The 50/30/20 rule, explained honestly
The 50/30/20 rule says: spend at most 50 % of your net income on needs, 30 % on wants, and put 20 % toward savings and debt. It's the most famous budgeting rule in the world — genuinely useful as a starting point, and genuinely misleading if you treat it as a law. Here's how to use it well.
How the rule works
Start from your net income — what actually lands in your account after tax. Then sort every expense into three buckets:
- Needs (50 %): housing, utilities, groceries, insurance, transport to work, minimum debt payments, childcare.
- Wants (30 %): eating out, streaming, travel, hobbies, upgrades — everything you could stop without your life falling apart.
- Savings & extra debt payments (20 %): buffer, investments, pension top-ups, and everything you pay on debt beyond the minimum.
The power of the rule is not precision — it's that it forces the one conversation most budgets skip: how much of my income is already spoken for before the month starts?
A worked example
Net income 35 000 kr a month: the rule suggests at most 17 500 kr for needs, about 10 500 kr for wants, and 7 000 kr for savings and extra debt payments. On a 2 400 € net income: 1 200 € needs, 720 € wants, 480 € savings.
Where the rule breaks
High-rent cities. In Oslo, Stockholm, Copenhagen, Amsterdam or Munich, housing alone can eat 40–50 % of a normal net income. If your needs land at 65 %, that's not a personal failure — it's your local housing market. The rule's job is to make that visible, not to shame you for it.
Low incomes. The lower the income, the larger the share that needs take. A rule built on percentages quietly assumes there's room to allocate; sometimes there isn't, and the honest answer is that the problem is income or housing cost, not discipline.
High-tax, high-service countries. In the Nordics, some things Americans budget as «needs» (healthcare, education, much of childcare) are already inside your taxes. Nordic needs percentages can legitimately run lower — which frees room for a higher savings share than the classic 20 %.
Adapt it: your own three numbers
Use 50/30/20 as a first measurement, then set your own split. Common honest variants: 60/25/15 during expensive life phases (small children, one income), 45/25/30 for high earners aiming to save aggressively, or a fixed-sum version — «needs are 19 400 kr, wants get 6 000 kr, the rest saves itself».
The only real rules are: the three buckets must sum to 100 %, savings must be a planned number rather than whatever is left, and the split must survive contact with a normal month — a budget you break every month is a wish, not a budget.
Measuring it without a spreadsheet
The tedious part of 50/30/20 is classifying every transaction. This is exactly what modern budget apps automate: ZivaFinance categorises your transactions automatically, splits fixed from variable costs, and shows the month against your budget as it happens — so you can see your real percentages in minutes, then decide what they should be.
Frequently asked questions
Is 50/30/20 based on gross or net income?
Net — what lands in your account after tax. In countries where pension contributions or union fees are taken before payout, use the amount you actually receive.
Do minimum debt payments count as needs or savings?
Minimum payments are needs — you must make them. Everything you pay beyond the minimum counts in the 20 % savings-and-debt bucket, because it's building your net worth.
What if my needs are way over 50 %?
First check for reclassified wants hiding in needs (car class, housing size, food delivery). If needs are still high, the levers are structural: housing, transport, insurance and debt costs — or income. The rule's job is to point at the real problem, not to prescribe guilt.
Is the rule outdated in 2026?
The percentages are less universal than when Elizabeth Warren popularised the rule, mostly because housing has outgrown incomes in many cities. The three-bucket idea — needs, wants, planned savings — is as useful as ever; just calibrate the numbers to your city and life phase.