Although I’ve thought about it for a long time, I find myself increasingly unsure about electoral reform, and questioning much of what is being said after Thursday’s election. I’m certainly no fan of the current system, with its awkward patchwork of safe seats and local tactical battles; and I’m distinctly unhappy to see the Tories handed the power to implement some really damaging policies; but I’m not sure the two things are really connected, or what a better system would actually look like.
It seems to me that, 5 years ago, people saw the hung parliament result as an endorsement of the current system – the two-party system was breaking down, and in future elections people’s confidence in minor parties would increase, and the trend continue. When given a choice, in a national referendum, people overwhelmingly rejected the Alternatve Vote – whether as too complicated, not good enough, or simply no better than the status quo doesn’t matter. It’s true that turnout was lower than for the General Elections before and after, by about 10 million votes; but my back-of-the-envelope calculation says that to change the outcome, 90% of those stay-at-homes would need to have voted Yes1, which seems a bit of a stretch.
Now, those people have realised that 2010 may have been an anomaly, not a trend, as the 2015 results fit much more closely to the norm established over many decades; the decisiveness of the victory came as a complete shock, and while not a landslide, it led to 3 party leaders resigning in one day. Suddenly, they’ve woken up to the fact that the system’s natural tendency is to restore a two-party contest, regardless of how many other positions are presented.
A lot of people have been sharing the statistic about how many votes each party got, relative to the number of seats they acquired. It’s certainly an arresting comparison, but it’s not all that clear to me why there should be a direct correlation between the two. Should the number of people who turned out to vote SNP in Scotland affect how many Brighton voters were needed to elect Caroline Lucas? Should a low turnout in Central Manchester mean that city is afforded less importance in national politics? It depends what question you’re asking, but if the underlying aim is to rule the country in a way which accounts best for the different views and needs of its population, then obsessing over the mathematics of ballot papers is kind of missing the point.
If a referendum proposed a national or large-region Proportional Representation system, I would probably vote against it; a glance at any result map will show you just how different people’s views are in different regions, and that local link is an important one. I’m also not a fan of “top-up”, “Additional Member”, or “Parallel” voting systems, where an MP for Eastbourne would sit next to an MP for Nowhere In Particular, there to make up the numbers based on national share of vote. Any form of PR ingrains the power of political parties, and massively disadvantages strong independent politicians, who play a vital role in debates; and examples from around the world show that the neat idea of X votes for Y seats soon breaks down when the parties work out how to tactically manipulate the exact system in place.
Rather than trying to mix local, independent representatives with a national, proportional, executive, I’ve long argued that Government and Parliament should be elected separately. There is a risk that the two houses could end up in deadlocked opposition, as with the US mid-term elections; and I’m not entirely sure how to limit the power of parties in the Upper House. But more importantly, this would involve not just voting reform, but a complete dismantling and redesign of the Westminster system, such as only a revolution is likely to deliver. I suppose if Scotland had voted for independence, the rump UK might have had the appetite for such a reform, but it seems unlikely to happen any time soon.
A less radical but rather appealing option is the Single Transferable Vote, which combines a degree of locality with a degree of proportionality using constituencies with a handful of representatives, rather than one or dozens. You could, for instance, divide the country into Counties, with up to a dozen MPs for each; these MPs would be elected as individuals, not from party lists, but people would be much more likely to have a local representative whose political stance they identified with. The big problem with STV is explaining it; many people seem to have rejected AV because they didn’t understand it – either explicitly saying it was too complicated, or through misunderstandings like thinking it gives people multiple votes.2 STV has the extra need to reassign “excess” votes, a stage of the process which is often poorly explained3, but generally involves assigning votes a fractional value based on how far over the winning line the candidate got from the first preference.
Again, we have to wonder what question we’re actually trying to answer by inventing ever more complex voting systems. Apart from sheer surprise, a big reason for the current hunger for reform is that opinion is very strongly divided: baffling though it is to those of us toward the left-liberal quarter of the Political Compass, it seems that a large number of people actually think the Tories are a good choice; but equally a large number4 of people think they are a very bad thing indeed. But this isn’t a problem caused by the current voting system, or even by the notion of voting for representatives; it’s inherent in democracy of any kind. That is, there are people in the country who hold fundamentally incompatible views about how it should be run; no amount of tweaking the system will ever lead to a government that satisfies both groups of people.
I don’t have a neat conclusion to this post, because I no longer believe in a neat solution to this problem (if I ever did).
- That is: starting with the real result of 6 million Yes and 13 million No, you need to add 9 million extra Yes votes, and 1 million extra No votes, giving a final tally of 15 million to 14 million. [↩]
- If you still have this misconception, look at it this way: you can run an AV count with piles of paper, no photocopier, and no calculator; each vote is always in exactly one pile. [↩]
- It’s completely glossed over in this video which is being shared a lot, for instance, by mentioning only the simplest possible scenario. [↩]
- though not, it seems, an equally large number [↩]