This June, there will be two elections in Ireland – one for the local authorities, and one for the European Parliament. The modelling I’m doing is purely General Election based, so is of relatively minimal utility for either, but we can still take a look at the polls and see if they give us any useful insights. I’ve already taken a look at the local elections (and may revisit it closer to the time, with the caveat that it is incredibly time consuming to look at permutations for all 949 LEAs), so in this post I will take a look at the European elections. We have had a couple of polls on this from Ireland Thinks, the latest of which came in mid-April.
Recent history of European Parliament polling
First though, some context. When we look at the last election in 2019, the best comparison point on national figures is the Red C poll published April 17th, for the vote on May 24th. On a raw vote level it wasn’t great – with Sinn Féin, Fianna Fáil, Fine Gael, Independents and the Green Party all being missed by more than the margin of error. While part of this can be attributed to movements during the last month of the campaign, as we’ll see below, closer polling also had substantial errors. The Greens, it is worth noting, were particularly underestimated here, far beyond a normal polling error.
It is of course worth noting that national level polling will always have limitations – even with the very large European Parliament constituencies. So let’s look at the regional polling from 2019 – we have regional polls published by both Red C and IPSOS, on the 16th and 10th of May respectively. These came much closer to the day of the vote, but the output was a mixed bag; perhaps in part due to the smaller sample sizes used (500-825).
Note that on the fourth chart above, the MoE is variable owing to mixed sample sizes; I have labelled the widest MoE. The spread here is a bit better than on the national poll, but overall there are still significant deviations. Again the underpolling of the Greens is particularly notable, as is the overpolling of Fine Gael. It is worth noting that in almost all cases, parties were consistently over- or under-polled across regions.
This again could be taken as a sign of changing preferences late in the process, or of a systemic polling error; it’s very difficult to say from a small sample. But this does provide some insight into the greater level of error in European polling versus General Election polling. It also should be noted that these errors do not necessarily replicate over time – one should not assume, for example, that because FG were overpolled in 2019, they definitely will be again in 2024.
Finally, a note on what a projection would look like, if a sample of two polls is taken at face value.
- In Dublin, the polls missed Ciarán Cuffe topping the poll for the Greens, although still likely giving a good enough performance for a seat, and would have had Lynn Boylan (SF) narrowly defeating Clare Daly (I4C), even with Daly’s better performance on transfers.
- In MNW, despite a big miss on Saoirse McHugh (GP), a projection would likely return the same result that we got in reality.
- In South, these polls would have indicated a seat for Liadh Ní Riadha (SF) over Grace O’Sullivan (GP); despite underpolling Mick Wallace (I4C), projecting from this poll would still give him a seat.
Of course, a sample of two polls shouldn’t be taken purely at face value, even if they are addressed at a specific constituency. At minimum the MoE should be considered, as well as the potential for outliers or fundamental sampling, weighting or methodology problems. This applies even more acutely for a single, national-level poll. Bear that in mind through the next two sections.
The Ireland Thinks poll
So with that in mind, let’s look at the 2024 poll itself. We don’t have any regional or constituency information, as only national and certain demographic data has been released, but we can still compare that with current polling averages (note: the polling average does not include today’s poll from Opinions because it came out while I was in the middle of writing this):
The first thing that jumps out here is that the results are very closely aligned with current polling, with all bar three groups within 1% of their national polling position. This shouldn’t perhaps be too surprising, if we consider the aforementioned poll in 2019, page 13 has a comparison of general, European and local election support for each party; for the most part GE and EE support is within a few points. But again, bear in mind what we saw above – there ended up being substantial gaps between the national poll and the constituency level polls (and, more crucially, results). Also, as seen above, even relatively small misses at national level can result in apparent surprises in the constituencies.
The most significant differences between this poll and national polling are with Sinn Féin (who are underperforming national polling) and the Greens and Aontú (who are overperforming, albeit by an amount that is within the MoE).
For SF, there could be a number of factors – polling declining faster than the polling average, an outlier poll, or simply that people are less inclined to vote for SF in European elections than in a general. The last one is the most interesting prospect – but we can’t verify that this is the case. For example, in 2019 and 2014, SF did better in the EE than the LE. But this is the lowest they have been in any poll for a long time; it won’t of course be until polling day that we know for sure what the cause is.
While within the MoE, the Greens and Aontú are also worth briefly touching on. European elections have been a significant strong point for the Green Party for a while; they have generally outperformed their results in any other form of election, perhaps because the electorate is more attuned to Green issues when it comes to Europe than domestically. Again while caution must be taken, it is extremely plausible that the Greens do better in the EE than the LE, and better than they will fare in the upcoming GE. Two of their candidates are incumbent MEPs, and the third is a Senator, all have relatively good profiles. It should of course be expected that their vote will decline from their very successful 2019, but the poll doesn’t point to a wipeout.
Aontú’s position is quite interesting. At the time of writing, they don’t have a candidate in Dublin, and their candidate in South is an unknown running for election for the first time. However, they are running their biggest hitter in Midlands North-West, party leader and Meath West TD Peadar Tóibín. This raises an extremely relevant question that we can’t answer without regional figures – how much of that support is for Tóibín? A relatively even distribution of that 6% would likely result in no seats, but if it’s heavily concentrated in MNW, that gives them a real chance. As to why Aontú might overperform – voters in EEs tend to drift a bit more away from the political “centre”. This has benefitted both the left and right on various occasions – an obvious comparison point is Independent Peter Casey coming very close to a seat in MNW in 2019 before falling on his face in a GE.
Is a seat projection possible from this data?
Similar to my thoughts on local elections, the answer is pretty much no, not really. I mean, sure it’s possible in a very literal sense, but I would have quite low confidence in its accuracy.
But for the sake of interest I will show what a result based on the Ireland Thinks poll might look like. I once again want to emphasise that this should not be taken as predictive in any way, and will list a number of caveats first:
- First and most importantly, we are looking at one poll. Any projections from this should be taken with a truckload of salt, even without all of the other factors I will mention below
- There have been some constituency changes between MNW and South, which mean the historical data is less reliable (as, unlike for a GE, we do not have information on the precise vote in the areas that moved constituency)
- This poll is national, not regional, so there may be significant upwards and downwards skews per region that is flattened out in national data – see above remarks about Tóibín
- The poll does not separate Independents 4 Change or Independent Ireland; it’s plausible that supporters of these could have selected “Independent”, “Other”, or as they were not listed in the poll, another party entirely
- Furthermore, it isn’t possible to differentiate between individual Independent candidates; while some assumptions can be made about I4C, there’s little that can be done to fix this
- Then modelling can only include declared candidates, which shouldn’t be much of a factor at this stage, but you never know who will emerge last minute and what impact it may have
- Transfers are even more difficult than usual to factor with new candidates, changes in parties, changes in constituencies etc etc
So with that in mind, and with the final statement that this should not be considered in any way predictive, and is just a demonstration of what a single poll could imply, here is what the model would interpret from the Ireland Thinks Poll:
- Dublin (4 seats): One each for Fianna Fáil, Fine Gael, Sinn Féin and the Social Democrats.
- Sinn Féin are running two candidates, and it doesn’t from these numbers look like they have a clear path to getting both home. However, it should be noted that this poll is quite low for SF, as discussed above. If that turns out to be a polling error, two seats is plausible, if difficult.
- Both the Greens and Independents 4 Change are quite close behind this group and if we get polling errors equivalent to what we saw in 2019 (something far from guaranteed), they could both hold their seats.
- While PBP/Solidarity are further back, they have a much stronger candidate than in 2019, and are only running one instead of two. Candidate strength is not straightforward to take account of, and could result in them being a lot more competitive than the modelling of this poll shows. It is also likely that their vote is overwhelmingly concentrated in Dublin.
- The Soc Dems should, on raw polling, pick up a seat here but there is of course a nuance on regional breakdowns and not knowing how much support is in Dublin. That said, GE polling indicates that Dublin is far and away the strongest area for them. (To avoid repeating myself, I will just say here that the vote distribution question should be considered when assessing all three SD candidates, but Dublin is, based on current polling indicators, far and away their best chance of a seat)
- Midlands North-West (5 Seats): Two for Sinn Féin, one for Fine Gael and Independent Ireland, plus Independent Luke “Ming” Flanagan.
- MNW is far and away the constituency with the most potential to surprise. SF and FG are both basically guaranteed one seat, and everything after that is incredibly close and chaotic, both in the model and also from just looking at things. I am not fully confident on that second SF seat.
- Fianna Fáil’s decision to run three candidates is baffling, and unless they have extremely tight transfer management, basically ditch two of the candidates or massively overperform polling, it’s tough to see them winning a seat in this configuration. The only saving grace here is that regional polling for Fianna Fáil is massively variable, so an overperformance isn’t impossible. Either way, they have made this about as hard for themselves as humanly possible.
- Independent Ireland is a bit of a shot in the dark here – for the modelling I’ve used Peter Casey as a base; this could end up being way, way off, and I would have very low confidence in this output (insofar as all of these outputs should be fairly low-confidence, as outlined above)
- Aontú could well be more competitive here than the model shows, based on above remarks about their vote distribution, and will compete for some of that Casey vote. I wouldn’t actually be surprised if they are highly competitive here.
- South (5 seats): Two Fianna Fáil, and one each for Sinn Féin, Fine Gael and Independents 4 Change.
- FF winning two seats is a bit weird here, but making assumptions from this one poll, their vote is pretty much the same as what it was in 2019, whereas FGs is quite a bit worse. Again I would point to the confounding factors listed above as potential contributory factors for this output, but FF, FG and SF should all win one seat here.
- Sinn Féin will be very much in the running to pick up a second seat here based on this poll, but Fine Gael’s job of holding their second seat looks relatively difficult.
- The I4C seat is pretty marginal, but again we have to make a lot of assumptions as they are not split out within the poll, and they did overperform their polling in 2019 anyway. There’s more potential for a miss on them than with other parties.
- Based on current polls it will take a miracle for the Greens to hold on to their seat here, even assuming they overperform. A decline in their vote might help the Soc Dems, who have pretty good polling in Munster at the moment, but would be a pretty long shot here either way.
*****
Thanks so much for reading! This website is done entirely in my spare time and run without ads, so if you want to donate, please do so via Patreon or Ko-Fi – any support is greatly appreciated and 100% of any donations received are invested into the costs of running and improving the site.
Irish Election Projections 