What the data actually shows
Studies of spending perception find that people's estimates of their own outgoings diverge from their actual outgoings, and the error is patterned rather than random. Small, frequent, and recurring costs — subscriptions, snacks, transport, the daily incidentals — are the ones most easily underestimated, because each is forgettable and they aren't experienced as a single salient event. Large or one-off purchases, by contrast, are vivid and easy to recall, so they get over-weighted in our mental summary.
Richard Thaler's concept of mental accounting helps explain the distortion: we don't track money as one running total, we sort it into separate mental 'accounts' and categories, and the small, scattered, low-emotion spending tends to slip out of view between those accounts. The result is a self-image of our finances assembled from the memorable bits, not the representative ones.
The mis-estimation also runs forward in time. Research on how people predict their own future spending finds they tend to underestimate it — expecting future months to be cheaper or more disciplined than they turn out to be — which feeds over-optimistic budgets and plans. Memory undercounts the past and prediction undercounts the future, and both lean the same way.
Why this feels different from how it actually is
It feels like we know where our money goes because we remember our spending in stories — the holiday, the big purchase, the bill that stung — and those stories feel complete. What they leave out is precisely the part that is hardest to recall: the dozens of small, ordinary transactions that don't attach to a memory.
Small recurring costs are also designed to be forgettable. A subscription that renews silently or a few pounds spent on autopilot generates almost no emotional or attentional mark, so it never enters the mental tally even though, repeated across a month, it can add up to more than the purchases we do remember.
And there's a comfort bias: it's easier to picture ourselves as careful than as leaky, so the version of our spending we reconstruct tends to be the tidier, more deliberate one. The gap between that story and the bank statement is usually where the surprise lives.
Memory undercounts the past and prediction undercounts the future — and both lean the same way.
What the research says to do about it
Track, even briefly. Because the error comes from memory rather than from how you spend, simply looking — categorising a month of real transactions from your statement or an app — tends to be revealing. You don't have to track forever; a short, honest audit replaces the reconstructed picture with the actual one and usually surfaces the small recurring costs you'd discounted.
Build systems instead of leaning on awareness. The research on misperception implies that relying on 'staying aware' is fragile, because awareness is exactly what's biased. Automating saving — paying yourself first, automatic transfers on payday — moves the outcome out of the reach of fuzzy in-the-moment perception, so the result no longer depends on accurately tracking everything you spend.
Audit the recurring layer specifically. Since subscriptions and small repeating charges are the most-underestimated category, periodically listing every recurring payment tends to catch the leaks memory misses — the forgotten subscription, the creeping renewal — that a one-time mental review never would.
What the research says does not help
Estimating from memory is the weak link, so basing a budget on what you think you spend tends to start from a biased number. The point of tracking is precisely that recall and reality diverge in a patterned way, which a memory-based estimate can't correct for.
Resolving to 'just be more aware' of spending rarely fixes it, because the small, low-attention purchases that drive the error are the ones awareness slides past. Vigilance is hard to sustain and is biased at exactly the moments it would need to work.
Focusing only on the big, memorable purchases can miss where the money actually goes. The research suggests the underestimated drip of small recurring costs is often the larger, more invisible part of the gap — so cutting one dramatic expense while ignoring the steady leaks may not move the picture much.
The gap between the story we tell about our spending and the bank statement is usually where the surprise lives.
What this looks like in real life
The month that didn't add up
You'd have sworn the month was quiet — no big purchases, nothing memorable. Then the statement shows the total is higher than expected. The gap isn't one dramatic buy; it's the dozens of small, forgettable transactions that never attached to a memory. That's the surprise living in the space between the story and the statement.
The subscriptions you forgot you had
Listing every recurring payment tends to surface charges that renewed silently — the forgotten subscription, the creeping renewal. Each is small enough to slip out of the mental tally, but repeated across a month they can add up to more than the purchases you do remember. Auditing the recurring layer catches what a one-time mental review never would.
Real numbers in context
The robust finding here is directional rather than a single clean statistic: across studies of spending perception, people's estimates of their own outgoings systematically diverge from reality, with small, routine, recurring costs underestimated and large or unusual ones over-remembered. Exact gaps vary by study and method, so treat any specific percentage with caution — the reliable point is the consistent direction and the role of mental accounting (Thaler) in producing it.
The same lopsided error shows up in prediction: research on forecasting future spending finds people tend to underestimate what they will spend, which is part of why budgets so often feel tight in practice. Put together, memory undercounts the past and forecasts undercount the future — which is the case for letting systems and tracking, rather than awareness, carry the load.