The final results are in and the scale and impact of changing voter preferences in the final days of this election has shocked us all.  No commentator, pollster, exit poll, academic analyst or bookie predicted the full extent of the movement in the end between Fine Gael and Fianna Fail.

In the run up to and during the campaign, polls did correctly identify that the government parties would most likely not be returned, accurately predicted the collapse of the Labour vote, the level of the Sinn Fein vote, and the strength of the Independent vote. Latterly they also showed the start of a move to Fianna Fail and some signs of a possible downward move for Fine Gael. Polling during the campaign on issues also helped inform commentators about the disconnect between the “Keep the Recovery Going” message, and the extent to which the recovery was felt by voters.

However, with late swing at play in the last few days, all polls have clearly underplayed the scale of the movement from Fine Gael to Fianna Fail. Some of the error recorded in our final poll vs. the result is of course due to that late swing of voters between Fine Gael and Fianna Fail; a movement that had been seen across all the pre-election polls taken up to a week out from polling day. There is also significant evidence, including the RTE exit poll indicating that 15% of voters made up their minds within 24 hours of the election. It is evident that these “late-deciders” came down on the side of Fianna Fáil more than the others. With another 20% deciding in the final week, this made this election particularly difficult to measure.

The final RED C poll did closely predict the support of many of the parties, well within the accepted margin of error.  Without any late swing our final poll figures for Fianna Fail and Fine Gael would also most likely have actually been within the accepted margin of error of any standalone poll measuring intention at that point in time, given that they are only just outside the error margin now.

However, despite detecting the beginnings of a shift to Fianna Fáil in our final poll, and having seen signs of a weakness in the Fine Gael vote as it declined in the early part of the campaign to 26%; it is clear that we didn’t capture the scale of this change in our final poll as accurately as we would have liked or expected.

We must acknowledge and explore the fact that our final polls did appear to in particular under represent Fianna Fail, and also somewhat over represent Fine Gael even before that late swing took place.

The fact that the rest of our predictions, including those for the other seven parties and Independent candidates, are close to the final first preference share suggests that the approach itself is not fundamentally flawed.  RED C has also built a reputation using this approach of accurately predicting elections and referendums over the last 12 years.

However, we do need to investigate further how we can ensure that we can better capture both elusive and/or “shy” Fianna Fail voters in our polls.

RED C are already in the process of analysing and reviewing the methods we have employed at this election, in order to ensure that we can continue to improve our contribution to the political discourse as a leading analyst of electoral mood.  We will of course keep you up to date on what we find, and how we feel these errors can be overcome in the future.