Relationship Balance Sheets Are a Trap (And What to Keep Instead)

A friend drove me to the airport at 5 a.m. — the genuinely unpleasant favor, the one that costs someone their entire morning — and when I tried to hand him forty dollars for gas at the curb, something crossed his face that I have thought about ever since.

It wasn't offense exactly. It was closer to disappointment, and it was gone in a second, replaced by the ordinary business of refusing money. But I'd seen it, and I spent the flight trying to work out what I'd just done wrong, because by every account I'd been raised on I had done something right. He gave, I reciprocated, the account balanced.

That was the problem. I had balanced the account.

And in doing so I had told him, in a language older and more precise than the one we were speaking, exactly what kind of relationship I thought we were in.

The 1979 result, stated plainly¶

Margaret Clark and Judson Mills published a paper in the Journal of Personality and Social Psychology in 1979 that ought to be far better known outside psychology than it is. Their distinction, in their own words:

“

Communal relationships, in which the giving of a benefit in response to a need for the benefit is appropriate, are distinguished from exchange relationships, in which the giving of a benefit in response to the receipt of a benefit is appropriate.

”

Read the two halves against each other, because the difference is one word.

In an exchange relationship, the trigger for giving is having received. You bought lunch, so I buy lunch. This is not cynical — it's the correct operating system for a contractor, a colleague, a neighbor you like but don't know, most of commerce, and a great deal of ordinary decent life. When the counterparty is an employer it is emphatically the right frame, and you should read their documents accordingly.

In a communal relationship, the trigger for giving is the other person needing it. Not what they gave you. Not what you owe. Need.

Then Clark and Mills did the experiment, and the finding is the part that stops you. Their hypothesis was that receiving repayment after you've helped someone increases attraction when an exchange relationship is preferred — and decreases attraction when a communal relationship is desired. Both experiments supported it.

Repayment made people like each other less.

The distinction wasn't invented from nothing, either. Clark and Mills credit Erving Goffman's 1961 differentiation between social and economic exchange as the stimulus for it — they use "exchange" instead of Goffman's "economic exchange" precisely because most benefits people trade have no monetary value at all. A ride at 5 a.m. is a benefit. So is attention, patience, and being the one who remembers.

The trap is not what everyone says it is¶

Here's where the popular version of this idea goes wrong, and where it gets genuinely useful instead.

The folk advice is don't keep score. Stop counting. Give freely. It's on a thousand posters and it is, as stated, both unachievable and slightly dishonest — because people in close relationships plainly do keep track of things, constantly, and pretending otherwise just moves the tracking somewhere it can't be examined.

Clark, Mills and Martha Powell tested this directly in 1986, in a paper called Keeping Track of Needs in Communal and Exchange Relationships. Subjects were given the chance to check whether another person needed help. The design let the researchers see who monitored whom, and under what conditions.

The results are precise, and they invert the folk advice:

• When no opportunity to reciprocate existed, people wanting a communal relationship kept track of the other's needs more than those wanting an exchange relationship.
• People wanting an exchange relationship tracked more only when reciprocation was possible — track the debt, collect the debt.
• For people wanting a communal relationship, whether reciprocation was possible made no difference at all.

The first experiment supported all three hypotheses. A second went further: even when nothing could be done to help, people wanting a communal relationship still tracked the other's needs more.

Sit with that last one, because it's the whole thesis in a single result. They kept watching when watching could not produce any benefit to anyone. That is not scorekeeping. There is no score to settle and no way to settle it. It's simply what caring about someone looks like when you measure it in a laboratory.

Why the wrong ledger feels so reasonable¶

The contributions ledger has one enormous advantage: it produces a number.

I did bedtime four nights this week and you did one. That's checkable, defensible, and it feels like justice. The needs ledger produces nothing you can win an argument with. It produces a state of attention, which is unquantifiable, unprovable and impossible to wave at someone during a fight.

So under stress — and this is the mechanism, not a moral failing — people switch ledgers. Exhausted, unappreciated, frightened, you reach for the instrument that can prove you're owed. The contributions ledger is available precisely when the relationship is least able to survive being audited.

And it has a second property that makes it worse: it's usually accurate. That's the cruel part. You really did do bedtime four nights. The tally is not a distortion — it's a true statement drawn from the wrong book, which is the hardest kind of wrong to argue with and the reason these fights are unwinnable by either side.

This is the limit case for everything this desk argues about applying accounting to a life. Double-entry genuinely is a usable theory of the self — it clarifies what you're spending and what you're building. But every accounting frame has a domain, and the boundary of this one runs exactly where the counterparty becomes a person whose regard you need.

Inside that boundary the instrument stops measuring and starts doing something — the same way reputation isn't a number that drifts but a value that gets written down in one event. Producing the tally is itself an act in the relationship, and usually a declaration that the other book is closed.

The turn¶

I want to be careful here, because there's a version of this argument that is actively harmful and it's the version that spreads.

"Don't keep score" must not become "never notice what you're carrying."

The communal frame is not a licence for one person to absorb everything while being told that counting would be petty. That's not a communal relationship — it's an exchange relationship where one party has been talked out of their ledger, and the research doesn't endorse it. Clark and Mills describe mutual orientation, not unilateral self-erasure, and a relationship where only one person's needs get tracked is failing on the communal metric, not passing it.

So the honest formulation is narrower and harder:

Track needs, including your own. The needs ledger has two columns too. "What do I need, and does this person track it?" is a communal question, not a selfish one — and it's answerable without ever producing a tally of who did the dishes.

If you keep the needs ledger and find your own column has gone unread for a long time, you have learned something real. Not that you're owed. That the relationship isn't running the operating system you thought it was — which is far more useful information than being right about bedtime.

What to keep instead¶

1. 1

   Identify which frame each relationship is actually in

   Most are exchange, and that's correct and healthy — the plumber, the colleague, the neighbour. Trouble comes from mismatch, not from exchange itself. Write down the five relationships that matter most and mark each. The uncomfortable ones are where you've assumed communal and they're running exchange, or the reverse.

2. 2

   Notice the moment you reach for the tally

   The reach is the signal. When you catch yourself mentally assembling evidence of your own contributions, that's not a fact surfacing — it's a ledger switch happening under stress. You don't have to suppress it. Just register: I have picked up the other book. That half-second of noticing is most of the available benefit.

3. 3

   Convert the tally into a need before you say it out loud

   "I did bedtime four times" is a contributions claim and it starts an audit nobody wins. The same information as a need — "I'm running on empty and need one night off this week" — is answerable. Same underlying fact, different book, completely different conversation. This is the single highest-leverage move here.

4. 4

   Refuse repayment gracefully, and let others refuse yours

   When someone tries to settle up with you on something you did communally, understand what they're doing: usually not coldness but discomfort at being in need. And when you're the one reaching for the forty dollars, ask whether you're repaying a benefit or discharging the feeling of owing. Those are different, and only one is about them.

5. 5

   Audit your own column once a season, in writing

   Not who did what — whether your needs are being tracked at all. Quarterly, briefly, honestly. This is the guardrail that keeps "don't keep score" from becoming an instruction to disappear, and it works for the same reason any ledger works: a thing reviewed on a schedule cannot quietly drift for three years.

The curb¶

I never brought up the forty dollars, and he never mentioned it, and we have driven each other to a lot of airports since.

What changed is that I stopped experiencing those rides as debt. That's the entire practical difference and it sounds like nothing, but it removed a small persistent hum from a friendship I'd had for eleven years — the low background calculation of whether I was ahead or behind, which I hadn't known was running until it stopped.

He was never keeping that ledger. He was keeping the other one, the one with my name at the top and a single column that says what does he need, and he had been keeping it the whole time, including on all the occasions when there was nothing whatsoever he could do about it. The 1986 experiment found people do exactly that in a laboratory. He was doing it at 5 a.m. on a Tuesday, in a car, for free.

The books balance or they don't. Some of them were never supposed to.

It's just business — except, and this is the point, right here.

Sources

1. Clark, M. S., & Mills, J. (1979). Interpersonal attraction in exchange and communal relationships. Journal of Personality and Social Psychology, 37(1), 12–24.
   +How this was checked

   The founding distinction, and the finding that repayment increases attraction under an exchange norm but decreases it when a communal relationship is desired. Credits Goffman (1961) as the stimulus for the distinction.

2. Clark, M. S., Mills, J., & Powell, M. C. (1986). Keeping track of needs in communal and exchange relationships. Journal of Personality and Social Psychology, 51(2), 333–338.
   +How this was checked

   Two experiments on who monitors whose needs. Communal-oriented subjects tracked more, regardless of whether reciprocation was possible — and continued tracking even when nothing could be done to help.

3. Margaret Clark's relationship lab, Yale University
   +How this was checked

   Ongoing programme of research on communal and exchange norms, where both papers above are archived.