Bank lending data reveals the science of loans and lies

Anything you say may be used against you — especially if it’s a fib.

The Sightgeist: Jon Kudelka.

I don’t lie. Honest to God. In fact, I swear on the life of my dear grandmother, who is in hospital right now, I am the most honest person you’ll meet. So, please believe me when I tell you, I don’t lie.

Obviously I’m lying. You can tell because I protested my honesty too much and, OK, I’m too old to have a sick grandmother. But there are a few other clues that gave me away. That is, my reference to God, the sick relative in hospital and, surprisingly, the fact I said please.

Now, I’ve always been sceptical of people who use God as an underwriter for promises but now we have real data to support this intuition. It comes from the fintech sector.

Some of those lending platforms that match lenders with borrowers have been digging through their data to find out who pays back loans and who doesn’t. Based on the words borrowers use when they apply for loans and whether they pay back those loans, the platforms have discovered that people who use words such as God, promise, hospital and thank you can’t be trusted. Less surprising are the words used by those who do pay back loans — interest rate, after-tax, minimum payment.

All this is obvious in retrospect. People who feel in control of their financial circumstances will make rational inquiries and those who feel least in control of their lives will plead, invoke God or embroider a sob story.

But perhaps it doesn’t matter why we lie or even the fact we lie, but more that it’s possible to find us out. That research comes from a forthcoming book called Everybody Lies and it explores how data on our online selves makes liars of us all. The subject of lying is hot with lots of heavyweight authors contributing titles such as The Truth About Trust, Telling Lies, Big Fat Lies and The (Honest) Truth About Dishonesty.

After a year in which the word of the year was fake news (Macquarie dictionary) or post-truth (Oxford dictionary), it’s no surprise the science of lying is growing.

We know there are lies, damned lies and statistics, but who knew there were white lies, black lies and blue lies? The sort of lies Donald Trump tells are blue lies which, according to Scientific American magazine, are “a psychologist’s term for falsehoods told on behalf of a group that can actually strengthen bonds among the members of that group”.

Expressions such as blue lies, core beliefs and alternative facts upset us in the media, if only because they challenge our core business. But should the rest of us care that the way we describe our husbands on Facebook (the best, my best friend) and the way we describe them on Google (jerk, annoying) differ? Should it matter that we tell dating sites we don’t care about race but never click on dates who are a different race to us? Should we worry about handing over our social media profiles to border officials in US airports if we don’t have any friends who live in Afghanistan caves?

Yes, we should, because we are laying a trail that institutions are using to create profiles of liars and they’re using those tools on us. Already a group of academics in England has developed a Pinocchio algorithm that trawls through texts, picking up language that flags users who are “possibly lying”.

While liars eventually will cotton on to words that expose them — use grand adjectives and make sure you use personal pronouns such as “I” and “me” — the rest of us increasingly will be judged not on what we say but on what they think we really mean when we say it.

We won’t know we’ve been refused a loan because we accidentally wished them “godspeed”. We won’t know we’ve been refused entry into a country because our friends on Facebook swear. We won’t know we’ve been refused a job because we liked the Facebook page I Love Being a Mum (evidently a sign of lower intelligence). And if they don’t take us at our word, what right of reply do we have?

Oh, by the way, I don’t lie. Or at least I find it hard to lie. Partly because I don’t like it, I’m not good at it and I figure if I spend my life asking people to tell me the truth I can hardly refuse to tell it myself. But perhaps I protest too much.