Problem: A party might be small because, despite having widespread support, it’s seldom sufficiently clustered to get its candidates over the threshold and into Parliament. An example of this is the Liberal Democrats, who got 2.4 million votes, but only eight seats. Another party might be small because it’s not standing candidates in every seat, but where it does, they get in. In that second case, they might not have many – or any! – votes other than in the seats that they won. The SNP are in this second position at the moment, with 1.5 million votes and 56 seats. Their electoral success would mean that their vote weighting was reduced, as they aren’t aggregating many votes from seats they didn’t win.
Response: Actually, I’m not sure why this is a problem – Weighted Votes will mean that each party’s influence in Westminster is proportional to its support across the UK. Their supporters will still benefit from them having an incentive to maximise their vote, which is the point of WV.
Thursday, 10 November 2016
Wednesday, 9 November 2016
Weighted Votes: Objection 3 - Over-powerful MPs
Problem: some parties have widespread support, but it’s not concentrated enough to get them over the line in any one constituency. If, by chance, they do get over the bar in one seat and get an MP, that MP is then representing not only their own constituents but also the interests of all their party’s supporters across the country. These are two very different – possibly conflicting – responsibilities, and there is a danger that an MP who wields a million votes will use them to give an unfair advantage to their own constituents, or will sell them to the highest bidder.
Response: Every MP has a dual role – representing her constituents, whether they voted for her or not, and (along with her parliamentary colleagues) representing the interests of all her party’s voters, whether they returned an MP of that party or not. The problem here is real. It’s just that, when this duality is condensed into one person, the conflict isn’t different in kind, it’s just more obvious.
As long as we require MPs (whom we elect to represent their constituents) to also represent their party, this conflict is inherent in the system. Right now, that conflict is used as an excuse to disempower supporters of thinly-spread parties - we tell them that, if they can't get an MP, they don't deserve a voice. That’s what PR is meant to cure. The conflict still exists, we just have to push it onto a different aspect of the system – in this case, the conflict of interest inherent in the MP-Constituency link.
As for the risk of corruption; yes, this is a real possibility, and it’s easier (and cheaper) for a lobbyist to buy one MP with a million votes than three Liberal Democrats, twenty-five Labour or thirty-nine of the fifty-six Scottish Nationalists. This will require laws on honesty and transparency to be passed and enforced, and not just on MPs but on journalists. But the MPs of the larger parties will have every incentive to keep the smaller parties honest, until we can choose a way to dilute the influence of over-powerful MPs.
Response: Every MP has a dual role – representing her constituents, whether they voted for her or not, and (along with her parliamentary colleagues) representing the interests of all her party’s voters, whether they returned an MP of that party or not. The problem here is real. It’s just that, when this duality is condensed into one person, the conflict isn’t different in kind, it’s just more obvious.
As long as we require MPs (whom we elect to represent their constituents) to also represent their party, this conflict is inherent in the system. Right now, that conflict is used as an excuse to disempower supporters of thinly-spread parties - we tell them that, if they can't get an MP, they don't deserve a voice. That’s what PR is meant to cure. The conflict still exists, we just have to push it onto a different aspect of the system – in this case, the conflict of interest inherent in the MP-Constituency link.
As for the risk of corruption; yes, this is a real possibility, and it’s easier (and cheaper) for a lobbyist to buy one MP with a million votes than three Liberal Democrats, twenty-five Labour or thirty-nine of the fifty-six Scottish Nationalists. This will require laws on honesty and transparency to be passed and enforced, and not just on MPs but on journalists. But the MPs of the larger parties will have every incentive to keep the smaller parties honest, until we can choose a way to dilute the influence of over-powerful MPs.
Tuesday, 8 November 2016
Weighted Votes: Objection 2 - Unrepresented Parties
Problem: It’s hard for any small new party to get a toehold in Parliament, and, under the raw WV scheme I’ve described, no-one is representing the opinions of their supporters.
Response: This is not a problem of WV, it’s a problem of how we choose MPs in the first place. We might decide, separately from the WV issue, that we will replace Plurality voting with STV in multimember constituencies, or have a second class of non-constituency MPs. But these are about who gets to represent a party in parliament, and can be decided separately from the issue of proportionality.
However, WV does open a different possibility, which is that a small party might assign its votes to another party that shares its values. Meibion Kernow, for example, might be OK having an arrangement of this sort with Plaid Cymru. Like any Confidence and Supply arrangement, it would have to be easy to dissolve it.
Some parties might not be able to find a partner, either because no other party was pure enough for them, or because no other party would accept the poison chalice of being associated with them. That then stops being the system’s problem.
Response: This is not a problem of WV, it’s a problem of how we choose MPs in the first place. We might decide, separately from the WV issue, that we will replace Plurality voting with STV in multimember constituencies, or have a second class of non-constituency MPs. But these are about who gets to represent a party in parliament, and can be decided separately from the issue of proportionality.
However, WV does open a different possibility, which is that a small party might assign its votes to another party that shares its values. Meibion Kernow, for example, might be OK having an arrangement of this sort with Plaid Cymru. Like any Confidence and Supply arrangement, it would have to be easy to dissolve it.
Some parties might not be able to find a partner, either because no other party was pure enough for them, or because no other party would accept the poison chalice of being associated with them. That then stops being the system’s problem.
Monday, 7 November 2016
Weighted Votes: Objection 1 - It’s untried
Problem: no-one else has ever tried this system. We have no precedent and no evidence for how it works.
Response: every voting scheme ever used was, somewhere, used for the first time. While it’s true that no other national government is elected this way, if it is ever used, someone will have been the first. Why not us?
But, of course, weighted votes are widely used – not to elect governments, but for meetings of shareholders and for TUC conferences. Company shareholders are no fools, and would not use weighted votes if it weren’t in their interests. Likewise, Trade Union leaders would not accept WV (the Card Vote system) if it didn’t work for them. I think we have enough evidence for WV.
Response: every voting scheme ever used was, somewhere, used for the first time. While it’s true that no other national government is elected this way, if it is ever used, someone will have been the first. Why not us?
But, of course, weighted votes are widely used – not to elect governments, but for meetings of shareholders and for TUC conferences. Company shareholders are no fools, and would not use weighted votes if it weren’t in their interests. Likewise, Trade Union leaders would not accept WV (the Card Vote system) if it didn’t work for them. I think we have enough evidence for WV.
Saturday, 5 November 2016
Weighted Votes: Final thoughts (before some additional thoughts)
Now, I've never seen Weighted Votes proposed for national elections, but it's the way companies work - the weight of a shareholder's vote is just the number of shares they hold. It seems to work pretty well for them. It's also used at TUC conference, where it's called the Card Vote - a union's delegation is issued a card with a number representing one vote, plus one vote per thousand paid-up members. It was this that set me thinking how it could be used for parliamentary votes.
While it could be extremely simple to implement WV – it would require the tellers in the division lobbies to add up a column of numbers, rather than just count heads – it could also be part of a more comprehensive change. So, for example, the electorate could have two separate votes at the polling station, one for your local MP and one for your preferred party (and this second vote would be the one that determines the party's weighting in Parliament). This is the system used in New Zealand, but they don't have WV, they use the party vote to allocate additional members from party lists.
The beauty of WV is that we could have proportionality right away, in a form that gives an extremely low Gallagher Index; and then add other components, such as the separate Party Vote, or Score Voting, or AV+, or rolling bye-elections, at a later date.
While it could be extremely simple to implement WV – it would require the tellers in the division lobbies to add up a column of numbers, rather than just count heads – it could also be part of a more comprehensive change. So, for example, the electorate could have two separate votes at the polling station, one for your local MP and one for your preferred party (and this second vote would be the one that determines the party's weighting in Parliament). This is the system used in New Zealand, but they don't have WV, they use the party vote to allocate additional members from party lists.
The beauty of WV is that we could have proportionality right away, in a form that gives an extremely low Gallagher Index; and then add other components, such as the separate Party Vote, or Score Voting, or AV+, or rolling bye-elections, at a later date.
Friday, 4 November 2016
Weighted Votes – how it might work
Each party is allocated all the votes that were cast for them, whether in places where they won or where they lost, and shares them among their MPs. So all the Green votes in (say) Hartlepool or Falkirk are bundled up and (in this parliament) Caroline Lucas casts all 1,157,613 of them.
This would mean that, in this parliament, a Conservative MP gets to wield a vote worth 34,244, a Labour MP’s vote is worth 40,290, and Douglas Carswell gets a hefty 3,881,129. Now, I hate the idea of Carswell having four million votes, but his party won those votes, and those voters deserve to have their voice heard. If they are not heard, the system is maximising the utility of people who support establishment parties. For the greatest happiness of the greatest number, our utility function must include the happiness of UKIP supporters.
Why do I hate the idea of one MP – whether Carswell or Lucas – representing the interests of a million or four million supporters? In one sense it’s no different from sharing that job across a dozen or a hundred or three hundred MPs. Except that, of course, having three hundred MPs means that extremists within a party don’t get things all their own way, they are moderated by their moderates. A single person wielding four million votes may be swayed by some idée fixe - or by lobbyists, or by misunderstanding, or by money – and just press on regardless in a way that a member of a larger party can’t.
What we have here is an inevitable contradiction. Proportional representation, combined with retaining the one-constituency-one-member link, unavoidably means that an underrepresented party has to have more power per MP than an over-represented party.
If we want to prevent some consequences of a system, we should look at what laws, and what transparency, will produce the effect we want. The arguments for getting money out of politics (including state-funding of parties) are well-known. I won't go over them here.
Overall, WV has all the benefits of other PR voting systems; but it still maintains the constituency link, doesn't require voters to learn a new system, doesn't create two classes of MPs, and enforces near-perfect proportionality . In WV the constituency link can be kept, because the power of a party in parliament is not governed by how many MPs it has, but by how many votes it has. This also reduces the importance of boundary changes and encourages honest (i.e. non-tactical) voting. As long as the party you support can get at least one MP, your vote counts.
(What if a party can't get even one MP? I would propose that a small party could do a deal with a slightly larger one that allows them to allocate their votes to that party’s MPs, though both TUSC and Britain First might have problems finding anyone willing to be associated with them. More about this in a later post.)
But, crucially (for me) WV means that the temptation to campaign only to floating voters, in marginal constituencies, is greatly weakened. The touchstone of electoral success would become the greatest happiness of the greatest number, not the greatest happiness of the most fickle.
For example, at the moment there is no benefit to the Labour Party in getting more people in Barnsley to vote Labour. Dan Jarvis already gets 55% of the vote, and would actually only need 33% to still win and be the MP, because of how the other parties divide up the vote. But with WV, it makes sense to get your voters out in numbers. Just edging it is not enough, you want the weight. The result is that candidates are competing for everyone's vote, not just their main opponent's soft fringe. Voters get wooed. Hard.
And, knowing that their vote has an effect, voters will be more likely to turn out. A Labour Party supporter in Maidenhead (Conservative majority of 29,000) has very little incentive to turn out and vote, other than as a slightly ineffectual protest. But under WV, their vote is just as useful (in terms of electing a government) as that of a Labour Party member in Gower (Conservative majority of 27). All those people, of whatever party, who say “why bother voting? My vote won’t make any difference” will now see that – whether they are voting for or against a dead cert in their own constituency – their vote will strengthen the hand of their favoured party.
This would mean that, in this parliament, a Conservative MP gets to wield a vote worth 34,244, a Labour MP’s vote is worth 40,290, and Douglas Carswell gets a hefty 3,881,129. Now, I hate the idea of Carswell having four million votes, but his party won those votes, and those voters deserve to have their voice heard. If they are not heard, the system is maximising the utility of people who support establishment parties. For the greatest happiness of the greatest number, our utility function must include the happiness of UKIP supporters.
Why do I hate the idea of one MP – whether Carswell or Lucas – representing the interests of a million or four million supporters? In one sense it’s no different from sharing that job across a dozen or a hundred or three hundred MPs. Except that, of course, having three hundred MPs means that extremists within a party don’t get things all their own way, they are moderated by their moderates. A single person wielding four million votes may be swayed by some idée fixe - or by lobbyists, or by misunderstanding, or by money – and just press on regardless in a way that a member of a larger party can’t.
What we have here is an inevitable contradiction. Proportional representation, combined with retaining the one-constituency-one-member link, unavoidably means that an underrepresented party has to have more power per MP than an over-represented party.
If we want to prevent some consequences of a system, we should look at what laws, and what transparency, will produce the effect we want. The arguments for getting money out of politics (including state-funding of parties) are well-known. I won't go over them here.
Overall, WV has all the benefits of other PR voting systems; but it still maintains the constituency link, doesn't require voters to learn a new system, doesn't create two classes of MPs, and enforces near-perfect proportionality . In WV the constituency link can be kept, because the power of a party in parliament is not governed by how many MPs it has, but by how many votes it has. This also reduces the importance of boundary changes and encourages honest (i.e. non-tactical) voting. As long as the party you support can get at least one MP, your vote counts.
(What if a party can't get even one MP? I would propose that a small party could do a deal with a slightly larger one that allows them to allocate their votes to that party’s MPs, though both TUSC and Britain First might have problems finding anyone willing to be associated with them. More about this in a later post.)
But, crucially (for me) WV means that the temptation to campaign only to floating voters, in marginal constituencies, is greatly weakened. The touchstone of electoral success would become the greatest happiness of the greatest number, not the greatest happiness of the most fickle.
For example, at the moment there is no benefit to the Labour Party in getting more people in Barnsley to vote Labour. Dan Jarvis already gets 55% of the vote, and would actually only need 33% to still win and be the MP, because of how the other parties divide up the vote. But with WV, it makes sense to get your voters out in numbers. Just edging it is not enough, you want the weight. The result is that candidates are competing for everyone's vote, not just their main opponent's soft fringe. Voters get wooed. Hard.
And, knowing that their vote has an effect, voters will be more likely to turn out. A Labour Party supporter in Maidenhead (Conservative majority of 29,000) has very little incentive to turn out and vote, other than as a slightly ineffectual protest. But under WV, their vote is just as useful (in terms of electing a government) as that of a Labour Party member in Gower (Conservative majority of 27). All those people, of whatever party, who say “why bother voting? My vote won’t make any difference” will now see that – whether they are voting for or against a dead cert in their own constituency – their vote will strengthen the hand of their favoured party.
Thursday, 3 November 2016
Weighted Votes: A modest proposal for voting reform
The arguments for Proportional Representation (PR) have been well-rehearsed – as have the arguments against. I’m going to assume we’ve listened to them all and decided that PR is a Good Thing – giving politicians an incentive to make us happy being the key point. But, now what? Which system should we institute?
One class of methods involves padding out the legislature with additional members. So, there are pure list systems, where no one member represents any specific constituency. There are additional member systems, such as MMP in New Zealand or AV+, where constituencies return members and then additional MPs are drawn from party lists to make up numbers.
STV and AV (including AV+) are very imperfect at producing even approximate proportionality; but, even if they were a lot better, they are blunt instruments, because any given politician either gets the job or doesn’t. A small change in the happiness of the electorate will go unnoticed by the system, because a change has to be big enough to trigger the removal of an MP of this party, and their replacement with an MP of that party.
Is there not a more supple, nuanced and responsive system? The Utilitarian Principle would have us look for a method that slightly punishes parties that do a slightly poor job, and seriously punishes those that do a seriously poor job. It would mean that a Tory voter in Barnsley who switches to UKIP has the same right to be heard as a Tory voter in Whitney who does the same thing. And really, none of the MP-based schemes can offer that.
Is there a method that can?
Yes. Yes there is. There is a system that we could introduce today that would instantly provide near-perfect responsiveness. It would give every voter the same power to rate our law-makers’ performance. It would enfranchise every citizen equally. It’s the Weighted Vote system.
More on this tomorrow.
One class of methods involves padding out the legislature with additional members. So, there are pure list systems, where no one member represents any specific constituency. There are additional member systems, such as MMP in New Zealand or AV+, where constituencies return members and then additional MPs are drawn from party lists to make up numbers.
STV and AV (including AV+) are very imperfect at producing even approximate proportionality; but, even if they were a lot better, they are blunt instruments, because any given politician either gets the job or doesn’t. A small change in the happiness of the electorate will go unnoticed by the system, because a change has to be big enough to trigger the removal of an MP of this party, and their replacement with an MP of that party.
Is there not a more supple, nuanced and responsive system? The Utilitarian Principle would have us look for a method that slightly punishes parties that do a slightly poor job, and seriously punishes those that do a seriously poor job. It would mean that a Tory voter in Barnsley who switches to UKIP has the same right to be heard as a Tory voter in Whitney who does the same thing. And really, none of the MP-based schemes can offer that.
Is there a method that can?
Yes. Yes there is. There is a system that we could introduce today that would instantly provide near-perfect responsiveness. It would give every voter the same power to rate our law-makers’ performance. It would enfranchise every citizen equally. It’s the Weighted Vote system.
More on this tomorrow.
Wednesday, 2 November 2016
Weighted Votes: The argument for proportional representation
The argument I suppose I’ve heard most often in favour of PR is that it’s “fair”, while the argument I’ve most often heard against it is that it produces weak and indecisive governments. I’ve also heard that it gives disproportionate influence to fringe parties (if they hold the balance of power in parliament), and allows extremists and fascists a voice.
I’m unimpressed by all of those arguments. The “fairness” argument is meaningless. The purpose of elections (I’m arguing) is not to create a parliament that resembles the electorate, it’s to force the legislators to try to keep us happy, to give them an incentive to seek the greatest happiness of the greatest number.
The same goes for the fringe party or extremists argument; if significant numbers of people think that expelling the Poles, or the Muslims, or the Jews, or the tailors, is the answer to their problems, then they must be seriously unhappy, which means the legislators are doing a horrible job. Frankly, the politicians that destroyed the stock of social housing and then blamed immigrants for the lack of housing had only themselves to blame when people voted to leave the EU. And you don’t cure that sort of poison by denying people a voice, you cure it by engaging with it, telling people the truth, and trying to make people as happy as you can.
The argument from the desirability of strong government, I think, misses the point of democracy altogether. Chile under Pinochet, Spain under Franco, Britain under Thatcher, Russia under Putin – all these are examples of very strong government. But what’s the point of having a strong government if it is not considering the happiness of its citizens? No, if you want to form a strong government, you should have to persuade the electorate to put their trust in you. That sort of strength should not be given by default just because you crossed an arbitrary line.
So, what is my argument for PR? It’s this. The current plurality system (colloquially known as “first past the post”) creates a distortion in the system. Some constituencies are so reliably held by one party – so-called “safe seats” – that the Utilitarian rating system doesn’t work. Some voters in that constituency will be deeply unhappy with the party that holds the seat, but their ratings are just deleted from the system. But even the voters who broadly like what the incumbent is doing can generally be safely ignored. The only people that need to be listened to are the voters at the margins, the persuadable ones, the “floating voters”, the median voters.
And this means that they are the ruling class. It is their utility that is considered. Rather than measuring the sum of happiness, we’re measuring a weighted sum, where extra weight is given to the happiness of a special group.
If any version of democracy is going to even remotely track the greatest happiness of the greatest number, it has to include everybody’s happiness in the aggregate score. Any scheme of voting that allows politicians to safely ignore a sector of society condemns that sector to non-citizen status. So a Labour MP, in a solid Labour area, can safely ignore what her Tory (or UKIP, or Green, or LibDem) constituents want, because the only people she needs to convince are the people who voted Labour last time but might not this time. Even if the majority of voters would be happier with (say) a state-owned rail system, she can safely ignore their wishes because the floater is king, and polls have told her what they want.
I’m unimpressed by all of those arguments. The “fairness” argument is meaningless. The purpose of elections (I’m arguing) is not to create a parliament that resembles the electorate, it’s to force the legislators to try to keep us happy, to give them an incentive to seek the greatest happiness of the greatest number.
The same goes for the fringe party or extremists argument; if significant numbers of people think that expelling the Poles, or the Muslims, or the Jews, or the tailors, is the answer to their problems, then they must be seriously unhappy, which means the legislators are doing a horrible job. Frankly, the politicians that destroyed the stock of social housing and then blamed immigrants for the lack of housing had only themselves to blame when people voted to leave the EU. And you don’t cure that sort of poison by denying people a voice, you cure it by engaging with it, telling people the truth, and trying to make people as happy as you can.
The argument from the desirability of strong government, I think, misses the point of democracy altogether. Chile under Pinochet, Spain under Franco, Britain under Thatcher, Russia under Putin – all these are examples of very strong government. But what’s the point of having a strong government if it is not considering the happiness of its citizens? No, if you want to form a strong government, you should have to persuade the electorate to put their trust in you. That sort of strength should not be given by default just because you crossed an arbitrary line.
So, what is my argument for PR? It’s this. The current plurality system (colloquially known as “first past the post”) creates a distortion in the system. Some constituencies are so reliably held by one party – so-called “safe seats” – that the Utilitarian rating system doesn’t work. Some voters in that constituency will be deeply unhappy with the party that holds the seat, but their ratings are just deleted from the system. But even the voters who broadly like what the incumbent is doing can generally be safely ignored. The only people that need to be listened to are the voters at the margins, the persuadable ones, the “floating voters”, the median voters.
And this means that they are the ruling class. It is their utility that is considered. Rather than measuring the sum of happiness, we’re measuring a weighted sum, where extra weight is given to the happiness of a special group.
If any version of democracy is going to even remotely track the greatest happiness of the greatest number, it has to include everybody’s happiness in the aggregate score. Any scheme of voting that allows politicians to safely ignore a sector of society condemns that sector to non-citizen status. So a Labour MP, in a solid Labour area, can safely ignore what her Tory (or UKIP, or Green, or LibDem) constituents want, because the only people she needs to convince are the people who voted Labour last time but might not this time. Even if the majority of voters would be happier with (say) a state-owned rail system, she can safely ignore their wishes because the floater is king, and polls have told her what they want.
Tuesday, 1 November 2016
Weighted Votes: The Utilitarian case for democracy
This might seem an odd place to start; surely we’re all in favour of democracy? Why do we need to hear arguments in favour of it?
Well, one of the problems I have come across in conversations about voting reform is that people don’t seem that clear on exactly why democracy is a good thing. And the result is that they argue for supposed “improvements” that are actually irrelevant.
So here’s my proposal. Feel free to disagree, but at least it will be a starting point.
The Utilitarian Case for Democracy
In 1789, Jeremy Bentham published An Introduction to the Principles of Morals and Legislation, in which he argued for a principle now known as utilitarianism.
This says that the right course of action is the one that produces the greatest sum total of happiness (also called utility) across the population. It had many logical problems - for example, with the idea of adding up different sorts of happiness experienced by different people – and came in for criticism from many people, not least the Church. In 1863 John Stuart Mill published Utilitarianism, which developed Bentham’s theory and closed some of its loopholes. Still, it had a number of critical problems, especially in how to calculate what the sum total of happiness would be for different courses of action.
As a rough guide for judging different policies, the principle of utility is fine, but we should forget any idea of calculating the precise numerical value of the utility of a policy. However, there is a procedure we can use; we can ask people how happy they would be with a policy. So that’s what we do, and we call this an “election”.
Well, one of the problems I have come across in conversations about voting reform is that people don’t seem that clear on exactly why democracy is a good thing. And the result is that they argue for supposed “improvements” that are actually irrelevant.
So here’s my proposal. Feel free to disagree, but at least it will be a starting point.
The Utilitarian Case for Democracy
In 1789, Jeremy Bentham published An Introduction to the Principles of Morals and Legislation, in which he argued for a principle now known as utilitarianism.
This says that the right course of action is the one that produces the greatest sum total of happiness (also called utility) across the population. It had many logical problems - for example, with the idea of adding up different sorts of happiness experienced by different people – and came in for criticism from many people, not least the Church. In 1863 John Stuart Mill published Utilitarianism, which developed Bentham’s theory and closed some of its loopholes. Still, it had a number of critical problems, especially in how to calculate what the sum total of happiness would be for different courses of action.
As a rough guide for judging different policies, the principle of utility is fine, but we should forget any idea of calculating the precise numerical value of the utility of a policy. However, there is a procedure we can use; we can ask people how happy they would be with a policy. So that’s what we do, and we call this an “election”.
I say, perhaps “I will abolish the bedroom tax!” Do you think you’d be happy with that? Vote for me.
Oh, sure, elections are only a crude measure of our expected satisfaction with the direction our legislators want to take, and there are different ways of organising it, different levels of detail, different ways of sharing out the jobs of government. But it can be a pretty effective process for – and this is the crucial point - getting lawmakers to take our utility into account.
As Bentham says:
Countries that don’t have democracy tend to be horrible places to live, because the government is maximising the utility of some special group. That might be the army, or the landowners, or employers, or the Mafia, or banks, or the clergy – whoever they need, to keep them in power. In this country, some people seem to think that the right form of the Principle of Utility is the one that calls for the greatest happiness of the greatest number of Oxbridge graduates, or Chief Constables, or billionaires, or Murdoch editors, or swing voters in marginal seats. I don’t. I want the Utility Function to maximise the sum of everyone’s happiness.
Now, once you’ve decided to go for the Universal Principle of Utility – and therefore to let as many people as possible rate you on your performance – there’s still the question of the consequence of that rating. On Trip Advisor, the main consequence of poor scores is that the hotel owner stays in charge but gets less trade. In UK politics, the consequence is that the Prime Minister loses her job. This is intended to incentivise her to consider our (perceived) utility – though it might also incentivise her to lie to us. Elections alone are no guarantee of honesty in politics. That's why we also have laws and journalists.
Oh, sure, elections are only a crude measure of our expected satisfaction with the direction our legislators want to take, and there are different ways of organising it, different levels of detail, different ways of sharing out the jobs of government. But it can be a pretty effective process for – and this is the crucial point - getting lawmakers to take our utility into account.
As Bentham says:
The dictates of the desire of amity, it is plain, will approach nearer to a coincidence with those of the love of reputation, and thence with those of utility, in proportion, cæteris paribus, to the number of the persons whose amity a man has occasion to desire: and hence it is, for example, that an English member of parliament, with all his own weaknesses, and all the follies of the people whose amity he has to cultivate, is probably, in general, a better character than the secretary of a visier at Constantinople, or of a naïb in Indostan. (An Introduction to the Principles of Morals and Legislation, p.77)Which is to say, that as a parliamentarian has occasion to desire the amity of many voters, so he will consult their desires.
Countries that don’t have democracy tend to be horrible places to live, because the government is maximising the utility of some special group. That might be the army, or the landowners, or employers, or the Mafia, or banks, or the clergy – whoever they need, to keep them in power. In this country, some people seem to think that the right form of the Principle of Utility is the one that calls for the greatest happiness of the greatest number of Oxbridge graduates, or Chief Constables, or billionaires, or Murdoch editors, or swing voters in marginal seats. I don’t. I want the Utility Function to maximise the sum of everyone’s happiness.
Now, once you’ve decided to go for the Universal Principle of Utility – and therefore to let as many people as possible rate you on your performance – there’s still the question of the consequence of that rating. On Trip Advisor, the main consequence of poor scores is that the hotel owner stays in charge but gets less trade. In UK politics, the consequence is that the Prime Minister loses her job. This is intended to incentivise her to consider our (perceived) utility – though it might also incentivise her to lie to us. Elections alone are no guarantee of honesty in politics. That's why we also have laws and journalists.
Monday, 31 October 2016
Weighted Votes: A summary of the arguments
The purpose of democracy is to make our legislators focus on the happiness of the greatest number. The better the electoral system does that, the better it is.
Plurality voting systems, such as our current one, distort this process by making a small number of voters’ votes crucial and the vast majority irrelevant.
Weighted Votes is a proposal for a system that corrects that distortion. It makes it worth a party’s while to campaign, even in unwinnable seats. It makes it worth your while to vote for your preferred party, even in unwinnable seats. It corrects for unequal sizes of constituencies, it discourages tactical voting in favour of honest voting, it can retain the constituency link while avoiding having two classes of MP, and it creates near-perfect proportionality of representation.
The way it works is that, in an election, all votes cast for each party – whether in seats they won or seats they didn’t win – are counted and divided by the number of seats that party has; the result is called the party’s weighting. Their MPs, when they vote in a parliamentary division, then have their vote counted with that weight.
For example, in the 2015 election the Conservatives got 331 seats for 11,334,920 votes, so each Conservative MP would get a vote worth 11,334,920 divided by 331, or 34,244, to cast in the division lobby. Likewise, every Labour MP's vote would be worth 40,290, SNP 25,972 and LibDem 301,986.
This can be combined with any electoral system – STV, AV+, pure list systems, Score Voting, whatever – as those systems are about who gets chosen to be a representative, while Weighted Votes is about how influential that representative is.
I am going to lay out my arguments for WV. I will start with some philosophical underpinnings, because I think decisions about electoral systems need to to be based on things like what democracy is for and why voter equality matters.
Once I’ve done that, I will then describe the system in more detail, along with what its good effects would be.
Plurality voting systems, such as our current one, distort this process by making a small number of voters’ votes crucial and the vast majority irrelevant.
Weighted Votes is a proposal for a system that corrects that distortion. It makes it worth a party’s while to campaign, even in unwinnable seats. It makes it worth your while to vote for your preferred party, even in unwinnable seats. It corrects for unequal sizes of constituencies, it discourages tactical voting in favour of honest voting, it can retain the constituency link while avoiding having two classes of MP, and it creates near-perfect proportionality of representation.
The way it works is that, in an election, all votes cast for each party – whether in seats they won or seats they didn’t win – are counted and divided by the number of seats that party has; the result is called the party’s weighting. Their MPs, when they vote in a parliamentary division, then have their vote counted with that weight.
For example, in the 2015 election the Conservatives got 331 seats for 11,334,920 votes, so each Conservative MP would get a vote worth 11,334,920 divided by 331, or 34,244, to cast in the division lobby. Likewise, every Labour MP's vote would be worth 40,290, SNP 25,972 and LibDem 301,986.
This can be combined with any electoral system – STV, AV+, pure list systems, Score Voting, whatever – as those systems are about who gets chosen to be a representative, while Weighted Votes is about how influential that representative is.
I am going to lay out my arguments for WV. I will start with some philosophical underpinnings, because I think decisions about electoral systems need to to be based on things like what democracy is for and why voter equality matters.
Once I’ve done that, I will then describe the system in more detail, along with what its good effects would be.
Sunday, 30 October 2016
On Weighted Voting
The purpose of this modest series is to propose a form of Proportional Representation that has not had a great deal of attention, but which I think has so many advantages that we should be seriously advocating it above all others.
I will start by analysing why we want democracy at all, and then why proportionality is the right way to do it. Then, on that basis, I will describe the Weighted Votes (WV) system, explain its benefits, and suggest how it might be implemented.
In brief, the Weighted Votes system applies to votes in parliament. Rather than making the number of MPs for a party reflect their share of the vote in the country, it makes their parliamentary voting strength equal their votes in the country.
I will put up a new post at noon every day from tomorrow, Monday 31st October until Saturday 5th November. The following week I will start to post a series of objections to WV, along with my responses to them.
I will start by analysing why we want democracy at all, and then why proportionality is the right way to do it. Then, on that basis, I will describe the Weighted Votes (WV) system, explain its benefits, and suggest how it might be implemented.
In brief, the Weighted Votes system applies to votes in parliament. Rather than making the number of MPs for a party reflect their share of the vote in the country, it makes their parliamentary voting strength equal their votes in the country.
I will put up a new post at noon every day from tomorrow, Monday 31st October until Saturday 5th November. The following week I will start to post a series of objections to WV, along with my responses to them.
Subscribe to:
Posts (Atom)