Sunday, 6 March 2016

Did Conservatives Facebook spending break the law?

So here’s a thing. During the 2015 general election, the Conservatives spent more than a million pounds on Facebook advertising. This spending was targeted incredibly carefully at the people the Tories wanted to most reach - the people who had been identified as swing voters in marginal constituencies. Was this spending legal and did it put individual constituency campaigns over the expenses limit? Sadly, we can never know. But it is a question which ought to be asked and if the Conservatives did not break the law in 2015, you can be sure that the way is now open for parties or individual candidates to do so in the future.

The overall spending on Facebook advertising was declared by the Conservatives as part of their national return. If this spending was all made on a generic “Vote Conservative” message then it would be perfectly legal (their declared national spending was about £3.5 million below the limit). But we know that this was not the case.

The whole point of Facebook’s advertising model is that they are able to target users very specifically. Their sales pitch is that you are paying to reach the people you want to. Their algorithms know who Facebook users are, what they are interested in and what their political preferences might be based on their use of the site. Moreover, they know the age and family circumstances and many other aspects of users’ lives. That's gold dust to political campaigns aiming to send a very specific message to very specific voters.

So if a political party comes along and tells Facebook which constituencies they are interested in and what demographics they want to target then the tech giant can make sure their message hits that particular band with better than 95% accuracy.

The Conservatives spent an average of £114,000 per month on Facebook advertising in the run up to last year’s general election. We know that from the expense returns filed with the Electoral Commission. We don’t know precisely how much was spent during the short campaign - the period of five weeks or so between the close of nominations and polling day itself when constituency spending is most tightly regulated to an average of about £16,000 per seat. But it is highly unlikely that the spending would have fallen during this most crucial period.

And no one outside Conservative Central Office knows precisely how many seats they spent their Facebook money on. But news reports before the election talked about a 40-40 target strategy - concentrating all their efforts on defending the 40 most marginal Tory held seats and gaining the 40 most marginal seats held by other parties. Things might have changed during the election with a few more seats becoming viewed as winnable, but this is likely to have been balanced by some held seats becoming viewed as safe. There is no point for the Conservatives of spending Facebook money to win the votes of users in safe Labour seats like Barnsley. Nor would they spend money seeking to add to a majority of more than 10,000 in seats like Surrey Heath.

If the total Conservative Facebook spend was divided between these 80 target seats (which is a reasonable supposition), then this would average around £15,000 per seat over the year before the election.

If we assume that the amount spent per month did not drop over the final period (and logic dictates it is only likely to have risen) then the average spent in each of the 80 seats during the short campaign would have been at least £1425. How many Conservative candidates declared spending of anything like this amount on Facebook advertising in their expense returns? None.

We know from the spending on by-elections like Newark in the run up to the election that the Conservatives did spend money on very specific, constituency and candidate oriented Facebook advertising. In that campaign a receipt was submitted for £3000. (Spending limits on by-elections are much higher at £100,000.) Are we really to believe that the Conservatives abandoned their candidate and seat specific advertising which served them so well at Newark for the general election?

The trouble is that we have no way of knowing for sure. Whilst leaflets and emails can be preserved for later examination, Facebook advertising is ephemeral. Unless a recipient has the presence of mind to screenshot it, we will never know what it said. And the whole point of the targeting algorithm is that the Conservatives will not have sent their Facebook adverts to people who are most likely to have recorded it in this way - their opponents.

What is absolutely clear though, is that this is a massive loophole in election expense law which would be easy to exploit in the future - even if it has not been until now. From everything that came out of the Conservative hierarchy, we know that they viewed their Facebook spending as worthwhile. They will do the same (or more) next time and the other parties will be playing catch-up.

For the record: I am not suggesting that Facebook have done anything wrong. They are a private company which has obtained all this data legally and with the consent of their users. They have the right to monetise their product and that is what they have done by selling advertising. Nor do I think that it is wrong for political candidates or parties to buy such advertising during elections. My concern is transparency. If the law states that parties can only spend up to a certain amount during the campaign in an individual constituency then we should expect there to be a mechanism to ensure that this is the case and that spending can be policed.

No comments: