How do you report on bank runs?

Looking ahead a couple of weeks, it might not be too long before the scent of bank runs is in the air. Given that, it might be a good idea to be thinking now, ahead of time, about the manner in which it’s sensible to report (and blog, and tweet, etc, to the extent that there’s any meaningful difference these days) on an incipient bank run.

A while ago, British charity Samaritans put together a set of guidelines for how to report on suicides, which have gained a degree of general acceptance. As a leading suicide prevention charity, they had noticed that the phenomenon of copycat suicides might be exacerbated by some features of the way in which they appeared in the media, and that responsible journalists could do a lot to minimise this effect, while still fulfilling their role in reporting the news.

Bank runs are intrinsically a phenomenon of copy-cat behaviour, and banking is a regulated industry, so it’s perhaps surprising that there’s no similar set of guidelines for responsible reporting on financial crises. There’s a general regulatory prohibition on spreading rumours, which people in the UK market were reminded about by the FSA in 2008, but this only applies to authorised individuals and it’s not very specific. Ofcom, the British media regulator, don’t seem to have produced anything either, which seems odd, as it means that the only serious look (as far as I can tell) which appears to have been taken at the role of BBC reporting in the Northern Rock collapse is Robert Peston’s own retrospective look at it on his blog. Peston’s thoughts are actually very insightful, but they are focused on the specific case and don’t really seem to generalise.

So what would such a code look like? Here’s my attempt at a sketch outline.

Accuracy

Scrupulous accuracy is more important than ever in the context of a bank run. Reporting of deposit flight should be based on specific and verifiable information, rather than anecdote. Extrapolation from aggregate statistics to individual banks should only be made with care and after checking for data issues, coverage changes etc. Rumours should not be passed on or “reported as rumour” unless they can be verified. Counterparties who are refusing to roll over short term deposits should only be cited if it can be demonstrated that their behaviour is typical.

Sensationalism

Language should not be excessively sensational. “Central bank lending” is a better term than “emergency lending”, and so on. The mere fact that a bank has requested official support should not be taken as a sign that it is insolvent. Pictures of queues outside bank branches and ATMs should only be used if they are representative — remember that there are queues at ATMs every Friday night in Central London. If there is no currently existing bank run, it is not sensible to use library pictures of an old one. Vox pop interviews with depositors should be used with care; people who are panicking are unlikely to contribute to a responsible report.

Incitement

Care should be taken to avoid the appearance of advising people to run on a bank. It is a fact that moving money ahead of a liquidity crisis is a rational thing to do, and this cannot be reasonably covered up or ignored and to do so is patronising and counterproductive. But if there is no other evidence of a bank being troubled than the fact that it superficially resembles another one which is, it makes little sense to tell its depositors that they should panic because other people might. To put this in concrete terms, during the Cypriot banking crisis, “move your money out of Cypriot banks because they are bust” was good advice, but “move your money out of Italian banks because the run might spread from Cyprus” was irresponsible.

Deposit guarantee schemes

All reporting on bank runs should contain a reminder of the deposit guarantee schemes relevant to the readership.

Non-intimidation

This principle stands in partial opposition to all the others. During financial crises, banks, governments and the public relations people employed by both, have a strong incentive to lie and to intimidate reporters. Nothing in these guidelines should be taken to justify this behaviour, which is nearly always counterproductive even in the short term. If a banking system has provable problems and is heading for crisis, it helps nobody for reporters to enable denial.

I am not a working journalist, so all of this should be taken as tentative. I circulated an early draft version of this a couple of years ago to a number of reporters, who I haven’t named here because the discussions were off the record.