Liam Weeks*
Since 2011 may be the last general election by PR-STV in Ireland (if some of the political parties get their way), it is worth looking at the value of the transferable nature of the vote. This is one particular feature that makes STV so voter-friendly and yet its value is sometimes dismissed by media commentators.
For example, there appears to be a myth that transfers don’t matter to candidates – that they need to be ‘in the frame’ after the first count, that is, in a winning position (top 3 places in a 3-seat constituency, top 4 in a 4-seat constituency and so on).
On RTE’s News at One yesterday Seán Donnelly said that only 11 candidates ‘outside the frame’ managed to win seats in 2011. The inference was that transfers affected only these 11 seats and not the 154 others.
This misses a crucial point. Just because the decisive places in the contest were not altered between the first and final count does not mean that transfers don’t matter. It is candidates’ ability to attract transfers that ensures they don’t slip out of the frame. For example, only 22 candidates were elected on first preferences alone. 14 others were within 0.1 of a quota after the first count. If we assume that these 36 candidates would have got elected without any further transfers, this means that 129 candidates needed transfers to win a seat. If they did not receive any they would all have fallen out of the frame.
The importance of transfers can best be observed by considering the cases of Fianna Fáil and Sinn Féin candidates, who were deemed ‘transfer-toxic’, but for different reasons. Sinn Féin had 5 candidates one place outside the frame and 2 that were 2 places off. Two candidates (John Brady and Larry O’Toole) fell out of the frame between the first and final counts.
12 Fianna Fáil candidates were one place outside the frame after the 1st count, with 9 candidates 2 places off. In addition, two candidates (Chris Andrews and Brendan Kenneally) were both inside the frame but slipped out come the final count. None of these 23 candidates could win a seat because they could not accrue enough transfers (even though some had a running mate). There were two exceptions to this pattern within Fianna Fáil. Both John Browne and Robert Troy were out of the frame after the first count but got back in via transfers (Browne had one running mate and Troy two). Troy’s feat was all the more remarkable considering he had barely over one-third of a quota on the first count.
Transfers do matter. Where at previous elections some of the 23 Fianna Fáil candidates might have attracted enough transfers to win a seat, this time round the electorate let it be known that Fianna Fáil was not their second preferred party (or even their third). Eliciting such information from voters is vital in the composition of a Dáil that most accurately reflects the will of the nation. This is something that should be considered when electoral reform is discussed in the new Dáil.
*Liam Weeks is an IRCHSS CARA Fellow at Macquarie University, Sydney. His research is funded by the Irish Research Council for the Humanities and Social Sciences with co-funding from the European Commission.
Irish Politics Forum 
I’m sure the political scientists here will be able to come up with many, but I think PR-STV served Irish voters better than most other systems in terms of what voters wished to express and to achieve. I suspect most voters, when the results were in, considered what they had achieved, like the Creator in Genesis, “and it was good”.
Let’s hope this diminishes this futile focus on the voting system and encourages a focus on the institutions and procedures of political governance.
Liam,
I would argue that the political parties – especially the three ‘establishment’ ones – have done their utmost over the years to shape PRSTV to suit thier own purposes. The ‘Tullymander’ of the 1970s is an obvious example, as were the FF inspired referenda and various more recent proposals to abolish PRSTV and substitute it with FPTP, which, had they been accepted by the people, would have copperfastened the likelihood of ‘permanent’ seat majorities for FF, provided of course they didn’t slip below a critical percentage of popular support. We still have a plethora of three seat constituencies, especially in the capital and other urban areas, which potentially distort the intention of an STV. It’s worth looking at the number of times in successive contests that the first three candidates in any three seat constituency are the ones who most often make it through in the end. The ‘strong’ party candidate in any threeseater has an incentive to either quota squat, or use any running mate purely as a ‘sweeper’, in order to maintain their own position.
Three seaters deliver a PRSTV FPTP result. It’s only when vote of parties collapses, such as FF in this election, or there is a national surge in favour of a particular party like Labour in 1992, that a significant change in personnel in these constituencies may occur. As the ‘Tullymander’ exposed, a swing away from one establishment party in favour of another will do the trick nicely as well, but generally it makes it very difficult for small or new parties to break into the running for seats in any constituency that does not have four or five seats.
I hope you are wrong about this being the last election to use PRSTV. That too will be a decision of the people. For all its flaws, PRSTV delivers a more just representation of the will of the people than any of the alternatives proposed by the parties as part of their political reform packages. Whenever I hear political parties droning on about how changes in our electoral system are necessary ‘in the interests of the country’ I’m reminded of those little green-brained invaders in ‘Mars Attacks’ whose every claim of ‘we come in peace’ was to be read as the exact opposite.
PR-STV has many wonderful characteristics. Few other systems allow voters to articulate their preferences in such an expressive manner. Voting is not forced to be constrained along party lines. And it is a very fair and proportionate system when constituency sizes are sufficiently large. But, as has been pointed out already, there are still way too many smaller constituencies. And it has been a long time since there were seven (and even nine) seat PR-STV constituencies in this country.
But PR-STV also has its flaws. It is quite similar in terms of proportionality and openness to open list PR systems. There are differences however. The ability to transfer is its principal distinguishing feature. I think this, though, can be a double edged sword. On the plus side, it does give the voter a very fine grained control over preferences.
And list PR systems are far more blunt. A voter can only choose a party and then maybe an individual within that party. But this does force the voter to explicitly pin his party colours to the mast. A small party can carve out a niche for itself, and perhaps develop a distinctive set of policies that cater to as little as 5-10% of the population. It doesn’t necessarily have to worry about voters that would never cast a ballot for it anyway.
Contrast that with PR-STV, where a party may indeed have a core vote, but also relies on transfers from all over the place. This doesn’t encourage policy specialization. It incentivizes parties and candidates to have broad appeal. They don’t want to upset the supporters of other parties too much because they may get transfers down the line from these same voters. This encourages parties to be “all things to all men”, innocuous and inoffensive, with a strong attraction to the centre ground. Not entirely a bad thing maybe. But it’s perhaps not a coincidence that Irish political life has been largely dominated for eighty years by a populist non-ideological middle-of-the-road political party, with the other two major parties being just a small stone’s throw away on either side of the political spectrum. Maybe that’s due to a conservative Irish psyche. Maybe it’s because of the electoral system. Or IMO due to both these factors with one reinforcing the other.
I wouldn’t like to see the end of PR-STV. It has many attractions and is quite an elegant system. But I wouldn’t be overly enamoured with it either. All of the voting systems seem to have their own strengths and flaws. There’s no perfect electoral system. But that being the case it might be sensible to use more than one system here. Maybe one system for the Seanad (large constituency PR-STV could be quite attractive there given its friendliness to non-party aligned candidates). Perhaps another (or maybe more than one) different and contrasting electoral system for the Dáil. That would be quite messy and something of a hodgepodge. But it might also be more robust and less susceptible to the particular peculiarities of any one voting system.
I don’t think changing the electoral system should be a priority. Far more pressing changes need to be made elsewhere in our political system. And even within PR-STV, increasing constituency sizes could be an easily made change to improve proportionality. PR-STV itself could be tweaked somewhat. But I still find myself reluctant to dismiss any changes to the electoral system as being an out and out red herring.
Not an out and out red herring, as you put it, but it could be used to distract attention from pressing reforms – and to kick the whole reform process into the long grass. In addition, having just used the system so precisely to achieve their objectives, I suspect most voters would look askance at any package of reforms that included changes to it.
@Liam, you’re right. I heard Donnelly say that and thought exactly that. He forgets that proportion of transfers and proportion of first preference vote are correlated. This doesn’t mean that they don’t matter.
@Finbar, no electoral system is perfect, but as you say, PR-STV has a certain elegance. I think you’re right to think about the constituency size. The small number of seats (District Magnitude in polsci speak) reduces its proportionality significantly. And the variation in the DM also matters because it means that in some constituencies certain parties have a chance but in others no chance.
I think a debate about district magnitude may be in order. Is 5 as in Malta an appropriate measure? perhaps 6 or 7? Maybe beyond that would be too much.
In terms of increasing our district magnitude; I think it is interesting that while it appears that we have a low effective threshold to representation (independent candidates may achieve representation with as little as 0.2% of the national vote share), niche parties such as the ICP, Coir, Eirigi, Fathers Rights, The Socialist Party, etc etc that tend to polarize the electorate tend to be damaged by our combination of low district magnitude (because this debilitates their ability to target demographics) and transfers.
Agreed. Perhaps something that should be looked at separately. It’s not a clearcut issue, especially when there’s a whole host of more straightforward reforms with more easily predictable consequences. There’s probably some merit to at least looking at the electoral system, but not if it’s just used as an opportunity to kick the can further down the road for other reforms.
if we had the German system – half single seat constituencies and a top up list- the result of the election, if one assumes maybe half a dozen independents or other swould have won in single seat constituencies, would have been FG 70 Lab 37 FF 34 SF 19 Others 6
Those who favor the STV have ignored how each party’s caaniddtes will be selected. It appears that the STV means party officials or the party leader will determine a hierarchy of caaniddtes to represent an area. This would mean more “top down” politics by political parties and less “grass roots” democracy. For real democratic reform we should enable greater scope for “citizens initiatives”.
Very interesting-I think Troy was successfulwith transfers because he was not part of “establishment” FF and was therefore more transfer-friendly