This Broadsheet.ie column appeared online on April 30th, 2019
Over the Bank Holiday weekend, two Sunday newspapers published political polls. They were detailed. They were professionally conducted. But above all else, they offered very different insights into the state of the main parties.
RedC, polling for the Sunday Business Post, reckons that Fine Gael is pulling well ahead of Fianna Fáil. According to its findings, the ratings for the top 4 groupings are, in decreasing order: FG 33%, FF, 23%, Inds 16% and SF% 14.
Not so, according to B&A, polling for the Sunday Times (Ireland). According to its research, voters are now shifting significantly from FG to FF, putting Fianna Fáil in first place with 29%, followed by FG on 28%, SF 21% and Inds 10%.
The field work for both polls concluded around the same time April 16/17, though RedC did its field work over a week, while B&A took about almost two.
For an informative and detailed comparison of the methodologies employed by both sets of pollsters, check out Prof Michael Marsh’s blogpost on the RTE website.
He makes the point that while both polling companies have often varied on the actual hard numbers, they have been in agreement for most of the past three years on the relative positions of the two main parties.
It is a point confirmed by trawl through the national polls conducted over the past few years. 45 out of the 46 polls conducted since July 2017 had Fine Gael ahead of Fianna Fáil. The recent B&A/Sunday Times poll is the first one in that time to put Fianna Fáil ahead. If true, this could be the first big shift in public opinion since Leo Varadkar became FG leader and Taoiseach in June 2017.
The important word is “if”. One poll out of 46 is just that – it’s 1 in 46. These are not great odds. Besides, if B&A is right, then mustn’t Red C be wrong? Afterall, though the two polling companies report significant shifts – they are in opposite directions.
We will know which one is right in just over three weeks when we have the results of the Local and European elections. In the meantime, it is useful to note that we have seen similar results before and they softened out in later polls.
In Sept 2018 RedC had FG 11% ahead of FF (33 vs 22) while as recently as January 2019, it had FG 10% ahead: FG 32: FF 22. This gap narrowed back to 6% within two months.
At around the same time (Dec 2018) B&A reported FG’s lead over FF down to just 2%, though it widened back out to 4% in early 2019.
So, what point am I trying to make here? Well there are two and they are political, not statistical.
First up is a point I have made here several times: namely that political parties do not do their polling in the way as the national newspapers.
This is not to say that these polls have no value or interest. They clearly have an interest – otherwise why would politicos be frantically texting colleagues or hitting refresh on Twitter every 30secs after 5pm on the Saturday evening looking for first sight of the numbers?
But it is just that, interest. They fuel the political natter and chatter in the days afterwards. They give political interviewers some nice opening questions for the following week… “Minister/Deputy/Senator… with your party …surging/static/plummeting …in the polls, do you need to change/reverse/dump… your… policy/leader/spin doctor?” [delete as appropriate]
Not to mention the hours of harmless fun they provide political nerds on hundreds of WhatsApp chats or Facebook private groups as they punch the new party totals into their excel spreadsheets and forecast their seat totals for a general election the date of which no one yet knows… including it seems Leo Varadkar.
Though I will come back to point one shortly, let me digress briefly to point two.
This form of political polling, the national poll of 1000 voters, is now effectively a branch of entertainment and – given how boring and turgid our politics can be – that is no bad thing. Just as long as we remember that it is more part of the entertainment end of politics, than the strategic.
It is also nothing new. The science of polling, and it is a science, was born out of the newspaper industry. In the late 19th century many American newspapers printed sample ballots which they asked readers to cut out, complete and send return so they could forecast the result. They came to be known as “straw polls” and political party operatives soon began to realise their usefulness.
The eponymous American pollster George Gallup started out as an academic who shifted into the business after devising “An Objective Method for Determining Reader Interest in the Content of a Newspaper”. This was an early form of focus group designed for newspapers where Gallup would observe people reading their papers, note which sections they liked and disliked and then advise the editor which parts to keep and which ones to drop.
Gallup quickly grasped the commercial potential of measuring public opinion, but he still saw it in the newspaper realms, regarding it in fact as “a new form of journalism”. He saw himself as taking the “pulse of democracy” fore newspapers between elections. A remark that led The New Yorker’s E. B. White to quip: “Although you can take a nation’s pulse, you can’t be sure that the nation hasn’t just run up a flight of stairs.”
Could this help explain Leo’s outpacing Micheál up to now?
Maybe not, but back to point one. As I have said, political parties do not poll in the same way as newspapers. They have long since realised that such broad canvass polling does not pick up on the critical factors that influence voting behaviours.
Parties do their polling constituency by constituency, based on named candidates. In Irish politics, the local and the personal matter a lot and it is an indication of the sophistication of Irish voters.
Ask 500 random voters which party they would likely vote for at some election in the future and you get one result but put a ballot paper in front of them with a list of their local TDs and other candidates and let them decide in private and you will likely get a slightly different result.
It is not just voter sophistication, it is also – to borrow a phrase from former campaign management colleague – an indication of voter promiscuity, i.e. their facility to vote across parties and groupings and switch between elections.
It is a sophistication that also allows voters to see the benign entertainment value of these polls, but not be driven by them. While people might complain that the frequency of polling means we more often talk more about the process of politics, the who’s in or out, who’s up or down, than its substance – their independence and transparency means they are at least based on some reality.
Something you cannot say about a lot of the insidious material that been peddled online in recent election campaigns in the US and elsewhere.