What’s Next?

Looking back at Nick Clegg’s conference speech it was interested to see he had a call in there for ‘fewer MPs’. I vaguely remembered this coming from him before and I’m still not sold on it. The idea seems to be that fewer MPs will require less money and drive down the cost of politics. This is technically true, but would also drive up the size of constituencies and require politicians to rely more on impersonal campaigning techniques.

There is a necessary cost of good politics and it’s not convincing that losing only 150 MPs makes enough difference financially to be useful - Clegg’s figure of £30 million is a fair amount of money but in the scheme of government losses there are more obvious funding problems than directly attacking things at the core of our already ineffective democracy. This single government IT project is £315 million over budget - offsetting the predicted gain from cutting 150 MPs for 10 years.

We’re not breaking the banks with MPs expenses - the story there is elected officials treating the public purse as something less than sacred, not that they’re causing the debt all by themselves. It’s a nice populist message of ‘too many politicians stealing our dough’ and to be honest it’s fairly harmless - our current system could probably work about as well with about 150 fewer MPs, but the problem there is our current system hardly works at all and reducing the number of MPs would create issues when trying to fix that.

Taking the current wisdom that three-member STV is our best bet for fair electoral reform in the near future, we’ll have to triple the size of constituencies. These would be shared between three MPs but the number of citizens in each locality would then be around 280,000. Whilst still not quite the ridiculous 700,000 per district you get in US congressional districts, it’s approaching a high figure. How high is too high is a matter of debate, but if we’ve followed Clegg’s plan and cut the number of MPs by 150 beforehand, that figure’s gone up to 360,000 people per constituency, at which point we’d be hard pushed to make the case constituencies are still ‘local’ areas. Reducing the number of MPs at this point would making the case for any multi-member system later much harder.

On the other hand, If we increase the number of MPs to 969 (and I think around 1000 would have to be an upper limit on what’s practical), the typical three-member constituency would cover almost exactly double the current constituency population - making boundary changes less of an obstacle. We’d then have a more proportional Parliament, with MPs sharing responsibility of areas not much larger than constituencies currently cover. Increasing the number of MPs might be a useful companion reform to sell multi-member constituencies to sceptics who find locality more important than proportionality.

Now yes, this would cost us - but assuming for a second the money couldn’t be found elsewhere there’s seems to me to be a another area of democratic excess just asking for the cull: The second chamber. Creating a functional, proportional and local primary chamber would be a better use of money than trying to create a second chamber that’s current design plan seems to be to make something democratic enough to justify its existence, but not so democratic it poses any awkward questions about the Commons. If cutting down to a single chamber could pay for an expansion in the number of MPs, it’d be worth it.

I’m one of those annoying people who when given any institutional problem in politics will usually try to trace it back to the electoral system. Gender representation? Electoral system. Party financing? Electoral system. Disengagement of electorate? Electoral system. How do we save the Union? Electoral reform. The trouble is I’m pretty sure I’m right (with some reservations I’ll come onto later). Democracy is about more than elections - you also need choice, something our current electoral system is pretty poor at providing. This Tuesday in the US there are 30 congressional seats that are so safe the opposing party isn’t even bothering to stand a candidate. Uncontested seats are only the most extreme end of the problem of safe and marginal seats, where only people in a few places realistically have a chance to influence the election and the rest can be written off as won or lost without any need to go deeper than that. This seems to me as a fundamentally unhealthy form of democracy. We have a system that demands we have as few political parties as possible, that those political parties must occupy a vast ideological tent and then allows these parties to only focus on a minority of the voters to swing elections without any mechanisms in place for this tactic to hurt them - of course people are disengaged from politics!

Another discouraging aspect of our current system is that it asks that people play down things they care about that their closest party doesn’t because otherwise the other party will get in and that’s would be even worse (This has been especially prevalent in the current US Presidential election). It encourages politics about preventing worst-case rather than trying for best case. It can’t be a coincidence that in the last election we barely had a majority of people turn out to vote - but more worryingly the majority of people who did vote voted for candidates other than the eventual winner. Most MPs only have the support of a minority of their constituents - there is something deeply flawed with a democratic system that meant a plurality of people who were engaged enough to vote didn’t get the representative they wanted.

There is quite clearly something wrong with this arrangement. The most obvious thing to do would be a move towards a system that distribute power in some way according to the actual support parties received at elections (My thoughts on exactly what kind of system will form a future post), but even this common sense idea comes up against resistance, and not just from ‘it may be broke but it hasn’t exploded yet so let’s keep going’ Conservatives. I was with a group of Lib Dems the other day who were sceptical about a PR system because the BNP could benefit from it. This bogeyman of extremist parties benefiting that’s constantly brought up against PR systems makes no sense, it may be far less likely for an extremist party to get into power – but once there there are currently no real checks on them. If we were serious about protecting people from the possibility of extreme government, we should create as democratic a system as possible, then put checks into the system. The solution isn’t crippled democracy but to create strong protections against an extremist government in the form of a bill of rights.

In general I’d say there’s often a failure to think about the big picture - that electoral reform is a vital cog of a new British settlement, but there’s a whole lot wrong with our system and we need a lot of different approaches in various areas to resolving it. We can’t fix all our problems with a magic electoral system, we’ll also need a new structure around it to manage the new reality that system may present. In the next few weeks I’m going to try to throw a bit of my thinking on various aspects of that here.

There’s news going round that the Government might seriously be looking at changing the electoral system to some version of the Alternative Vote.

I’m a bit torn on this, on one hand I’m glad the problem is being thought about; on the other hand I think AV is a bad solution to the problem. I’m disappointed that there are only currently plans for one additional preference, this gives us the same dynamic we have for Mayoral elections, where you vote your conscience with your first vote, and then cast your ‘real’ vote for one of the main parties. This gives the illusion of greater choice and legitimacy; whilst in the end very little has changed. The system needs at least three votes to avoid the problem, as this allows the possibility for votes to condense around a common third party candidate before one of the main parties. This will do nothing to increase representation of parties already under-represented at Parliament and in fact the system can produce more un-proportional outcomes than under FPTP. This is a system that seems to fix the problem by giving MPs greater legitimacy without actually changing very much. Of course, the appearance of progress without actually having to sacrifice anything is doubtless too tempting a prospect to resist.

In my opinion AV is only appropriate when there can only be one of a position, like a Mayor, President or other singular executive position. MPs’ aren’t like that and there’s no need for only one MP to represent an area, in fact the people are better represented and Parliament will be more proportional if an area is represented by multiple MPs. For a legislative body we really should be looking at multi-member proportional systems like the Single Transferable Vote or (my pet project) the Delegated Vote instead of systems that just solidify existing control.

I see two possible ways this could affect future reform:

1. AV will bring in a slightly more proportional chamber (by no means a certainty), so that a transition to a more proportional chamber will be less of a shock on the main parties and a more proportional system like STV or DV could transition with less opposition than from our current state.
2. AV removes the most potent weapons in our electoral reform arsenal (the absurd number of MPs without majorities) and superficially addresses the concerns in our current voting system and pushes all call for reform far into the future.

I hope for the former, but find the latter far more likely. It depends if we want the strongest possible electoral reform argument, or the possibility of a Parliament that might be slightly more friendly to it. Alternatively, we could end up with a Parliament that’s not much different to what we have now, but believes it’s fixed the problem.

Those of my readers who have been around a while might remember I threw an electoral system out there a while ago. I’ve had a few months to tweak it and address the problems people had with it and as a little bonus I put together a short video with nifty graphics to explain it. The main site can be found here.

Watch the Video :

The Delegated Vote is a proposed electoral system that doesn’t give representatives equal power, but instead gives them voting power equal to the number of votes they receive. Using this foundation principle, the value of the vote becomes consistent across constituencies allowing votes to flow across constituency boundaries and proportionality to be expressed exactly.The following four points are the foundations of the system (they can also be explained by watching the video above):

MPs have different voting powers – By getting rid of the idea that all MPs need to have an equal vote, it makes it possible for all voters to have an equal vote. Currently votes in different constituencies can be said to have different values because of the different influence they have on the outcome. By rooting the value of MPs vote in individual votes, votes at election retain a universal value. This allows us to express a proportional outcome far more precisely.

Top Candidates standing in an individual constituency automatically elected – As their voting power differs, this creates a proportional outcome. The goal is that a majority of voters are represented locally. At present, the system states that the top two candidates are elected; it might turn out to be better if it were the top three.

Constituency Pooling – As the vote has a universal value, votes can move across constituency boarders. Parties without concentrated support gain seats by standing a candidate in multiple constituencies and pooling the votes between them. If they gain a seat is decided by whether the number of votes they receive falls between an upper and lower boundary.

Upper and Lower Boundaries – The upper and lower boundaries are created from an average of the top 40 and bottom 40 vote counts of MPs elected through a single constituency. These exist so that all MPs fall within a broad range and there is not a lower criterion for entry given to MPs entering through multiple constituencies. Similarly it prevents there being any advantage for parties to stand in multiple constituencies instead of a single one, refocusing them towards local representation. The object is more to guide party behaviour than disqualify candidates.

The foundations of the system can be seen in The Value of the Vote, whilst Gerrymandering Rehabilitated explores the change in the nature of constituencies this brings about that makes full proportionality possible. Comparison with STV puts the system up against the best current proposed replacement for our voting system.

The purpose of the site is feedback and discussion so please leave a comment on any of the pages or get in contact at alex@alexparsons.co.uk

Do you get a more proportional seat distribution (in terms of parties) if you give seats out at random instead of using our current FPTP system? Insanely, yes.

First I worked back from the 2005 results to work out what the proportional outcome in seats would be. Then I compared the actual returned seats and got the total number of differences in party allocation (274). This number doesn’t mean much by itself (it’ll count both the party the seat should have belonged to and the seat it now does) but it’s useful for comparison purposes as distributions less than this number will be closer to a proportional result.

The next task was to write a short macro to throw out 646 votes at random from the 27,148,510 cast. After running this 200 times the results were consistently more proportional than the actual 2005 results, giving an average distribution difference of 61.69.

That isn’t that impressive as we’re working from a very large sample. The next step was to randomly pick a vote from each constituency’s results as opposed to the whole country. Surprisingly, this still gives a far more proportional outcome with an average distribution difference of 68.24.

So as it turns out, picking votes randomly out of a hat is more democratic in terms of party-seat distribution than giving it to the person with most votes. Crazy huh?

There’s an interesting letter in the Telegraph from the 2005 Tory intake calling for US style recall elections. It sounds nice but the world it belongs in is the one where most MPs enjoy the support the vast majority of voters and a recall can reflect a genuine shift in attitude. The problem is that we don’t live in that world and most MPs don’t have enjoy participially high levels of support in their constituency, 426 MPs had more of their constituents voting for other candidates than themselves and 56 had over 60% of their constituents vote against them (conversely only 35 had 60% or more vote for them). The fact that the system allows a majority of voters to be represented by people they didn’t vote for should really be one of the first places people look when trying to work out why no one trusts our political system.

Implementation wise I can see problems. As support for most MPs is quite low, the only way to prevent abuse of this system would be to have a prohibitively high level of signatures to authorize the recall. For this to happen they’d have have committed an offense so large that their removal would be a foregone conclusion and the recall would effectively be  a rubber stamp. Even more worryingly, unless all other parties bow out and allow a single alternative candidate to stand (which has happened), there is every possibility they could still win.  Adding a little bit of democracy could result in a profoundly undemocratic outcome, where an MP elected with 70% of the vote could lose over half their support and still retain their seat with 30% of the vote if the rest of the vote was divided between rival parties. Trying to patch over the holes in our political system just reveals it’s deep inadequacy.

Recall elections are a good idea when applied to executive positions like the example given of the California governorship, but in the UK we lack the same culture of elected executives (City mayors are the ones I can think of). In practice, they can’t be truly effective in legislatures.

So I like the sentiment, but I’d rather they were thinking about things that really would improve our political system. I don’t think the Tories are fundamentally a lost cause in this area, I think they’re are in a good position to deliver on some of my wish list; after last Autumn they could easily find enough support for fixed term parliaments and their experimentation with primaries leaves me hopeful they’ll extend that. I doubt they’ll be brave enough to talk about term limits and electoral reform, but they really should be if they’re serious about fixing the system.

I’m revisiting my electoral system (last seen here) with some substantial revisions and improvements, and I’m trying a side experiment to see if effectively pulling 646 random votes out of a hat produces a more proportional government than our current one. Amazingly it seems so far it does, we actually do better by random chance than FPTP. I’m sure statisticians will be unsurprised but that’s going to keep me amazed for a while.

Our current system where the PM decides they have a good chance of winning (having just won a war or adjudged the interest rates perhaps) is clearly stupid. Not only does it lead to time and money wasting non-events like this November, it’s just a power asking to be abused.

The danger with fixed terms is it gives campaigns something to work towards and allows the for absurd US campaigns that have been in full power for months, despite the election being a year away. The advantage of the British system is that campaigns are short and (relatively) inexpensive.

We need more than a fixed date, we need a random number generator. You could set criteria to prevent it happening a week later, perhaps you could also make more full use of the percentage of the vote the government received (or opinion polls) to make less popular governments shorter (an automatic national vote of no confidence).

Whilst I quite like the idea of an complex and mysterious process (it would have to be to prevent second guessing), there are some quite obvious flaws here. The usual standard of Open Source for the program would be important to prevent it being obviously highjacked, but with said mystery there’d always be that uncertainty.I doubt we’re ever going to get to the point where people feel good trusting a computer with this vital function, and this distrust will be important in the inevitable robot uprising.

Still, It’d be pretty neat.

I currently have a post over at OurKingdom discussing the problems with FPTP and my alternative electoral system.  My full length piece on the system can be found here.

Depending on feedback, I might rewrite it as a series of smaller blog posts. Alternatively I’ll find out it’s an old idea that’s been dismissed and feel a little silly about the whole thing, but isn’t learning fun!