Why Zero-Based Budgets Break on Unplanned Expenses, and the One Category That Fixes It

Zero-based budgeting asks you to give every unit of income a specific job before the month begins, so that income minus assignments equals zero. It is internally coherent, it is the method behind several popular apps, and plenty of people swear by it. It also has a characteristic way of collapsing a few weeks in, and the collapse is usually blamed on the person rather than on the design.

This post argues that the design has a hidden requirement, shows the published data on how often that requirement is violated, and sets out the one change that lets the method, or any method, survive contact with an ordinary year.

How the collapse happens

Take a hypothetical. Say you are four weeks into a carefully built zero-based budget, and a dental bill for $180 arrives. It belongs to no category. The method’s answer is correct as far as it goes: move money from elsewhere. So you take some from food and some from clothes.

The trouble starts the following week. You still need to eat, so the food adjustment was never real. You spend what food actually costs, the category goes negative, and you patch it from somewhere else. By month-end, actual spending bears little relationship to the plan, and you have spent several evenings maintaining a document that has become fiction. This is the point at which a budget gets abandoned: in this example, after about five weeks.

Nothing in that story is laziness. The method did what it was designed to do. The design simply assumes something that is not true for most households.

The hidden requirement

Zero-based budgeting needs the unexpected to be rare. When a surprise arrives once or twice a year, shuffling money between categories is a small correction. When one arrives most months, every month becomes a rebuild, and rebuilding a budget over and over is how people stop having one.

So how rare are surprises? The best public figures are American, and they are not reassuring.

Finding	Figure	Source
Households that had at least one financial shock in a 12-month period	60%	Pew Charitable Trusts, Survey of American Family Finances (7,800+ households), 2015
Median cost of a household’s most expensive shock	$2,000	Same
Households that struggled to make ends meet after their most expensive shock	55%	Same
Adults who would cover a $400 emergency with cash or its equivalent	63%	Federal Reserve, Economic Well-Being of U.S. Households in 2025 (published May 2026)
Adults who could not pay a $400 emergency expense by any means	12%	Same
Adults with three months of expenses set aside	55%	Same

Sources: Pew, “How Do Families Cope With Financial Shocks?” and Federal Reserve, Savings and Investments chapter. Checked September 17, 2026. The Pew survey is a decade old; I have not found a more recent study that asks the same question at the same scale.

Pew counted shocks such as a major car or home repair, a hospital visit, a pay cut or a job loss. Those are the large ones. If six in ten households meet at least one of those in a year, the smaller kind, such as a replacement phone charger, a vet visit, a school trip, a parking fine, is more or less a monthly event. A budget with no line for them is a budget that assumes something which happens most months will not happen this month.

The category that fixes it

The fix is one extra category. Call it “things I have not thought of.” It has four rules.

1. It gets a real allocation every month, sized from your own history as described below. A token amount defeats the purpose.
2. It is not earmarked. The moment you name what it is for, it has stopped covering the unnamed.
3. It is not called savings. This matters more than it sounds. A category labelled savings gets defended, and dipping into it feels like failure. This one needs to be spendable with no sense of having broken something.
4. Whatever is unspent at month-end moves to savings. The money is not wasted if the month is quiet. It just was not savings until the month was over.

This is different from an emergency fund. An emergency fund is a stock of money for the $2,000 event. The unthought-of category is a monthly flow for the $40 to $200 events that are individually unpredictable and collectively routine. You want both, and the second is what keeps the monthly plan honest.

It is also different from a sinking fund, where you save monthly toward a known future cost such as car insurance or holiday gifts. Sinking funds handle the irregular-but-foreseeable. If a bill catches you out every year at the same time, it belongs in a sinking fund, not here.

How to size it

Look backwards. Pull as many months of bank and card statements as you can stand, ideally six to twelve. Mark every purchase that was unplanned at the time it happened. Total them per month.

The individual items will look random: the dentist, a washing machine part, a last-minute train fare. The monthly totals will be much steadier than the items suggest. The randomness is in which thing breaks, not in whether something does. It is the same reason an insurer can price a policy without knowing whose boiler will fail.

Take the median month, not the average, so one large event does not inflate the figure. That number is your starting allocation. Events large enough to distort the average belong to the emergency fund.

I am deliberately not offering a rule-of-thumb percentage. I have not found one with a credible source behind it, and the number that matters is sitting in records you already have. If you do not have clean records, a thirty-day purchase log will give you a first estimate, with the caveat that one month is a small sample. For people whose income moves as much as their spending, the same backward-looking approach applies to the income side, as covered in budgeting with lumpy income.

Granularity versus attention

The wider point about budgeting methods is that they are mostly variations on two decisions: how granular your categories are, and how strictly you hold them. The name on the method matters less than whether its granularity matches the attention you actually have.

Method	Granularity	Strictness	Attention required	Where it tends to fail
Zero-based	High	High	Weekly or more	Unplanned expenses force repeated rebuilds
Envelope (cash or digital)	Medium	High	Weekly	An empty envelope mid-month, with a real need still to pay for
50/30/20	Low	Low	Monthly	Too coarse to show where a leak is
Pay-yourself-first	Minimal	Low	Near zero once automated	No visibility into spending at all

This table is my characterisation of how the methods are usually described by their proponents, not a measured comparison.

Zero-based sits at high granularity and high strictness. That costs a lot of attention. If you have the attention, I think it is the most informative of the four, because every unit of money has to be argued for. If you do not, a coarser system you keep is worth more than a precise one you abandon.

One workable arrangement splits the two jobs. Try a four-category summary day to day: committed bills, everyday essentials, reserves and goals, and flexible spending. Keep transaction detail and a bill calendar alongside it. Then, once a year, run a detailed tracking month to get the resolution a coarse system cannot give. A low-effort system that runs continuously, plus a high-effort audit that runs occasionally, covers most of what either extreme offers.

If you want to keep zero-based

You can. Add the unthought-of category as a first-class line, size it from your history, and treat a raid on it as the system working rather than failing. Keep a few sinking funds for the irregular bills you can foresee. And set a rule for yourself that a category can be rebalanced once a month, not every time reality disagrees with the plan. If you find yourself rebalancing weekly, the categories are too fine for the life you are actually living.

This is general information about budgeting methods, not personal financial advice.