Poll of Polls with seat projections: Is Fianna Fáil support underestimated?

Eoin O’Malley (20 February, 2011)

The three most recent polls by Red-C,  Millward Brown and Ipsos-mrbi in today’s Sunday Business Post and Sunday Independent and tomorrow’s Irish Times are broadly consistent and are also consistent with recent trends. Fine Gael’s position is solidifying (not surging) and Labour is slipping slightly (not falling). The Greens look unlikely to return any seats. Fianna Fáil and Sinn Féin also seem not to be moving. In fact it was the week between the 2nd of February and the 10th February is where there seems to have been movement (see below) not the week just passed. That’s not to say the final week of the campaign can’t see some significant changes. The results of the weighted poll of polls are

The weighted average estimates presented here are possibly more valid than any single poll because it should average out errors (assuming the errors in the polls are not all systematically in one direction). It means that the results of one poll won’t have a major impact on our overall assessment of the state of the parties. We will expect that some polls will have error above and beyond the margin of error. Even with the best will in the world some samples produced by polling companies are ‘rogue samples’ and produce results that are not good estimates. About one in twenty will be like this.

Polling isn’t all science – an important part is the art of the questioning and thinking about how the method might bias the estimates. We might consider that all polls are underestimating Fianna Fáil for a number of reasons. One is that people will be shy of admitting to vote for Fianna Fáil, particularly in face to face interviews because of a stigma attached the party now (this would affect Millward Brown and Ipsos-mrbi). Red-C and OI are open to correcting for this through political weighting (matching what party people say they voted for in the last election with the actual result). Both however say it’s not an issue because people’s responses are broadly similar to the actual result. Of course this assumes that people who are ashamed to say they would vote for them now, would also be ashamed to admit having voted Fianna Fáil in the past.

Another issue is that our elections are conducted using candidate-based ballots, whereas the polling companies ask what party you will vote for. It’s quite possible that many would say they’d never vote for Fianna Fáil but would be happy to vote for Pat Carey because he’s a decent fellow. They might not even consider these positions as being inconsistent. Some local polls which use ballot papers allow us to compare broadly. One in Cork NC and Tipp South seem to be broadly consistent with the regional trend from the national polls, but one in Dún Laoghaire indicates that Fianna Fáil is (at about 18%) much higher than the Dublin figures from other surveys would show. The difference here might be that there are two well-known candidates in Dún Laoghaire, where as the two more outspoken Fianna Fáil TDs for Cork NC and Tipp South are standing down or as independent.

It’s difficult to come to any firm conclusion about whether Fianna Fáil estimates are about right or systematically below the real vote intention. But perhaps we shouldn’t be surprised to see Fianna Fáil closer to 20 percent in the final results. This could be opinion poll underestimation manifesting itself as a surge on the last few days of the campaign. However this might not translate into a big jump in seats (from what is estimated here). This is because Fianna Fáil is not receiving transfers as it had in 2002, when it was both popular and transfer friendly.

On the weighted poll of poll figures, using Michael Gallagher’s method of projecting polls to seats we would expect to see Fianna Fáil return 25-29 TDs (incl. Seamus Kirk); Fine Gael would win 65-69 seats, a good deal lower than the estimates Adrian Kavanagh predicts; Labour would win 33-37 seats and Sinn Féin 17-21 seats. It’s difficult to say much about the smaller parties because local considerations and the geographical concentration of the vote matter most, but we can assume Others would win between 17 and 20 seats.

There has been some evidence in local polls and in the SBP-Red-C question on second preferences that Fine Gael is getting a lot of transfers and as by far the biggest party might be expected to get an even bigger seat bonus than is factored in. Even if the bonus gave it 12 seats more than it ‘deserved’ on a proportional basis (about what Fianna Fáil got in 2002), Fine Gael would still, at 75 seats, be eight seats shy of a majority.

On that basis, and given that it’s far from clear what type of independents might be elected, the Fine Gael-Labour coalition is still the most likely government to emerge. But it will not be a coalition of equals. This is not comparable to the government formed in 1992 between FF and Labour, which though the numbers might be broadly the same (FF had 68 and Labour 33) Fianna Fáil had a very poor election by its then very high standards and Labour had its best result ever. In 2011 Fine Gael will not be minded to concede as much as Reynolds was willing to give to Spring, and even though it still might be Labour’s best result ever and it could be the second largest party for the first time, Labour will be disconsolate at its failure to make a bigger breakthrough.

This is even more reason why Labour might be better advised to stay in opposition as I argued here.

In weighting the polls I take into account a number of issues (see here for details).

There is an element of arbitrariness in the weighting (and because I don’t have the full details of the polls, I’ve guessed things such as the dates of fieldwork), but in general it should give a better indication of the state of the parties.

Below are the data and weightings used. I’ve also included polls from earlier in the campaign to show the trend over time. We can see that a shift took place between the 2nd and 10th February.