more on this in Part 2).