How to merge two separate ledgers into one without double counting
Two people have been keeping their own records for months and now want a single household book. What to do about the overlap, how to remove the transactions that exist twice, and whether to bring the history across at all.
Two people have each been keeping their own record for six months. Both are reasonably diligent. Now they want one household book, and the obvious move is to pour both files into it and carry on.
That move produces a book that is wrong in ways which are hard to see and harder to undo. Some purchases exist in both records. Transfers between the two of you appear as expenses on one side and income on the other. The two category schemes do not agree about what a category is. And the totals for the overlap months will be inflated by an amount nobody can quantify afterwards, which means the year-end figures are quietly useless.
The merge is doable. It is mostly a sequencing problem, and the first decision is the one people skip.
First decide whether the history is coming at all
There are three defensible answers and only one of them involves any work.
The first is to start the shared book from a date and leave both histories where they are. Each person keeps their own past, can still look at it, and the household record begins clean on the first of a month. This costs nothing, produces no double counting, and is the right answer far more often than people expect. The main objection to it is a desire for a complete year, and a complete year assembled from two incompatible records is not actually complete, it is just long.
The second is to bring across one person’s history — usually whoever was tracking the shared costs — and to treat the other record as personal. This gives continuity on rent, energy, groceries and childcare, which is the part that has any analytical value, without merging two sets of lunches.
The third is a full merge of both records. This is worth doing if you genuinely intend to compare this year with last year at the household level, and it is the only option that requires the rest of this page.
Find the overlap and pick a rule for it
The overlap is any period where both people were recording. Inside it, three kinds of duplication occur, and they need different treatment.
The same purchase recorded twice
One person pays for the weekly shop, mentions the amount, and the other records it too because it is a household cost. In a merged file this appears as two entries with the same date and the same amount, or nearly the same amount if one of them rounded. Same-day plus same-amount matching finds most of these, and the ones it misses are usually a day apart because somebody recorded it the next morning. Widening the match to a two-day window catches those at the cost of a few false positives you will have to eyeball.
Reimbursements recorded as spending
One person pays for a weekend away and the other transfers half. If the transfer is recorded as an expense by the sender and as income by the receiver, the household net is correct by accident but every category total is wrong. If it is recorded as an expense on one side and nothing on the other, the household has apparently spent one and a half times what it spent.
The rule that fixes this permanently: money moving between two people in the same household is not income and it is not an expense. It is a transfer, and it does not belong in a household ledger at all. Strip every internal transfer out of both files before merging, and strip out cash withdrawals at the same time, because a withdrawal is also just money changing location.
The same subscription on both sides
Annual and monthly subscriptions get recorded by whoever pays and sometimes also by whoever uses. These are easy to find because they repeat: list every recurring line in both files, and any name appearing in both is a duplicate until proven otherwise.
Reconcile the two category schemes before importing, not after
Two people who each invented a taxonomy will have between them roughly forty categories, and about a third of them will be near-synonyms: shopping and household, eating out and restaurants, transport and travel, kids and childcare. Merging first and tidying later means editing hundreds of entries by hand.
The cheaper order is to agree a small shared scheme first — five to eight top-level categories is plenty for a household — write down a mapping from each old category to a new one, and apply the mapping to each file before it goes in. Anything you cannot map confidently goes to a single catch-all rather than to a guess, because a wrong category is worse than an unclassified one: it is invisible.
This is also the moment to decide the awkward boundaries once, while nobody is annoyed. Is a takeaway groceries or eating out? Is the annual car insurance transport or insurance? Are term-time lessons education or activities? The answers do not matter much. Having two different answers matters a great deal.
The order of operations
- Agree the shared category scheme and write the mapping down where both of you can see it.
- In each file separately, delete internal transfers, reimbursements between the two of you, and cash withdrawals treated as spending.
- Apply the category mapping to each file, sending anything doubtful to a catch-all rather than to a guess.
- Import the file belonging to whoever recorded the shared costs. That one becomes the spine of the household book.
- Import the second file, then sort the overlap period by date and scan for pairs with the same or nearly the same amount within two days of each other. Delete one of each genuine pair.
- Compare a single month you both remember well against a bank or card statement for that month. Not to make it match exactly, which it will not, but to confirm the merged total is in the right region rather than fifty percent high.
- Write down the date the merged book becomes authoritative, and stop editing anything before it.
That last step is the one that saves the most time later. Without a stated boundary, somebody will keep tidying the old months forever, and the figures will change every time either of you looks at them.
What the merged history is and is not good for
A merged record is reasonably reliable for large, regular, single-source costs: rent or mortgage, energy, council tax, insurance, childcare, subscriptions. Those are recorded once by one person and are hard to duplicate.
It is much less reliable for small, frequent, cash-like spending, where the two of you had different recording habits and different tolerances for skipping things. A comparison that says your household spent less on coffee in the spring may be describing a change in coffee or a change in who was bothering to record it. If the merged history is going to be used to argue about anything, it is worth saying out loud which of the two categories the number falls into.
Frequently asked questions
Should we merge our old records or start fresh?
Starting the shared book from a date is the least work and produces the fewest errors, and it is the right choice unless you specifically intend to compare this year with last year at household level. A merged history assembled from two incompatible schemes is long rather than complete.
How do I find duplicate entries when combining two ledgers?
Sort the overlap period by date and look for pairs with the same or nearly the same amount within a two-day window, since one person often records the next morning. Recurring lines such as subscriptions are easier: list every repeating item in both files, and any name in both is a duplicate until you confirm otherwise.
Should transfers between partners be included in a household ledger?
No. Money moving between two people in the same household is neither income nor expense, and recording it makes every category total wrong. Strip internal transfers and cash withdrawals out of both records before merging; the purchase is the expense, the transfer is just money changing hands.
What do we do about two different category schemes?
Agree a small shared scheme first, five to eight top-level categories, and write a mapping from each old category into it. Apply the mapping to each file before importing rather than tidying afterwards, and send anything doubtful to a catch-all instead of guessing, because a wrongly categorised entry is invisible.
Is merged history reliable enough to compare years?
For large regular costs such as rent, energy, insurance and childcare, reasonably so. For small frequent spending it is not, because the two of you had different recording habits and a drop may reflect who was bothering to record rather than what was bought. Say which kind of figure you are looking at before drawing a conclusion from it.
Try it for one month
Fambook gives a household one shared ledger: anyone can add an entry in seconds, every entry says who spent it, and the month adds up in one place instead of two. Records with no signal and syncs afterwards. Recording, categories, budgets, statistics, CSV import and export, sync and sharing for two people are free — the subscription only buys you less typing.