At what stage did you decide who you would vote last week in the local and European elections?

Did you make up your mind ages ago, or only really pay attention and make up your mind in the few days before the election?

When you made your final choice, was the candidate from the party you would normally vote for, or did you vote for a local candidate who works hard in the area, and not really think about the party?

Evidence suggests voters are now much more volatile, likely to make up their minds who to vote for in the final few days before an election, and are far less likely to only vote for a candidate from one party.

The myriad of candidates at both the local and European elections also mean people tend to be drawn to candidates they know or have heard of working hard in the local area.

With all that in mind imagine trying to poll how people will vote a couple of weeks out from the election itself?

Polling local and European elections in the past has been shown to be quite a poor predictor of how people actually behave at local elections.  This is in contrast to polling for general elections, which has been shown to be very successful in Ireland over the last 25 years.

This year any polls conducted which asked about the party someone might vote for have been shown to be quite poor, while those asking specifically about candidates have performed much better.  Why is this?

Dr Stefan Muller, Assistant Professor, School of Politics at UCD, suggests “Irish local elections are unique. Incumbency plays a huge role in Local Elections. Our research shows that even incumbents who got elected on the last count face a huge advantage in rerunning and getting re-elected compared to the candidate who lost on the last count. Canvassing, name recognition and community ties are highly relevant.”

None of the polls at these local elections, got close to predicting how low Sinn Fein support would drop in the last 2 weeks running up to the election. They did however confirm the downward trend in support for the party, seen in the General Election polls over recent months.  This it appears was further exacerbated by factors that drive local election success, such as strength of candidate on the ground, campaigning, incumbency and of course turnout.

But we need to be clear that general election polls and local election polls are not the same thing.  They are looking at two very different elections for all the reasons set out by Professor Muller. Senior Fianna Fail politicians have been very vocal at calling out polls conducted online in particular as underrepresenting their support at this election. The comments suggest that general election polling conducted online over the past year, which has seen their party support fluctuate at around 15%-17% of the first preference vote, is not up to standard, because they actually achieved 23% of the first preference vote at the local election.

The majority of polls they are referencing were asking about how people would vote at a General Election, but their stronger performance was at the local elections.  It is clear, from all the evidence above, that you can’t point to General election polls and say they are wrong because they didn’t predict the result in the local election. As Micheal Martin said himself results in one, don’t equate to the other.

It is not inconceivable that the party could perform better at local elections, than at a future general election; particularly in light of the fact that Sinn Fein also did so poorly compared to the last General Election and recent opinion polls.

It is also the case that online polls, of which they specifically complain, either correctly predicted Fianna Fail’s share of the vote at the last General election or even overestimated it.  Furthermore, telephone polls in the year running up to the last election in 2020, regularly overestimated their support vs. the final result.

It is also important to recognise that online polling of “specific candidates” at these European elections, as opposed to ones asking about party support have also been shown to be pretty accurate, well within sample error for all Fianna Fail candidates.

Having said all of that, the European candidate poll results from online polls did tend to slightly “underrepresent” Fianna Fail candidates, but within the margin of error. Thus, there is some evidence that right now, online polling is possibly slightly underrepresenting Fianna Fail voters. This is however to a much lesser degree than claimed by the party, who are looking at the difference between local election results and General election polls.

As pollsters we are constantly reviewing our methods to ensure the best accuracy.  No method is perfect.  Face to face polls have to overcome issues with people not answering the door and evidence that people are more likely to tell the interviewer what they want to hear.  Telephone polls have to overcome the fact that less and less people answer their phones to unknown numbers.  Online polls rely on large panels of people who take surveys to respond, which can occasionally be weaker in certain demographic cohorts.

We work hard to mitigate against all these factors using as random a sampling approach as possible, looking at sample structure through quotas and weighting on key variables such as education, income, working status and region. We also look at ways to predict turnout to give greater weight to those most likely to vote.

If there is some evidence that current online polling method needs to be tweaked or supplemented on these aspects to further improve their performance, you can rest assured we will certainly be doing so.  The aim always being to show trends in support that help us to understand likely voter behaviour into the future.