Saturday 11 December 2010

Arguments against reform: would it be too inconvenient to count the votes?

John Rentoul has done an excellently thorough demolition job on the reasons not to choose AV being propagated by Jonathan Isaby of ConservativeHome, though they would appear to represent the collective wisdom of the No campaign and the anti-change Tory-Labour alliance of MPs who put their name to them.

A number of the arguments are clearly disingenuous and contradictory. To an extent, such opportunism (arguing against AV as not proportional, and then making anti-PR arguments against AV)is par for the course.

But there is one argument that is so absurd that No campaigners should face ridicule and lose face as democrats should they continue to use it.

It is to his enormous credit that Jonathan Isaby mounted an excellent and successful campaign - which I supported - to count votes on election night.

Clearly, councils had alighted on the idea of not paying people to count votes after the polls closed at 10pm because it would save some money. Opponents of the change did not argue that it would cost less to count through the night - it wouldn't - but simply set out why there were important democratic reasons to not make a relatively trivial cost saving.

As Isaby blogged:

At first we met with great intransigence from a considerable number of returning officers - many of whom seemed to regard the counting of the votes as something of an inconvenience.

Now he is offering an "argument" against AV of precisely that form:

9. AV IS EXPENSIVE: Under AV we won’t be able to count ballot papers by hand on election night if we want a quick, decisive election result. Local councils will have to purchase electronic counting machines that are very expensive and prone to malfunction.

This is implausible, in terms of the expensive "electronic counting machines".

But it is also a completely absurd argument in a democratic country. The difference with the 'keep election night' campaign is that the issue at stake would now not be when votes would be counted, but whether votes should be counted at all, were this to be become cumbersome and inconvenient.

Reasonable people can take different views about issues such as whether only first preferences should count.

A paradigm FPTP v AV test case is where candidate A would beat candidate B 41-39-20 under FPTP but where A would lose to B 45-55 in a two-candidate race, or if voters expressed transferable preferences.

One side (FPTP) says candidate A should win because "only first preferences should count": they want to give greater weight to the intensity of voter preferences. If there are 15 candidates, and the leading one has 11%, the electorate have made their choice. It is not considered legitimate to find out whether the winner, or another candidate, has majority support.

The other side (AV) says the candidate with the most support among the electorate should win: they want the breadth of support to count, so that the question is which of A or B gets closest to 50% of everyone.

Whatever view you take of this, what matters is the principled argument about which is the better way to work out who the voters would prefer to be their MP. To argue for the "cost" issue as having weight is a total absurdity.

Unless some people think like this:

"I don't think first-past-the-post is as good.

I'd rather we did count all of the voters' preferences. But I don't think we can afford to do it now. Maybe we could bring in AV when the economy has returned to growth.

On balance, what with all of the other costs of a General Election, I just don't think we can right now afford the staffing costs of a count which might last six hours rather than four.

For me, that's the decisive factor. I'll voting for the system where it is less expensive to count the votes".

After all, it would definitely be quicker and cheaper - under first-past-the-post - to count a random sample of half of the votes, and then project the result, on the grounds of "deficit reduction" savings.

Academics may well be able to show that we would get the same result in well over 99% of cases if we counted 30,000 out of 60,000 votes. It might be that counting 10% of the votes would do pretty well if there was accurate randomisation. (This is how US networks can "call" Congressional races as returns come in).

But this is not an argument even the most hawkish of anti-spending advocates would ever offer. The objection would be one of principle: count all of the votes, as a rather more fundamental issue of respect for the voters than the question of whether those counting the votes should stay up at night.

The only argument here is what we should think of as "votes" - not whether we should then count those that we decide should count if it seems inconvenient or costly.

Indeed "it would be too expensive to count votes" is a shocking and offensive argument if made by those in what remains one of the world's largest economies, especially if they ever again want to (rightly) call for proper democratic practices in much poorer countries.

Mugabe's Zimbabwe is much more cash-strapped than Britain. What would ConservativeHome think if ZANU-PF were now to take inspiration from the No to AV campaign's logic, for example, suggesting they can't afford paid staff at polling stations, or to count votes? (I am sure it would definitely save public money to ask the political parties to supply volunteers instead).

It is legitimate for the No to AV campaign to continue to try and argue that "second preferences shouldn't count" but they have to win that argument on its merits. The "cost" argument against counting them should be considered absolutely worthless as supporting evidence.

Anybody repeating the "it would be too expensive to count the votes" argument should be ashamed to call themselves a democrat.


Anonymous said...

Given the fact that half the Yes camp is trying, like some dodgy second-hand car dealer, to sell us something they know is a hunk of junk I find it a bit rich of you to get on your moral high horse about the 'opportunism' of the No camp.

Having said all that I actually think its the Yes camp that should be ashamed of calling themselves democrats. I agree money should be no object to democracy but then again AV is less democratic than FPTP so I think its legitimate to ask why we should lash out more money on it.

The cost argument really boils down to that. As a proponent of the Yes side you don't agree so the cost doesn't bother you, equally naturally proponents of the No camp are unlikely to want to lash out extra money on something they view as being rubbish. In that sense I agree money isnt the real issue...the real issue is the anti-democratic nature of AV and the pack of lies that is being used to sell it too us.

Richard Shaw said...


Your comment makes a couple of suggestions which are really quite baseless.

Firstly claiming as fact that "the Yes camp is trying... to sell something they know is a hunk of junk", implying dishonesty on their part and therefore FPTP must be 'better'.

Even if some of the Yes2Av folks think AV isn't the best system in the world in comparison to AV+, STV or more proportional systems they regard it as much better than FPTP. Like AV in practice: if you can't have your preferred system (e.g. STV) then you may still rank AV above FPTP.

Your next point: AV is less democratic than FPTP.

How does that work? How is ensuring the winning candidate is the one with a majority of votes less democratic than a system where the winner is the candidate with the largest minority of votes? How is a system where every vote counts less democratic than one where if you don't vote for the top two candidates your vote is wasted - having the same effect as having not voted at all?

Before you next accuse anyone of lying I suggest you follow some classic advice from Wikipedia: [Citation Needed].