Once more on myth and reality of STV

In their short version of why voters should reject STV, No STV claim:

“STV is complicated, confusing, prone to errors and delay, it reduces local accountability, increases the size of ridings, allows MLAs to avoid direct accountability for their decisions, increases party control and allows special interests to dominate party nominations.”

The first three claims above are matters of fact -- so are they true?  The answers are no, no and no. 

Comparative studies of election administration demonstrate that voters have no problem voting with STV (Farrell 2001).  Nor is STV ‘prone to errors and delay.’ 

In some cases, STV-using jurisdictions have a preliminary result before bedtime on election night, while in others they come the next day. When we turn to more systematic research about how STV works, the No STV vague claims cannot be sustained. 

One recent academic study ranked STV-using Ireland 14th out of 98 countries in terms of the quality of its democracy (Campbell and Polzbauer 2008), while the Economist magazine ranked it 12th out of 1563 countries in its survey of the quality of worldwide democracies, with a particular focus on election transparency. Those that know anything about international elections seem to think the Irish STV system is working just fine. 

What of the other claims above?  Does STV increase party control over nominations, allow special interests to dominate, allow one candidate and their supporters to dominate multi-member nomination meetings or diminish local accountability? 

Here too there is considerable research on what actually happens in jurisdictions that use STV and it does not concur with the No STV version of the events.  Experts on Irish politics note that nomination contests are local affairs that exhibit keen competition between different candidates.  The central party is brought in only to negotiate how to divide up the multi-member ridings so that the different candidates of the same party won’t step on each other’s toes. 

Ireland also bucks the trend in western countries of moving away from direct voter contact, with over half of voters claiming to have had a face-to-face visit with a candidate -- that twice as many than in the UK in the same period (Marsh 2004).  This local focus is assumed as important by both voters and candidates in the Irish system (Marsh 2007).  Indeed, many scholars complain that the Irish politicians are too locally focused!  It would appear that the evidence on local politics under STV directly counters the unsubstantiated claims of No STV. 

In their recent TV ad, No STV boldly claim that the ‘downtown’ will dominate representation in multi-member STV ridings, and on their website they claim that more populous urban areas will dominate the rest in the cities.  Though, again, they provide no evidence for this assertion.  Of course, that’s probably because they can’t -- STV doesn’t work that way. 

No STV are trying to use people’s knowledge of the unrepresentative results of the ‘at large’ systems used in most B.C. municipalities for civic elections.  That system does allow one area of town to dominate, usually the wealthier side.  But the logic of STV militates against such results because voters cannot be so easily marginalized or ignored as they can under multi-member plurality or ‘at large.’  Because STV increases the number of votes that can have an impact on the results, politicians would be foolish to ignore pockets of voting support outside of the downtowns or the less populated parts of urban ridings. 

The math working here is simple: a FPTP election typically sees anywhere from 40-60 per cent of the votes wasted -- a lot of those voters can just be ignored.  But under STV 80 to 90 per cent of the votes will help elect someone, so any potential voters are ignored at the politicians’ peril.

Spurious associations

One of the ‘a-ha’ issues that the No STV campaign have struck upon is the idea that there is something funny going on with the counting under STV.

On the face of it, this claim is pretty amazing.  I mean, B.C.’s No STV seem to think they have discovered something rotten at the root of the STV system that no one has found before.  The claim is truly incredible if we recognize that STV has been used in Ireland for nearly a century and we are supposed to believe that no one there -- from the public, to the politicians, to the media, to the election officers -- discovered what our home grown experts have!  We are veering dangerously close to conspiracy theory here. 

Nonetheless, let’s examine the claims.  No STV highlight three issues that they think are problematic with STV: that voters cannot control the fractional value of their vote transfers, that the different thresholds that will be used in the multi-member ridings are unfair and that STV’s counting process is unsound.  We’ll take each in turn.

Under STV, a voter has one vote, though that one vote may help a number of people get elected.  If your first preference has a lot of support, more than they need to gain the quota, only part of your vote will stay with them while a fraction of your single vote will go on to help your second preference, and so on. 

But No STV complain that the voter cannot control how much of the fraction goes where, that this will depend on what other voters do.  As one glib characterization by No STV puts it, it is like ordering a steak and beer but allowing your neighbor to decide how much you’ll actually get. 

The only problem with this claim is that it actually applies to every single voting system ever designed or known to humankind.  Think of FPTP -- the value of your ballot in affecting the outcome in a local race is crucially dependent on what other voters do.  This reflects a basic and unavoidable truth about representation and elections: though voting is an individual act, its results are necessarily collective.  In fact, people using our existing system consider what other voters might do all the time when they strategically calculate who to give their support to (think potential Green party voters).  So No STV is passing off a basic structural fact of all voting systems as a problem for STV. Nice trick.

No STV also complain that the quota for election will not be the same in all ridings, that seven member Victoria will have a quota of about 12.5 per cent while two member northern ridings will face about 33 per cent.  This is a funny claim coming from the group telling voters to support our existing FPTP system on election day. 

While the proposed STV system for BC would feature six different set thresholds, the existing FPTP system has effectively 85 different thresholds for election.  Because FPTP is winner take all, candidates can win with all different thresholds, depending on the competitive dynamic in any given riding -- could be 50 per cent, could be 40 per cent, could be less or more.  If STV is unfair then by the No STV logic the existing FPTP system they are supporting is super unfair. 

Another issue raised by No STV involves alleged counting problems with STV.  They feature a newspaper editorial on their website that calls STV counting ‘a shell game,’ clearly inferring that there is something bogus going on. 

Math, logic and STV

One No STV campaigner, on the main site and his own, cites the problem of ‘non-monotonicity’ or the mathematical possibility that a voting system may not work in a logical way, claiming it occurs all the time under STV. 

Without delving into the technical details of this complex debate, suffice to say that the key claim is that under STV there may be cases where voting for your first choice second might actually better help them to win than marking them first. 

If true, it would be a very serious strike against the claims that STV allows voters to avoid strategic voting and manipulation.  But what we have here is the classic problem of ‘a little knowledge is dangerous thing.’  The non-monotonicity debate is an obscure issue discussed largely by mathematicians in theoretical terms.  Researchers studying real elections can’t find any examples of it occurring (Allard 1995; Gallagher 2005).  So No STV takes something that exists only in theory, for which research has found no practical examples, and manages to turn it into something that happens all the time.  Nice research! 

Insincere arguments

I’m not sure any of the No STV arguments can be characterized as sincere but there are a few that stand out as obviously false.  One is that No STV takes no position on any other voting system.  Given the complaints they’ve made about STV, particularly as concerns local representation, it is pretty clear that any other PR system would also fail to pass muster. 

For instance, any mixed-member proportional (MMP) system, as recommended in Ontario in 2007, would come under the same kinds of attacks for making ridings bigger, particularly in northern areas, etc.  What people need to understand is that the No STV is the Yes FPTP campaign, though they try to pretend otherwise. 

Another insincere argument is their feeble effort to counter the ‘wasted vote’ idea that STV supporters talk about.  They claim ending wasted votes would mean that every voter would have to see their chosen candidate win, something they claim is akin to betting on the every horse in the race so as not to lose.  But this is a false analogy -- in fact it reproduces the ‘winner take all logic’ that STV supporters are criticizing. 

The wasted vote argument is simply why shouldn’t voters be able to make common cause with others, if they can.  Our single member ridings orphan typically half the voters in privileging geography over everything else.  If 9 per cent of voters want Greens as elected members, who are we to say they can’t have them?  Why are such votes wasted, simply because they are cast over a broad geographic area?  These are serious questions about representation that are simply dismissed by No STV.

But probably the most insincere argument posited by the No STV forces is the idea that if voters cannot become the equivalent of an STV returning officer, they should not vote for the system.  This is arguably the most patently ridiculous claim they have made. 

If demonstrated expertise in the details of policy were the criteria for public involvement in policy debates, there would be no public involvement in policy debates.  It’s like saying if you’re not a nuclear physicist, you don’t have a right to an opinion on nuclear power.

On policy, the public takes the big picture, particularly in terms of results -- that is their way into the debates.  And that is what they should do with STV as well -- think about the broad results the system might produce and consider whether they like the sound of them. 

What is so unbelievable about the No STV position here is that they are a group with considerable experience with public engagement on policy -- as a former communications director for a Premier, a special advisor to a Premier, a former deputy minister, and a number of former politicians.  They know how policy consultation works, how complicated policy is, and how limited the public’s purchase is on its details. 

Yet they peddle this simplistic approach to assessing STV.  It is simply not believable, given all their experience.

Dennis Pilon teaches politics at the University of Victoria and is the author of The Politics of Voting: Reforming Canada's Electoral System.

Comments