June 2022 Projections Update

Seat Map

Three polls this month, but Ireland Thinks still don’t publish breakdowns, so our constituency figures rely on RedC and B&A. General trends continue, as Sinn Féin have continued to push up their ceiling slightly, once again equalling all-time highs. Really the only thing keeping Fine Gael in touch right now is Dublin, where they remain within 4.2 points of SF. Elsewhere, the gaps are immense – 18.9% in Connacht-Ulster, 19.7% in Munster and 17% in Rest of Leinster. Fianna Fáil aren’t faring much better; the seats total currently disguises poor underlying numbers which are very slowly, but steadily, declining – this month represents their lowest national total (18.6%) since last November.

On the smaller parties, hard to know what to make of the Social Democrats – strong showing in RedC counterbalanced by a dismal one in B&A. Even if RedC are right, they still don’t have obvious pick-up targets, but they pulled off some surprises last time, so let’s see. The Greens are in a doom spiral at this point; unless things improve I expect their projection to hit zero again in the next month or two.

Labour are in a weird spot – doing well in Dublin, but really struggling everywhere else.

While this has left their national total pretty much back where it was in January, it means that all of their seats outside of Dublin will be hard to hold. But on the flip side, they now have the potential to be competitive in a number of places where they don’t hold seats. Dublin Rathdown might be their best shot at a pickup, but they will also be looking at being viable Central, South-Central and South-West – though all of these are still very much outside chances right now.

PBP/Solidarity are the same as ever – highly geographically concentrated support with no clear sign of a breakthrough; their main challenge on these numbers is defending Cork North Central and Dublin Mid-West. Aontú’s numbers in Connacht-Ulster may warrant attention, but RedC are only pollster showing this at the moment. Keep an eye on it – again they have no obvious pick up opportunities, but if this is anything more than an outlier this month, they could make an outside run at, say, Cavan-Monaghan. Independents had a good month, as we’ll see in the seat changes, and as said many times before, I expect them to outperform the model.

There’s not much else to discuss in depth, so I’m going to cover off two fairly technical things – one related to the census and potential boundary changes, one related to the DBS by-election – and then one brief issue polling matter. If you would prefer to skip straight to the seat changes , click here.

Accounting for the census and future boundary changes

With the publication of preliminary census results, the day I have been dreading is approaching – there will be a constituency commission, there will be around 11 new Dáil seats and, based on those initial numbers it is likely that every single constituency will need to be redrawn to some extent. This is going to have a significant impact on the projections, as the 2020 base vote figures used to calculate swing will no longer be accurate.

So I want to talk a little bit about how I plan to account for this, and I have an illustrative example (albeit in reverse) stemming from a discussion with the Tortoise Shack’s Tony Groves. We’ll cover that later.

Firstly, thanks to the brilliant work of Dr. Ian Richardson, we have FPV %s for every Electoral Division in Ireland, and turnout % for nearly all of them (Limerick City, for example, is missing). All we need to do is cross-reference these with polling schemes and notices of poll and we can tell exactly how many votes we need to add or subtract from the 2020 baseline. There may be some challenges obtaining the schemes and so on, but this can be worked on.

This creates a process that while somewhat onerous in terms of data collation, is at least mathematically straightforward. For example, let’s imagine that Kimmage C moves from Dublin Bay South to Dublin South Central. From the polling scheme we know in 2020 that there were 1,934 registered voters in this ED, and turnout was 46.2%, so 894 voters. Of those, 29.08% voted for SF. So we would subtract 260 votes from SF in DBS and add that to the total in DSC, and do the same for other parties, and then apply the existing swing calculations.

As above, this is a fairly simple calculation and while it’s not 1:1 perfect as it can’t factor in the effect of different candidates and so on, I think it’s as good as we can get. The challenge of course will come from the sheer number of EDs likely involved given the scope of redraws we will likely be looking at.

Now, I mentioned above that I could illustrate this, so I will. The aforementioned conversation with Tony was related to the redraw of Dublin Central and Dublin North-West in 2020. Four EDs from DNW were moved to DC (along with one from DBN), and the calculation here was to work out what would have happened in 2020 if these EDs had not been moved, particularly on incumbent Fine Gael TDs Noel Rock and Paschal Donohoe.

Firstly, we use the formula above to work out how many votes were cast in each ED.

Then we take the FPV% in each ED and then use that to calculate exactly how many votes in Dublin Central in 2020 came from the newly added EDs, and subtract it from the 2020 total to give us a measure of what would have happened if Dublin Central had not been redrawn. ^{(note: I have excluded several minor candidates from this screenshot so it will fit, but they were calculated)} Similarly, this number was then added to DNW to get the outcome there if it had kept the four EDs it lost.

Note that the transfers in the above diagram, based on 2020 transfers, should be considered more illustrative and not necessarily what would have happened. Similarly the candidate dynamics of a 3 and 4 seater are quite different. Regardless, this should give an idea of how I will approach projecting the redrawn constituencies once we know what they look like.

Accounting for the by-election in DBS

I’ve mentioned a few times that in the aftermath of the Dublin Bay South by-election, the model isn’t really able to account for this in terms of projecting how Ivana Bacik (LAB) will do in DBS. I wasn’t happy with this state of affairs and spent some time trying to factor it in. Predictably, I ended up with two completely different ways to measure it, I’ll explain these below and why I chose the one I did (which is ultimately somewhat arbitrary, so if you prefer the other one, just imagine a different outcome).

I took the by-elections that had happened between 2011 and 2020 and measured the change between the prior election, the by-election and the subsequent election. I excluded Roscommon-South Leitrim in 2014 (as the constituency was abolished by the next election) and Dublin West in 2011 (because the winner resigned before the next GE). This left 9 by-elections to look at.

There are two ways to look at things. Firstly, measuring how much better or worse the party that won the by-election did in the subsequent GE when compared to the prediction from national swing from the previous election.

To illustrate: in 2016, Joe O’Brien (GP) got 4.6% FPV. He won the 2019 BE with 22.9% FPV, then held the seat in 2020 with 13.2% FPV. The Greens increased their vote from 2016 to 2020 from 2.1% to 7.1%, a c.263% change. This implies O’Brien’s vote should have been 12.1%. As he got 13.2%, this put him at 109.1% of his projected vote.

Repeating this calculation across these 9 BEs, on average the party that won the BE got 108.6% of the projected vote in the next GE. So this would allow us to increase Bacik’s projected vote by a small amount.

The second way is to directly compare the BE vote to the subsequent GE. To use O’Brien as an example again: he got 22.9% in the BE and 13.2% in the GE; that is to say , 57.6% of his BE vote stuck around for the general. Averaged across 9 BEs, we get a figure of 97.9% for winning parties. This would imply that Bacik will keep the bulk of her by-election vote.

I’ve opted to take the former approach over the latter. As above, this choice is somewhat arbitrary, but I’m more comfortable with it for a number of reasons:

• The former method is a bit less swingy than the latter, and takes better account of the effect of national conditions
• Some of these numbers are a bit skewed by parties with multiple candidates; the effect of this seems lesser in the first method
• The second method implies that Bacik will get 29.6% FPV in the next GE. That’s two seat territory and I am not convinced that’s plausible
• Generally speaking I will err on the side of caution in modelling and I’m not sure why this should be different
• I did briefly consider taking an average, but that figure would, in essence, be so far divorced from the underlying factors that it may as well be completely made up, so I decided not to

The upshot of this is that I have Labour in a stronger position than before in DBS, not enough to look secure on current numbers, but now more likely to hold their seat than before.

None of this is perfect of course, and by opting for the latter I may be understating Bacik here. But I believe some account should be made. If someone else has come up with a better way to account for BEs when making projections, please let me know!

Incoherent views on NI

Small note on this, but the last RedC poll asked some questions about Northern Ireland. I’m not an expert on the subject, but the party support breakdown reveals some interesting incoherence among voters in the Republic. While FG, FF and SF voters are broadly in line on a number of issues, the differences are worth discussing.

Firstly, it is noteworthy that Fine Gael voters are more aggressive towards the UK than Sinn Féin voters are on a number of topics – 8% more FG than SF voters think that the UK government is being dishonest about the protocol and 10% more want the EU to sanction the UK, and 8% less think the EU is being too rigid. However, they are much less likely to believe – by 21% – that this makes a stronger case for a United Ireland.

My speculation is that this is an ideological artefact – despite taking a harsher stance towards the UK than SF voters are, FG voters are much less likely to think that this in an argument for UI. This is a bit incoherent to be honest – if the UK misbehaving isn’t an argument for UI, what exactly is it an argument for? – and I wonder how much of it is to do with FG supporters’ strong positive identification with the EU as much as anything else.

Seat Changes

Three constituencies saw changes in June, after we had no movement last month.

Clicking on the linked name for each constituency should jump to the relevant section of this page. Changes on this page indicate changes from May 2022’s projections; changes on constituency pages indicate changes from current composition.

Note: The projections reflect, and always have reflected, most likely outcomes. So if a final seat is more likely for candidate X over Y, the model will show X winning the seat.

This does not mean the scenario where Y wins doesn’t exist, or even is necessarily unlikely (there’s a lot of marginal calls!). It also does not mean that every single “most likely” scenario will come true; statistically that in and of itself is probably not going to happen. This is true from from a simple probability point of view, even if we ignore deficiencies in underlying data. A projected result merely means that the model thinks X winning is the most likely outcome.

• Donegal (IND +1 FG -1)
• Sligo-Leitrim (IND +1, FG -1)
• Wexford (FF +1, LAB -1)

This finally feels right. While it’s not directly related to Joe McHugh’s retirement – something the model can’t account for – I suspect that might be fatal for Fine Gael here. Thomas Pringle (IND) moves back to being narrowly favoured to hold his seat among a strong polling month for Independents.

I have had Pringle ahead once before, but that didn’t feel quite right, but I think his odds of holding his seat increase significantly in a situation where he is competing with a random Fine Gael Councillor rather than the well-established McHugh.

Of course I fully expect that Donegal will not exist as a single constituency once the commission is done, as the population is now at a level where being split into two three-seaters makes sense – which I imagine Sinn Féin will only be delighted about.

This one is still up in the air for reasons discussed before, but within the parameters we have, it remains very closely contested. With Independents polling well, and FG not so much in Connacht-Ulster, this flips back to Marian Harkin. However, it is worth noting that (a) we have no idea if Harkin will actually run again and (b) there were two other Independent candidates who absorbed over 8% of FPV last time as well – if they don’t run again, that will have significant impact.

Labour’s relatively static overall numbers are masking a significant downward shift in Rest of Leinster, offsetting it with increases in Dublin. If this polling reconfiguration holds true, all of their seats outside of the capital suddenly look very vulnerable. While veteran Brendan Howlin is the most likely to hold on, assuming he runs again, he’s still in an uncomfortable position. Right now the polling puts him behind a second FF candidate, albeit only very slightly.

As has been mentioned a few times, his personal vote is likely to be a saving grace and allow him to beat the projections, and it’s also worth noting that he did outperform the Labour trend in 2020 and 2016. It’s possible that he has a higher floor, but trying to definitively determine this isn’t possible. Either way this is a highly competitive constituency that could come down to fine margins, whatever way it goes.

Other notes

• Cavan-Monaghan: I mentioned Aontú suddenly doing well in Connacht-Ulster in RedC, and if there is somewhere they can capitalise, this would be it. On the RPA, they aren’t particularly close yet, but what if they actually did hit that 6% mark provincially, what about then? Also no, but it would be fairly close. So I don’t expect this to change soon at all, but it is interesting.
• Dublin Bay South: I’ve indirectly discussed this a bit above with Bacik, but to put some context behind how close this is right now, a swing of less than 0.1% from the Greens to FF will put Jim O’Callaghan ahead of Eamon Ryan, and if Labour’s support in Dublin goes up by 0.5% and the other two remain static or drop, Bacik becomes favoured to hold her seat. Also if FG can’t find competent candidates, I’m not so sure their second seat will happen.

Thanks so much for reading! Normally this is that awkward bit with the blurb about donations to the website, but I want to say something else this month.

Over the last while, Ireland has witnessed a deliberate, increasingly organised series of attacks on the rights and wellbeing of trans people. This is a concerted attempt to erase their presence in public spaces, remove hard-earned legal protections, prevent healthcare provision, intimidate them into silence and, ultimately, put trans people in physical danger. It is no coincidence that this confected “panic” has been pushed so hard during Pride month, in what should be a month of celebration and rebellion.

Much of this is coming from astroturfed groups backed by shady finances and with open links to far-right groups who are not just transphobic, but also racist, misogynist and anti-semitic. While it is no surprise that this effort has found support among Ireland’s terminally incompetent fascist movement, the real danger is that this is being facilitated by some of the largest platforms in Irish media, who are behaving in a way that is at best willfully ignorant, and at worst, actively complicit.

Do not fall for this shit. “Debates” over people’s humanity and rights never, ever lead anywhere good. Listen to trans people, elevate their voices and provide the support and supports they ask for.

If you are a regular donor, or wish to donate to the website via Patreon or Ko-Fi, I will be matching any donations raised this month and giving the total amount to Transgender Equality Network Ireland.

Alternatively if you don’t trust me (fair), you can donate directly to TENI here.

Trans rights are human rights.