If we can’t see Facebook ads, how can we challenge them?
Online advertising paid for by foreign sources has the potential to undermine debate in our democracy, writes Eoin O’Malley
IN the summer of 1994 I found myself somewhere in rural west Limerick, canvassing for my father’s failed attempt to get elected to the European parliament.
I canvassed a lady with an English accent, and clothes that marked her out as ‘alternative’. She asked about the candidate’s opinion on hare coursing, an issue that was current, but not one I knew much about. It didn’t come up in our usual haunts in Limerick city. I didn’t even know what my father’s views were — but I was pretty sure I knew hers.
I exclaimed that I found hare coursing a horrific excuse for a sport, leaving her satisfied.
A few houses away, a man with working hands and a strong Cork accent asked about the candidate’s view on the same issue. “He’s sound on it,” I said, winking. He too was satisfied that he and the candidate were as one.
I didn’t lie, but I’m not sure I told the whole truth.
At the time the bigger parties, especially Fianna Fail, would have had intelligence on who lived where, how they voted and what issues were important to each. They would have been better prepared than we were going to each door. A lot of that micro knowledge is gone because parties just don’t have as many members as they did, and people’s political concerns aren’t as obvious.
Legendary political operations, such as that operated by Bertie Ahern, allegedly knew who to get out, on what issues they could be mobilised to vote, and who not to waste their time on.
There is of course a name for this type of canvassing: microtargeting.
In the last two British elections, the Brexit referendum and Trump’s election, there are claims that political campaigning has moved online and that much of the online work is microtargeting.
The Tory party spent £1.2m (€1.37m) on Facebook ads in 2015 and the Labour Party may have spent a similar amount this year. These are pretty small numbers, especially when you consider that the Leave side in the Brexit referendum reckon it was the Facebook ads that swung the result for them.
Microtargeting starts with a political campaign putting different voters into segments according to their personal and political preferences, based on polls.
In the Brexit campaign, Leave identified six groups. Two were solidly Remain, accounting for about 30pc of the electorate. Two were certain to vote to Leave, accounting for a third of voters.
The rest were in one of two floating groups: the first was the ‘Disengaged Middle’ (middle-class, in their 30s and not interested in politics). They didn’t regard the EU as important and didn’t know much about it.
The other group they called ‘Hearts v. Minds’ (mainly female, late middle age, poorly educated, with children, and in low-paid jobs). They didn’t like the EU but were unsure that leaving was such a good idea.
Campaigns work by making information available to voters. We all have predispositions, and some information confirms us in our views, some persuades, some repels us.
Both persuadable groups were susceptible to economic arguments that put forward the idea that Britain would do well outside the EU.
The ‘Disengaged Middle’ didn’t feel threatened by immigration, and were a bit queasy about arguments against immigrants. The ‘Hearts v. Minds’ were attracted to the anti-immigrant rhetoric of Ukip.
As in Ireland, the UK does not allow broadcast advertising for political aims. But advertising online is unregulated.
Microtargeting involves a sort of ‘narrowcast’ advertising. Companies such as Facebook claim they can identify target groups because of the vast amounts of information they have — often information that people have willingly shared with the company.
A political campaign can then offer an ad on Facebook pages tailored to the people it wants to persuade — and unlike ads you might see in a newspaper, those ads can be very different depending on the different reader.
Those who don’t like immigrants can be shown ads using anti-immigration arguments, while the other group can be channelled arguments tailored to their predispositions.
This poses danger to democracy — because a basic idea in democratic debate is that the debate should happen publicly.
It’s when those arguments come in contact with all voters that they are tested. But if you can privately say one thing to one group and another to a different group, you can deceive both.
Here in Ireland, in the Lisbon Treaty referendums of 2008 and 2009, the Yes side stuck with a dull campaign based on a message of ‘jobs and growth’.
The No side was supported by a medley of groups, with differing and often contradictory arguments. Some saw the Treaty as a sure-fire way your children would be conscripted into a European army. Others thought those children would be lucky to make it to that army because the EU was poised to impose abortion on the country. In short, different groups made different claims to their target audience.
However the fact that all voters — be they Yes or No supporters — could see the various arguments meant that they were possible to refute. And one of the greatest fears with Facebook ads is that, because they are not made public, they cannot be challenged.
There will be a referendum in Ireland on abortion next year. The louder shouters on both sides in the debate claim that nefarious outsiders will try to influence the outcome.
There’s very little to stop outsiders doing so. While party election spending is regulated, spending by third parties is not as regulated.
It would be easy for a third party based abroad to influence the referendum through Facebook ads without ever breaking a law here.
There’s an understandable rush to have a referendum on abortion — but it might be best to sort out narrowcast advertising first. Better that than have the losing side claim the referendum was unfairly influenced, without anyone really knowing whether it was.
‘There’s very little we can do to stop outsiders from influencing any referendum we hold on abortion...’