Close elections are bad elections; what we need is social consensus

Screenshot_20191223-225856 mt Twote:
Close elections are bad elections. What we need is social consensus. Anything important decided by a narrow fraction of the most disengaged voters is destabilizing.  Professionalization of politics is the problem, not the solution.
Complaining about something like politics becoming more professional, or more efficient, is about as much use as - and indeed, is very similar to - complaining about evolution. But I'm used to disagreeing with mt over politics, so skip that bit as a detail and come to the "close elections" thought.

About which, as I half said in a different context (oh, there are so many thoughts in the world to write down!) in Men spake from God being moved by the Holy Ghost / Every man in his own language, I feel I've said before but can't find, so will say again:

Politics tends to produce close results1, but not consensus2. Nominally, pols are supposed to seek and promote consensus, and mt is still yearning after that idea, but I see little evidence that it is a major part in practice. And politics is backed by coercion: at least in majoritarian states: if you win, you get the power to impose your - sorry, the people's - will; this tempts far too many people. This tends to leave the middle ground barren and dead, populated with the corpses of those attacked by both sides. The solution, of course, is to move as much as possible out of the purview of politics, since anything done there will inevitably be fought over as a zero sum game. And move it out into the free market, where individual decisions are indeed made by agreement.


I should have reffed Aristotle's politics and the quote from Hayek: It is not democracy but unlimited government that is objectionable...


Does J R Oppenheimer ask: can science provide better models for democracy?


1. In an idealised society with a spectrum of opinion categorisable as from Left to Right, imagine two parties competing for votes, sure of all those on the side of the spectrum away from the other party; inevitably, they migrate to the centre. This was Tony Blair's chief idea, and lest you dismiss it too readily, notice that he was the only Labour to win an election for uncounted moons. I prefer conviction politicians - Blur IMO had no real ideas about what to do once he achieved power. The idealisation applies but in more blurred form in more complex scenarios.

2. Having commented on closeness, I feel the urge to comment on consensus too, since it is so strikingly absent. At least in the UK and USA; I'm rather less familiar with our Continental friends or those further abroad. But it is hard to know what to say. The UK is split by Brexit as one axis and but whether you dislike Corbyn or Bojo more on another, and that's not helpful. The USA is, apart from the perhaps-superficial Trump/Populist split, also divided between "free market" vs "progressive", but muddily, possibly leaving room for consensus in the details even if on principles there can be no meeting.

3. My picture shows Henry Worsley, who has nothing at all to do with this article. But I love the picture; I got it from the New Yorker. They shoved it into my fb feed for months on end and I finally got round to reading it. To my surprise, it doesn't show a USAnian: to me, he looks like one of the cowboy pilots from Catch-22, or the mad ones from Dr Strangelove. But no, just an Englishman.


Anonymous said...

I do think the US first-past-the-post systems with 2 parties is particularly bad: each party picks a candidate who represents the average motivated voter of their party, and then the election picks the candidate who represents the majority party. Thereby getting a candidate who is often at the 75th percentile or more of their state (if we represent candidates on a 1 dimensional axis).

A ranked-choice system would result in moderate candidates having a better chance. And I think that would make a difference.


(the UK system has just proved itself also broken...first past the post with multiple parties often means the median candidate of a _minority_ party can get elected...)

William M. Connolley said...

One could perhaps argue that FPTP is bad at consensus but that PR isn't? But I have no experience of PR.

David B. Benson said...

William, what are you attempting to state? FPTP? PR?

Nathan said...

I thin MMM means the preferential voting system in Australia.

So you could have voted Lib Dem 1, Greens 2, Labour 3, someone else 4, and so on.

You would find the candidate that wins would be closer in preference to the majority of voters.

David B. Benson said...


Nathan said...

Anon at the top

William M. Connolley said...

first-past-the-post = FPTP. Proportional representation = PR.

Graeme said...

The only difference is that in FPTP coalitions settle before the election. In PR, they settle after the election. In FPTP, the parties have to try to reach a majority of the electorate before the election. Under PR, anti - democratic tactics as espoused by the Libdems recently can actually result in a share of power

Phil said...

Yes, the jungle and the ToC is much better than a government.
Err, maybe not.

There are many issues that there is a consensus on. Elections are almost never about such issues. When they are, the election isn't close.

Graeme said...

What are your criteria, Phil? 100 small parties fighting to pull a coalition together?

Nathan said...

There's other choices than FPTP and PR


Our system is great, but you may not want the compulsory voting (I like it, but it's not essential)

Anonymous said...

Here's some data for the US House of Representatives showing one analysis based on "DW-nominate" scores:


I'd argue that the complete lack of any dots in the middle of the diagram shows the lack of moderate views, and leads to a system in which moderate solutions are generally not explored (though I'd argue that Obama tried to be a moderate and suffered because of it). And I'd argue that at least part of the cause of that gap is that the US is a two party, first past the post system, where moderates can never win the primary and where people vote party-line for the general election. And therefore a ranked-choice voting system could lead to more moderate candidates being elected. Ranked choice voting also eliminates the "spoiler candidate" problem (e.g., Nader in 2000 tanking Gore, or LePage winning two gubernatorial elections in Maine because of vote-splitting from anti-LePage voters - see https://en.wikipedia.org/wiki/2010_Maine_gubernatorial_election).

I don't understand parliamentary systems as well, but it seems like Labour plus Liberal would have beaten Conservative, so the FPTP multi-party systems leads to results that don't reflect the preference of the electorate. Either ranked-choice voting or proportional representation would fix that. I'm not as convinced that this would necessarily lead to more moderate behavior, relative to my conviction that ranked choice would help a lot in the US electoral system.


William M. Connolley said...

> Labour plus Liberal would have beaten Conservative

It doesn't work like that. There are Labour votes who would never vote Liberal; and vice versa. My (brilliant) proposal is that in addition to "for" votes, you could use your (still single) vote as "against" a candidate, without having to specify anyone to be "for". That would solve the tactical voting problem, and the Lab-not-vote-Lib problem.

Sam said...

I'm fairly convinved that the desire for electoral reform tends to be driven by two main forces. Firstly, lots of people who want PR tend to do so for partisan reasons since it'll benefit their preferred party (lib dems in particular). The SNP are a notable exception here, since they're pro PR, but under PR would probably get about half their current seat total, so I guess we can consider that a principled stand. The other thing driving it is that people see that we haven't been getting good government for the last decade or so, and so people want a policy lever to try and change that, and the easiest one is electoral reform. But I don't see much evidence that different electoral systems lead to better outcomes. Italy has experimented with almost every electoral system under the sun in the last few decades, and their politics is still a right old mess. Indeed people never seem to think through the implications of changing to a different voting system.

I mean, if we in the UK were to go to a PR system, you'd suddenly have 10-15% of the MP's from brexit party / ukip, and probably a similar number from further left greens or something like that. At least in the case of the brexit party lot, I'm not at all convinced that having these people in the legislature would improve the quality of governance. There's an argument that both lab and cons are currently stuffed fairly full of nutters anyway, which is probably true, but that's more to do with both parties being messed up in different ways (both of them have pretty terrible leadership selection procedures - give me hunt over johnson any day).

Where reforms are needed, are both in the parties themselves, and in the wider set of institutions around government. A more powerful lords (though maybe with slightly reformed selection procedure) and less powerful executive would be a good start.