Very few households can answer simple questions about their own money

A quiz almost nobody passes

Ask what a household spent last month, what its total borrowing costs it in a year, or what it pays for its various recurring commitments, and most people will hesitate. This is not a comment on anyone’s competence. It includes people who are numerate, organised and doing perfectly well.

The reason it feels shameful is that the questions sound like they should be easy, and the difficulty is assumed to be personal. It is not. The information is genuinely difficult to assemble, and nothing in the ordinary arrangement of household finances assembles it for you.

That distinction matters, because a problem believed to be a personal failing tends to produce avoidance, while a problem understood as structural tends to produce a system.

It is also worth being precise about which questions actually matter. Nobody needs to know what they spent on a given Tuesday. The questions that influence decisions are few, and they are the ones most reliably unanswerable.

The numbers are scattered by design

A household’s money lives across several institutions, in several forms, under several logins. Income arrives in one place, spending leaves from two or three, borrowing sits somewhere else, and any longer-term savings are somewhere else again. No single view exists unless somebody builds one.

Each provider shows its own fragment accurately and none of them shows the total. This is not conspiracy; it is simply what happens when a household deals with many organisations, each of which knows only its own part.

Add to that the irregular items — the annual charges, the occasional large purchase, the costs that arrive twice a year — and a single month is not representative of anything. Answering honestly requires a year of data, which is exactly the amount of data nobody has to hand.

Recall is unreliable in a predictable direction

Even where the information could be reconstructed, memory does a poor job of it. Spending is remembered by episode rather than by total, so the large and unusual purchases are recalled and the steady accumulation of ordinary ones is not.

The result is an estimate that is systematically low, because the forgotten part is made up of the transactions that were unremarkable at the time. Somebody asked what they spend on ordinary daily items will almost always name a figure below what the record shows.

That gap is not dishonesty and it does not go away with effort. It is a property of how episodes are stored, which is why the useful method is to look at the record rather than to think harder.

The same asymmetry applies to income for anyone whose earnings vary, since the good months are the memorable ones. A household estimating both sides from memory can be wrong in both directions at once, and the two errors do not cancel.

Not knowing has specific costs

The consequences are concrete rather than moral. A household that does not know its total is unable to tell whether a change would help, so decisions get made on intuition and their effects are never verified. Recurring commitments continue past their usefulness because nothing lists them.

It also makes the household easy to price badly. An arrangement whose cost is unknown cannot be compared with an alternative, which is why introductory terms and quiet increases work so well: they rely on nobody holding the current figure in mind.

And it removes the ability to answer the most important question a household ever asks itself, which is whether it could absorb something going wrong. That answer requires knowing outgoings, and outgoings are exactly the number missing.

A small number of answers is enough

The remedy is not a detailed budget, which is more work than most households will sustain and more precision than the decisions require. Four figures will do: what arrives in a typical month, what leaves, what is owed in total, and what is held in reserve.

Those four can be assembled in an evening from statements rather than from memory, and they can be refreshed once or twice a year. That frequency is enough, because the numbers do not move quickly, and a stale figure is enormously more useful than no figure.

It is worth doing even when — especially when — the household expects to dislike the answers. The numbers are already true, and the only thing not knowing them changes is whether they can be acted on. If they turn out to reveal a problem of size, free regulated debt advice and regulated financial advice both exist, and both work considerably better with four accurate figures than with none.