Analysis: How accurate was the election exit poll? – RTE.ie


The expression goes that “nature abhors a vacuum” and the world of politics is no different – particularly during an election.

After the polls close, people want to know what has happened, even before the ballot boxes are opened.

To fill that void, exit polls are used to give an indication of a possible result.

These are substantially different to opinion polls because they interview people after they vote. So they are considered to be a more accurate barometer of trends.

In the case of the European elections in Ireland, the gap between the polls closing and the first results being announced is unusually long at 48 hours.

For politicians, pundits and the media, the first information they can use as guidance comes from exit polls.

In the local and European elections last month, the only such survey conducted was carried out by REDC for RTÉ and TG4.

All of these surveys come with health warnings. In this case the exit poll had a margin of error of +/- 4% for the European elections and +/- 3% for the local elections.

It was conducted among 3,246 respondents at 156 polling stations.

The survey correctly identified some major trends such as the increase in support for the Greens, the party’s Ciaran Cuffe topping the poll in Dublin and the strong Yes vote in the Divorce Referendum.

It also gave a better indication of the final outcome than opinion polls conducted during the campaign, which is unsurprising given the larger sample size.

But not all of its projections were accurate, prompting criticism from some political parties.

It gave indications for each of the 59 candidates running in the European elections and was wrong by 2.5 points in four cases.

It overstated Cuffe’s support by 5.5 points and his party colleague Saoirse McHugh by 3.4 points, understated support for Independent Luke ‘Ming’ Flanagan by 4.4 points while Independent Peter Casey was understated by 2.5 points.

The exit poll was wrong by two points in four other cases. Fine Gael’s Frances Fitzgerald, Barry Andrews and Ann Rabbitte, both of Fianna Fáil, were all understated by two points. Matt Carthy of Sinn Féin was overstated by two points.

In the local elections, the exit poll gave indications for 14 parties or groupings and was wrong by more than 2.5 points in three cases. It understated Fianna Fáil by 3.9 points, overstated the Greens by 3.5 points and overstated Sinn Féin by 2.5 points.

In the Divorce Referendum the poll suggested a Yes vote of 88% versus an outcome of 82%.

The survey was criticised by some political parties and particularly Fianna Fáil, which said the exit poll underplayed its level of support.

After such exit polls are published, a review is held to measure the projections against the actual results and attempts to explain the reasons behind the errors.

In recent weeks REDC has conducted an analysis.

It said that “most results” for the parties or candidates at the local and European elections were “within the guided margin of error”.

It said it was “naturally disappointed” that some of the results were outside the margin of error.

It is worth noting that some academics point out that the margin of error is proportionate to the vote won by a candidate. 

In analysis on the exit poll, Emeritus Trinity College Professor Michael Marsh explained: “If a candidate actually gets 50% of the vote, the sample estimate – the poll figure – should be between 46 and 54.

“However, it should be closer to the real outcome as that moves away from 50. If a candidate won 20%, our poll figure should be more in the range of 17-23, the range getting tighter as the figure gets further from 50%.”

Using that standard, the REDC exit poll was outside what many statisticians would regard as the margin of error for four of the 59 European candidates, three of the 14 parties and the referendum result.

In its review, REDC said there had been an unusual turnout in the election – particularly in Dublin, where it was lower than expected and behind the level of turnout in previous elections.

It also varied across the capital at 51% in Rathfarnham and 27% in Tallaght South. This led to an over-representation of the Greens.

REDC said: “Many areas with low turnout were also in pockets where the Green vote was strong.”

Another factor was that some people left it unusually late to vote. Some polling stations reported turnout as low as 13% by midday.The effect of this was that people who voted in morning were over-represented in the exit poll.

A further issue was there were three elections taking place at the same time.

REDC showed ballot sheets with candidates for the European elections but was unable to do so for the local elections.

Instead it asked people which parties they voted for the local contest. In its analysis the polling company said that with such a large number of candidates “this clearly can also lead to confusion among voters”.

People often vote for candidates rather than party in the local elections. So when members of the public are asked which party they voted for they are “often unsure” which party the candidate represents.

Exit polls in other elections by different companies have also deviated from the final result as pointed out by Professor Marsh. So the issues which arose following the REDC survey are by no means unique.

In the absence of an exit poll, parties would be left to rely on tallies which are unofficial observations of how candidates are performing.

But many politicians point out that tallies are less accurate than exit polls in local and European elections.

Also, exit polls collect data useful to social scientists and political parties so they can better understand the electorate.

The REDC survey asked questions about the age of voters for particular parties and their views on other topics such as a United Ireland.

While some politicians are critical of the predictions made in the exit poll, even critics point out that the surveys of this nature perform a valuable function.

Perhaps, like nature, politicians too dislike vacuums.

“Disclaimer – This story is not been edited by ViralAssam Staff and is auto-generated from syndicated feed”


Like it? Share with your friends!

What's Your Reaction?

hate hate
0
hate
confused confused
0
confused
fail fail
0
fail
fun fun
0
fun
geeky geeky
0
geeky
love love
0
love
lol lol
0
lol
omg omg
0
omg
win win
0
win
Choose A Format
Personality quiz
Series of questions that intends to reveal something about the personality
Trivia quiz
Series of questions with right and wrong answers that intends to check knowledge
Poll
Voting to make decisions or determine opinions
Story
Formatted Text with Embeds and Visuals
List
The Classic Internet Listicles
Countdown
The Classic Internet Countdowns
Open List
Submit your own item and vote up for the best submission
Ranked List
Upvote or downvote to decide the best list item
Meme
Upload your own images to make custom memes
Video
Youtube, Vimeo or Vine Embeds
Audio
Soundcloud or Mixcloud Embeds
Image
Photo or GIF
Gif
GIF format