Everyday Money
Budget Calculator
Sort a month of take-home pay into needs, wants and saving, see the three shares it actually produces, and compare them against the 50/30/20 framework and what US households measurably spend.
How the month divides
Needs share
65.2%
$3,910 of $6,000 take-home pay. The framework allows $3,000, a figure the average US household also exceeds — housing, transport, food and healthcare alone were 71.2% of its spending in 2024.
The three shares
- Needs $3,910.00 against a $3,000 benchmark, $910 above it.
- 65.2%
- Wants $950.00 against a $1,800 benchmark, $850 below it.
- 15.8%
- Saving $700.00 against a $1,200 benchmark, $500 below it.
- 11.7%
The month as a whole
- Take-home pay
- $6,000.00
- Assigned 92.7% of take-home pay is assigned to something. The three shares above are each rounded to one decimal, so they need not total exactly 100.
- $5,560.00
- Unallocated Not yet assigned to anything, and therefore not saved. This is the money that absorbs whatever comes up — and the cheapest share to change, because directing it costs nothing.
- $440.00
- Short each month
- —
$6,000 of take-home pay divides into 65.2% needs ($3,910), 15.8% wants ($950) and 11.7% saving ($700). $440 is unassigned, which is not the same as saved — it is the money that absorbs whatever comes up. Directing it to saving would take that share from 11.7% to 19% without changing anything you spend, which is usually the cheapest improvement available on a page like this. Needs run $910 above the framework's $3,000. That is the usual position rather than a diagnosis — housing, transport, food and healthcare alone came to 71.2% of average US household spending in 2024, against a benchmark that allows 50% — and it is the column that moves least, since four of its lines are decided on a timescale of years. The 50/30/20 split used here is the framework from Warren and Tyagi's All Your Worth (2005), not published guidance — no agency sets these percentages, and the CFPB's own budgeting material names none at all. It is one month of arithmetic: nothing here is projected, compounded, or adjusted for costs that arrive annually.
The average household does not fit inside this benchmark either
Shares of average US household spending, from the Bureau of Labor Statistics Consumer Expenditure Survey for 2024. These four categories are the bulk of anyone's needs column, and together they come to 71.2% of spending — against a framework that allows 50% of take-home pay for needs. Measured against pre-tax income instead, the same four are 53.7%, still over the line.
| Category | Share of spending |
|---|---|
| Housing | 33.4% |
| Transportation | 17.0% |
| Food | 12.9% |
| Healthcare | 7.9% |
| All four | 71.2% |
| What the framework allows for needs | 50% |
So a needs share above 50% is the ordinary result, not a diagnosis. It also explains why the useful response is rarely "spend less on needs": four of those five lines are decided on a timescale of years — a lease, a car, an insurance policy — while the wants line is available this month.
What this calculates
This adds up a month of spending in three groups and divides each by your take-home pay, which turns a list of amounts into three percentages you can compare against something. The needs group is what you would struggle to stop paying: housing, transport, food at home, insurance and the contractual minimum on any debt. Wants is everything discretionary, entered as one total. Saving is money you deliberately move somewhere — into savings, into a retirement account, or onto debt above the minimum. It reports the three shares, the gap between each share and the 50/30/20 framework in dollars rather than in percentage points, and whatever is left unallocated. That last figure is kept separate on purpose: money not yet assigned to anything is not saved, and adding it to the savings line would report a household as saving 20% when it is saving 12% and losing the rest to whatever comes up.
How it works
needs share = needs total / take-home pay wants share = wants total / take-home pay saving share = saving total / take-home pay unallocated = take-home pay - (needs + wants + saving) gap vs needs target = needs total - 0.50 x pay gap vs wants target = wants total - 0.30 x pay gap vs saving target = saving total - 0.20 x pay A positive gap is more than the framework allows for that group; a negative gap is less. Two properties worth stating, both measured: The three gaps always sum to exactly the negative of what is unallocated, because 50 + 30 + 20 accounts for the whole of your pay. So a budget with nothing left over has gaps that cancel out, and one with money left over is under target somewhere by exactly that amount. The three shares are each rounded to one decimal, so they need NOT total 100. A fully allocated budget can print 99.9 or 100.1 and nothing is wrong. The page never claims the three add up, because at one decimal they do not reliably do so. There is no rate, no projection and no compounding here. This is one month of arithmetic.
The division takes a second. What it is worth measuring against is the part that needs care, and there are two honest answers to it — one from a book, one from a survey.
Start with where 50/30/20 actually comes from, because almost nothing that repeats it says. It is the framework in "All Your Worth: The Ultimate Lifetime Money Plan", written by Elizabeth Warren and Amelia Warren Tyagi and published by Free Press in 2005. That is its provenance in full: a book, by two authors, twenty years ago. It is not published by a regulator or an agency, and the Consumer Financial Protection Bureau's own budgeting material names no percentages whatsoever — its method is to record your income, log your spending by category, and reconcile the two, leaving every allocation decision to you. So the split is a well-known opinion rather than a standard, and this page treats it as a line to measure against rather than a test to pass.
Now the measured answer, which is where this page departs from every other budget calculator. The Bureau of Labor Statistics runs the Consumer Expenditure Survey, and in 2024 the average US consumer unit spent $78,535 against $104,207 of income before taxes. Of that spending, housing took $26,266, transportation $13,318, food $10,169 and healthcare $6,197. Those four come to $55,950 — 71.2% of all spending, and 53.7% of pre-tax income. Housing alone was 33.4%. Housing, transport and food together were 63.4%.
Read that against a framework that allows 50% for needs. The average American household is over the needs cap on four categories, before a single discretionary dollar is spent, and before counting childcare or anything else that is difficult to stop paying. This is not a marginal miss. It means that for a very large number of households the 50% figure is not a target being missed through poor discipline; it is arithmetically out of reach at their income and their rent. A calculator that reports "your needs are 65%, the rule says 50%" and stops has told them nothing except that they are failing a test the average household also fails.
What follows from that is a different way to read your own result. On this page's default figures — $6,000 of take-home pay, $3,910 of needs, $950 of wants, $700 of saving — the shares are 65.2%, 15.8% and 11.7%, with $440 left unallocated. The needs share is 15.2 points over the framework, which in dollars is $910. The useful question is not how to get $910 out of that column, because four of its five lines are housing, transport, food and insurance and none of them moves this month. The useful question is which of the three numbers can actually change, and by how much.
Usually the answer is the saving share, and the lever is the unallocated money rather than the wants column. $440 is sitting unassigned in the default scenario: moving it into saving takes that share from 11.7% to 19% and changes nothing else about the household's life. That is the single highest-value edit available on this page, and it is invisible on any calculator that quietly counts leftover money as savings — because then the figure already reads 19% and there is nothing to notice.
That is why the unallocated line is kept separate and named. Money left at the end of a month is not saved. It is available, which is a different state, and in practice it is the money that absorbs the car repair and the higher grocery bill. A budget is a set of decisions, and the distinction between money you assigned somewhere and money that merely survived is the difference between a plan and a hope.
A word on which column a payment belongs in, since it changes the answer more than people expect. Two lines do the most damage when they are misfiled. Restaurant meals are wants, not groceries — food at home and food away from home are separate lines in the BLS data for good reason, at $6,224 and $3,945 in 2024, and the second is discretionary in a way the first is not. And debt payments split: the contractual minimum is a need, because missing it has consequences, while anything above the minimum is saving, because it is a deliberate decision to build net worth. Filing a $500 payment with a $200 minimum entirely under needs overstates that column by $300 and understates saving by the same amount.
Finally, the framework's own definition matters: it is specified on after-tax income. Running it on gross pay makes every share look better than it is, by the size of your tax and payroll deductions, and that is a large distortion rather than a rounding error — the BLS average household's $78,535 of spending was 75.4% of its pre-tax income. If you enter your salary here instead of what lands in your account, the three shares below are all flattering and none of them is wrong arithmetic, which is the hardest kind of error to notice.
A worked example
$6,000 of monthly take-home pay: $1,850 housing, $720 transport, $650 groceries, $430 insurance and healthcare, $260 in minimum debt payments, $950 of wants, and $700 deliberately saved.
The needs column totals $3,910. Against $6,000 of take-home pay that is 65.2%, with wants at 15.8% and saving at 11.7%. The three do not reach 100 because $440 was never assigned to anything.
Measured against the framework, needs are $910 over the $3,000 the 50% figure allows, wants are $850 under their $1,800, and saving is $500 under its $1,200. Those three gaps sum to -$440, which is exactly the unallocated amount — that identity always holds, because 50 + 30 + 20 accounts for the whole of your pay.
The first instinct is to attack the needs column, and it is the wrong instinct. Look at what is in it: $1,850 of housing, $720 of transport, $650 of groceries, $430 of insurance, $260 of minimum payments. Four of those five cannot change this month, and the fifth changes by maybe $80 with real effort. There is no version of this month where $910 comes out of that column, and a calculator that points at the 65.2% and says "reduce your needs" is issuing an instruction that cannot be followed.
Now look at the $440. It is unassigned, which means it is not saved — it is available, and available money is what absorbs the unexpected car repair. Move it into the saving line and that share goes from 11.7% to 19%, with nothing else about the household changing at all. No spending is cut. The saving share nearly doubles because the money was already there and merely undirected. That is the highest-value edit on this page, and note that it is completely invisible on any calculator that folds leftover money into savings for you: those report 19% from the start, and the reader never learns there was a decision to make.
Worth seeing what misfiling does, because it is the most common error here. Suppose the $500 actually paid on debt each month goes in entirely as a need, rather than $260 of minimum in needs and $240 of overpayment in saving. Take the extreme version — file a $560 payment under needs and drop $300 from saving — and the shares read 70.2% needs and 6.7% saving. Identical money, identical month, and the budget now looks close to unfixable when what actually happened is that $300 of deliberate net-worth building got recorded as an obligation. Restaurant meals filed as groceries do the same thing in the other direction, making the needs column look inescapable when part of it is discretionary.
One more comparison worth making, and it is the reason this page does not simply grade you. In 2024 the average US household spent 33.4% of its outlays on housing, 17.0% on transport and 12.9% on food — 63.4% on three categories. Add healthcare at 7.9% and the four come to 71.2% of spending, or 53.7% of pre-tax income. A household at 65.2% needs is not an outlier against a 50% standard; it is close to the middle of a population where the standard is not reachable. That does not make the framework useless, but it changes what it is for: a direction to move in, not a line you are failing to clear.
What this assumes
- Income means take-home pay: what reaches your account after tax and after payroll deductions. The 50/30/20 framework is defined on after-tax income, so entering gross salary inflates all three shares by the size of your deductions.
- Anything deducted from your pay before you see it is outside this calculation entirely — including health premiums paid through an employer and retirement contributions made by payroll deduction. Those never reached the take-home figure, so counting them again as spending or saving would double-count them.
- The 50/30/20 split is the framework from Warren and Tyagi’s All Your Worth (2005), used here as a benchmark because it is widely known. It is a book’s proposal, not published guidance, and no agency or regulator sets these percentages.
- Needs are taken to be housing, transport, food at home, insurance and healthcare you pay yourself, and the contractual minimum on debt. Where a payment sits is a judgement, and moving one line between groups changes the shares without changing the money.
- Debt payments are split: the minimum is a need, and anything above it counts as saving, on the basis that a voluntary overpayment builds net worth. Filing the whole payment as a need overstates that column.
- Money left over is reported as unallocated rather than as saving. Saving means money deliberately moved somewhere; a surplus that has not been assigned is available rather than saved.
- Each share is rounded to one decimal place independently, so the three need not total exactly 100 even when every dollar is allocated. A fully allocated budget can display 99.9 or 100.1, and the drift runs in both directions about equally often.
- One month is assumed to be representative. Annual and irregular costs — insurance paid yearly, car registration, holidays, medical deductibles — do not appear unless you divide them by twelve and enter them.
- The BLS spending figures quoted on this page describe average US consumer units in 2024. They are context for reading your own shares, not a target, and averages conceal a very wide spread.
What it does not model
- This cannot tell you whether your budget is good. It reports three percentages and the distance from one book’s framework; whether that distance matters depends on your income, your rent, your household and your obligations, none of which is arithmetic.
- The framework it measures against is very likely unreachable for a large share of US households. Measured: housing, transport, food and healthcare alone were 71.2% of average household spending in 2024 against a 50% needs allowance, so being over that line is the normal case rather than a diagnosis.
- Wants are entered as a single total, so nothing here identifies which discretionary spending to change. That is deliberate — itemising ten wants fields produces a longer form and the same total — but it means the page cannot point at a specific subscription.
- Where a line belongs is your judgement and the page cannot check it. Groceries against restaurants, and minimum against overpayment on debt, are the two that most change the result, and a plausible-looking set of shares can come from a consistent misfiling.
- Nothing here is forecast. There is no rate, no projection and no compounding: this is one month divided three ways, and it says nothing about where the saving share leads over a decade.
- Irregular and annual costs are invisible unless you convert them to monthly amounts yourself. A budget that balances in eleven months and breaks in the twelfth looks identical here to one that works.
- Variable income is not modelled. For commission, shift or self-employed pay there is no single take-home figure, and running this on a good month produces shares that do not describe the year.
- No priority order is suggested between saving, investing and clearing debt. A 24% card balance and a savings account are not interchangeable uses of the same dollar, and choosing between them is outside what a budget split can answer.
- This is not advice about what to spend. Showing what your money currently does, and how that compares with a published framework and with measured averages, is not a recommendation about what any share of it should be.
Questions
Where does the 50/30/20 rule actually come from?
From a book. It is the framework in "All Your Worth: The Ultimate Lifetime Money Plan" by Elizabeth Warren and Amelia Warren Tyagi, published by Free Press in 2005 (LCCN 2005042483, ISBN 074326987X). That is the whole of its provenance: two authors, one book, twenty years ago. It is not issued by a regulator, an agency or a standards body, and the Consumer Financial Protection Bureau’s own budgeting material names no percentages at all — its method is to track what comes in, log what goes out by category, and reconcile the two, leaving the allocation entirely to you. None of that makes the split worthless; it is a clear, memorable starting point and being deliberate about three groups is better than being deliberate about none. But it is one well-known opinion rather than a standard, which is why this page shows your distance from it beside what US households measurably spend, instead of scoring you against it.
My needs are way over 50%. Is my budget broken?
Probably not, and the measured figures are worth seeing before you conclude otherwise. In the Bureau of Labor Statistics Consumer Expenditure Survey for 2024, the average US consumer unit spent 33.4% of its outlays on housing, 17.0% on transportation and 12.9% on food — 63.4% on three categories. Add healthcare at 7.9% and those four come to 71.2% of all spending, which is 53.7% of average pre-tax income of $104,207. So the average American household is over a 50% needs allowance on four categories before buying anything discretionary. If your needs share is 60 or 65 percent you are not an outlier failing a standard; you are near the middle of a population for whom that standard is arithmetically out of reach. The honest reading is that the needs share mostly measures your rent and your commute, both of which are decided on a timescale of years, and that the figures worth acting on this month are the other two.
Should I use my gross salary or my take-home pay?
Take-home pay, and this is not a small distinction. The framework is defined on after-tax income, so running it on gross salary makes all three of your shares look better by exactly the size of your tax and payroll deductions — a household whose deductions are a quarter of pay would see a 65% needs share report as around 49%, which is flattering and useless. Enter what actually lands in your account. The corollary is that anything deducted before you see it should be left out of the spending lines as well: health premiums paid through an employer and retirement contributions taken by payroll deduction never reached the take-home figure, so entering them again as an expense or as saving double-counts them. If most of your retirement saving happens by payroll deduction, your saving share here will understate your real rate, and that is the correct behaviour for a calculation based on money you actually receive.
Why is money left over not counted as saving?
Because it has not been saved yet — it is available, which is a different state. This is the most consequential design decision on the page. On the default figures $440 is unallocated: it is not committed to anything, and in practice unassigned money is what absorbs the car repair, the higher grocery bill and the birthday nobody planned for. Reporting it as saving would tell a household it saves 19% when it deliberately saves 11.7% and loses the rest to whatever comes up. Keeping it separate also makes visible the single best edit available on this page: move that $440 into the saving line and the share goes from 11.7% to 19% with nothing else about the month changing, because the money was already there and merely undirected. On a calculator that folds surplus into savings for you, the figure already reads 19% and you never learn there was a decision to make.
Are restaurant meals a need or a want?
A want, and it is one of the two lines that most often distorts a result. Food at home and food away from home are separate lines in the BLS data — $6,224 and $3,945 respectively in 2024 — and the distinction is real: you must eat, and you need not eat out. Filing restaurant spending under groceries inflates the needs column and makes the budget look more fixed than it is, which matters because the needs share is the number people conclude they cannot change. The other line to watch is debt. The contractual minimum is a need, because missing it has consequences; anything you pay above the minimum is a deliberate choice to build net worth and belongs in the saving line. Filing a whole $500 payment with a $200 minimum entirely under needs moves $300 out of saving and into obligations, and the measured effect on this page’s defaults is severe — one such misfiling takes the shares to 70.2% needs and 6.7% saving on identical money.
Why do my three percentages not add up to 100?
Two reasons, and both are expected. The first is that they genuinely do not have to: if some of your pay is unallocated, the three shares sum to less than 100 by exactly that proportion, which is the point of reporting the unallocated figure separately. The second applies even when every dollar is assigned — each share is rounded to one decimal place independently, so a fully allocated budget can print 65.2 plus 15.8 plus 19.0 and land on 100.0, or land on 99.9 — or 100.1, since the drift runs both ways about equally often. That is measured, not theoretical. Nothing on this page claims the three add up, because at one decimal place they do not reliably do so, and a calculator that forced them to total exactly 100 would have to misreport one of the three to get there.
What should I do if the budget does not balance at all?
The page tells you the size of the gap and deliberately does not tell you how to close it, because that answer depends on facts it does not have. What it can do is separate the arithmetic from the judgement. If spending exceeds take-home pay, the shortfall is being covered by something — savings, credit, or family — and identifying which is the first useful step, since a gap filled by a credit card compounds and one filled from savings does not. Look at where the overage sits before deciding: an over-target needs column is mostly housing and transport, which move on a timescale of years rather than weeks, while an over-target wants column is genuinely available this month. And if the needs column alone exceeds your income, no reallocation inside this page resolves it — that is a question about income or about housing, not about budgeting, and the arithmetic here is only useful for showing which of the two it is.
Is a 20% saving rate the right target?
It is the figure the framework proposes and this page has no basis for endorsing it as correct for you. What is worth knowing is what that 20% is meant to cover: in the framework it includes both saving and debt repayment above the minimum, so a household aggressively clearing a card is doing the same work as one filling a savings account. Whether 20% is right depends on how much you have already accumulated, how far off retirement is, what your employer contributes, and whether you carry high-rate debt — a 24% balance and a savings account are not interchangeable uses of the same dollar, and this page takes no view on which comes first. For what it is worth as context, the BLS figure for personal insurance and pensions was 12.5% of average household spending in 2024, most of it Social Security deductions rather than voluntary saving, which suggests that a deliberate 20% of take-home pay is well above what the average household manages.
Sources
- BLS — Consumer Expenditures, 2024 (news release) (opens in a new tab)
Every measured spending figure on this page, for reference year 2024, released 19 December 2025. Average annual expenditures per consumer unit of $78,535 against average income before taxes of $104,207. Housing $26,266 (33.4% of spending), transportation $13,318 (17.0%), food $10,169 (12.9%), personal insurance and pensions $9,797 (12.5%), healthcare $6,197 (7.9%), entertainment $3,609 (4.6%), apparel and services $2,001 (2.5%). Food at home $6,224 and food away from home $3,945, which is the basis for treating restaurant spending as discretionary. The derived figures this page quotes are arithmetic on those published numbers: the four need-like categories total $55,950, which is 71.2% of spending and 53.7% of pre-tax income; housing, transportation and food together are 63.4% of spending; total spending is 75.4% of pre-tax income. The release notes that shares do not sum to 100.0 because of rounding.
- CFPB — Budgeting: how to create a budget and stick with it (opens in a new tab)
The claim that no US consumer-protection agency publishes a percentage split for budgeting. This page was checked directly and names no percentages whatsoever — no 50/30/20 and no numeric proportion for needs, wants or saving. Its method is process-based: record all income including multiple jobs and benefits, log and sort spending by category, map bill due dates, then consolidate into a worksheet to see whether income covers expenses "while also having enough to save". Allocation decisions are left to the reader throughout. It is cited to support a statement about what guidance does NOT exist, which is the page’s editorial spine.
- Formula reviewed
- Formula version
- 1
Version 1 means the formula has not changed since this page was published. If it changes, this number moves and the change is described here.