Tuesday, 11 May 2010

So who won? Resolving the democratic dilemma

Regular readers may be relieved to hear that I'll be back onto regular economics soon. If nothing else, even I will be bored of the election before long.

But I have a couple more political articles to come first.

The debate over forming a coalition to govern the UK seems to be mainly focused on the question of "who won?" Was it the Tories, because they are the largest single party with 36% of the vote (and 47% of the seats)? Was it a combination of Labour and Liberal Democrats, because together they represent over 50% of people? Or, similarly, the Conservatives and Liberal Democrats, who have about 59% between them? Should we reform the voting system so that the majority of votes always reflects a majority of parliamentary seats?

All this debate ignores a deeper point. This is not a football match. The electoral system is not intended as a talent contest to award a prize to whoever can perform better and impress more people. It's meant to provide a way for the preferences and needs of the population to best influence the decisions of society as a whole.

Even had the election been decisive, and were there no third party to confuse the issue, we'd have (say) 55% of votes and 65% of seats held by one party. Sure, a clear victory. But does this mean the wishes of 45% of the population should be ignored?

The Tories have certainly "won" in some sense, with 36% of votes and a clear swing in the nation's preferences towards them. Of course, 30% would have voted Tory no matter what, but 6% of people clearly switched their vote from Labour to the Conservatives. That 6% have effectively decided the election (or not quite decided it, as is becoming clear).

So if Labour and the Liberal Democrats do form a coalition (though this now looks unlikely now due to an improbable alliance between Labour backbenchers and Conservative newspaper owners) they have to acknowledge that a big chunk of the country has voted Tory. The choices of these people should be listened to.

So, of course, should the choices of the 64% who voted another way. But how can we listen to both sets of people, when their policy preferences are often opposed?

What is clear that any government which follows the will of 55% of the people and ignores the preferences of the rest is not maximising societal utility. Of course, the 45% deserve less influence than the 55%, but not none at all.

There is no obvious way to achieve this with a purely representative system, and especially not a geographical one. Compromises are possible - a Lab-Lib coalition with Ken Clarke as business secretary, say? Not very practical, I fear. But there could be another way.

In economics, we don't expect the provision of shops, products or services to be governed by majority rule. Just because more people use Tesco than any other supermarket does not mean Sainsbury's, Asda and Morrison's - and that deli round the corner - can't exist too. Even if Tesco and Sainsbury's together could get over 50% of the market, they would not get to shut the others down.

Coca-Cola might have over 50% of the UK cola market, and your local pub is more likely to sell you Coke as a result, but Pepsi, Tango and milk still exist.

Instead, people choose where to allocate their resources and any conflicts or competition are resolved this way. If you and I both want to live on a particular piece of land, we simply bid for it and whoever's willing to pay more gets it. If I win because you'd rather spend your money on Pepsi instead of on a bigger house, you can do that. Ultimately, somewhere in the world, because of our choices, a piece of land will be used to grow sugar cane or provide accommodation for a family.

In this way, whatever's important to you gets your attention and your resources. And we all get a better deal as a result. I can trade what matters to me against what matters to you, and we will both end up with more of what we want. The world's limited resources are allocated efficiently.

Now, what if we could do the same with policies?

Personally, I don't feel strongly about whether Trident is renewed or there's a tax on £2 million houses. On the other hand, I'd really like a permissive policy on immigration, closer integration with Europe and a bit more monetary stimulus. I don't mind taxes staying where they are, or going up a bit. But I think money in some parts of the health service could be spent a bit more efficiently through market mechanisms, freeing up some resources for better education.

There's no party I can vote for which expresses these views. But I'd be quite happy to trade with someone who feels strongly about different things, so we can both get a bit of what we want.

I'm not calling for a Swiss-style referendum system - that's too simplistic, and leads to policies being considered in isolation without taking into account the trade-offs between them. Indeed the strength of the party system is that each party has put forward in its manifesto a particular combination of trade-offs which they feel will lead to a better society.

But it's clear from the result that none of the packages on offer are widely supported. And this will always be the case - until we put some kind of scoring mechanism on policies, so that people can show how much they really care about different things.

So what I propose is a points system. We each get 100 tokens to allocate to different policies. Each of these policies has a price, and we choose which ones matter to us.

How is the price set? By the tokens of those willing to bid against it. If I pay 50 to stop Trident being renewed, you need to pay at least 51 to get it.

Of course these things bring financial costs too, and if Trident is approved then about £3 billion a year in taxes will have to be found to pay for it. These are intrinsic parts of the policies, so you can't bid for Trident without also being willing to accept the cost of it. If you've spent all your tokens on Trident, you may find that someone else's tokens get to decide whose income tax is going to pay for it.

A system like this would need lots of technology to implement it - an eBay-like bidding system is an example of a similar, though simpler, algorithm. And perhaps most voters, at least today, would not be ready for this level of complexity. So I'd still be happy to see a representative democracy where our elected MPs cast these votes on our behalf - especially if they're given a formal duty to represent the broad interests of their constituents and not of their parties.

A simple version of this system could even be used now, to produce a policy programme for a coalition government. It seems reasonable to believe that:

  • 36% of people want a smaller government, a cap on immigration and disengagement from Europe, and more local control over schools and hospitals
  • 29% of people want roughly the status quo, with an emphasis on equality and help for the poorest, and a fairly strict defence and justice policy
  • 23% of people want an immigration amnesty, pro-European policies, substantial redistribution in the tax system and more civil liberties
With this knowledge, it shouldn't be impossible to bash out a policy programme that roughly balances the priorities of the electorate.

In return for more redistribution, Labour voters would probably be willing to give up their party's security policies and tight central control of public services. Liberal Democrats would have little choice but to give up on Europe and immigration, but could get their way on civil liberties and some tax reforms. And Tory voters would be able to cut public spending and enhance local control through private involvement in some aspects of education and health, but might not be able to get their way on the form of tax cuts.

European policy would probably remain centred around the centrist position that Labour has implicitly pursued, with a bias towards the Tories' preferences as they won more votes than the pro-European Liberal Democrats.

Sure, the complexity of the horse-trading involved in this kind of compromise would make the last few days look like a game of Countdown. But compromise is what we elect politicians for.

A proportional representation system would still be an improvement on the current situation - it would help to make these negotiations more balanced, as the number of MPs putting forward each position would better reflect the electorate's choices.

But the core problem with any winner-takes-all system - FPTP or coalition - is that there is no single identity for the country, and nobody has made a single decision. There is no one thing that the population "wants". Indeed, Kenneth Arrow's famous Impossibility Theorem says that there is no way to perfectly reflect the preferences of a population in the decisions of a government. But a system that accurately transmits voters' wishes into a policy compromise would surely be a step closer to the ideal.

Of course nothing like this is going to happen in the near future. But I hope that whoever ends up in power will have the wisdom and restraint to govern in the interests of the whole country - not just the minority, or even majority, who has chosen them.


Anonymous said...

This will astonish you!



Money as debt


project camelot magnetic motor





Monatomic Gold



isomorphismes said...

So in my view, Arrow's theorem tells us a worst-case bound on agreement. But in a country like the UK opinions are sure to be more homogeneous than the nastiest arbitrary arrangement you can think of.

I haven't come across any literature that takes this view on the theorem but that's the way I've always seen it. To adapt it to your point: it may be possible to winsorise the most extreme voting blocs and satisfy all 4 axios.