Why keeping score with friends feels gross

A friend covers your lunch. A few days later you send them the exact amount, to the cent, because that's what fair means. And somehow the gesture lands cold, and you can't quite say why.

Psychologists have been circling that feeling since the late seventies. Between friends, a precise repayment carries a second message about what kind of relationship this is, and it's rarely the message you meant to send.

Two kinds of relationships

In 1979, Margaret Clark and Judson Mills published two experiments in the Journal of Personality and Social Psychology that this whole corner of the field still leans on. Their claim was that people run relationships on two different sets of books. In an exchange relationship, benefits are debts: you repay them promptly and everyone keeps loose track. In a communal relationship, you give because the other person needs it, and nobody is supposed to be counting at all.

The test was simple. A subject helped a woman with a task, and she either repaid him right away with one of her extra-credit points or just said thanks. When a closer relationship seemed possible, the repayment backfired: liking for her fell from 194 to 177 on the study's 240-point scale. When the situation was strictly businesslike, the identical repayment pushed liking up, from 176 to 193. Same favor, opposite effect. A second experiment, this time with a partner who was new at school and hoping to make friends, found the mirror image: asking to be repaid warmed things up between acquaintances and chilled them between would-be friends.

The mechanism Clark and Mills proposed is the uncomfortable part. Repaying instantly implies the favor was given in expectation of return. It hands the other person a receipt, and the receipt reads: we are not close.

The pen experiment

Clark ran a sneakier version in 1984. Pairs worked a joint task, circling number sequences in a big grid for a shared reward, and each person picked a pen. Choose the same ink color as your partner and the two contributions blur together for good. Choose a different color and there's a running record of exactly who did what.

People expecting a transactional relationship took the different-colored pen almost every time, 88% of pairs in the first study and then 94% and 90% in the two that followed. People hoping to become friends with their partner did the opposite. Just 12.5% picked a pen that would keep their contributions countable, significantly below chance. They went out of their way to make bookkeeping impossible.

One wrinkle worth keeping. Established friends, tested in the later studies, mostly didn't track but didn't sabotage the record either. Clark's reading: destroying the ledger is a signal, and you send it when the relationship is new enough to need one.

The $9.99 problem

Now put payment apps in the picture. In 2019, Kim, Zhang and Norton published seven studies in the Journal of Experimental Psychology: General under the title “Pettiness in Social Exchange.” They showed real Venmo and PayPal users realistic mock transaction histories. One person paid friends in round numbers: $10, $35, $20. Another paid to the cent: $9.99, $34.95, $20.06. Asked who they'd rather be friends with, 81% picked the round-number payer.

It gets weirder. A $5.15 gift card produced less desire for friendship than a $5 one, despite being objectively more generous. A prospective date offering two hours and four minutes of help won approval 44.2% of the time; two hours flat won 61.5%. The precision alone did the damage.

Two details from the paper matter most here. The pettiness penalty only showed up between people; when the payment went to a company, round versus precise made no difference at all. And in a survey of 100 Venmo users' real transactions, 52.6% of friend-to-friend payments were precise amounts anyway. The thing that reads as petty is the thing more than half of us do, because the alternative is eating the difference or asking someone else to.

Roommates split evenly on purpose

None of this means friends stop caring about fairness. A 1980 experiment by Austin in Personality and Social Psychology Bulletin had roommates and strangers work on a task where performances clearly differed, then divide a reward. Roommates split it evenly no matter who had done the work. Strangers were the telling contrast: they divided by merit when merit favored them and discovered a sudden taste for equality when it didn't.

Friendship locked the rule in place; strangers picked whichever rule paid better. The even split between friends is a choice, and everyone knows what it says: the relationship outranks the arithmetic.

The olive-oil line

Flat life is where that choice gets stress-tested, because the arithmetic still has to happen. In 2020, Clark, Tuffin and Bowker interviewed 37 New Zealand flatmates aged 20 to 35 for a study in the New Zealand Journal of Psychology. Two sources of conflict dominated: clashing standards of cleanliness, and what the authors called freeloading and miserliness.

Freeloading covered unpaid bills, food that went missing, housework left to others. One interviewee described a flatmate's partner showering at the flat while everyone else covered the utilities. It kept ending with someone moving out. A flatmate told to start buying groceries or stop eating everyone else's food answered that he'd sooner leave, and then he did. But the penny-pinching side got no warmth either. The flatmate “who draws a line on the olive oil” came up in the interviews too; participants allowed that frugality is sometimes necessary and still found the miserliness unattractive.

The housework fights are their own genre; the money half is the part we care about here. Hold it next to the earlier findings and the bind snaps into view. The flat demands exact fairness. The flat also shames whoever produces exact fairness in person. You need the count kept, and you can't afford to be seen keeping it.

Where dvup comes in

Kim's company result points at the exit. Precision only reads as petty when it comes from a person; nobody in the 2019 studies cared whether a payment to a company was round or precise. So don't be the accountant. Appoint one.

That's the job dvup does. The app holds the ledger so nobody has to hold the different-colored pen. Expenses go in as they happen, and opposing debts collapse into a single net flow. The number you settle at the end came from software rather than from a friend's private tally. If it's $9.99, that's the ledger's doing, and your friend just paid what it said.

Splitting, balances, and settling up are free in dvup, with no cap on how many expenses you track. Receipt scanning gives you 5 free scans; after that, a short ad earns another, roughly one a day, and dvup+ removes the ads and the wait. And if your crew just splits everything evenly, that's one tap; we've written before about when the even split backfires at bigger tables.

Questions