What’s left, and what it becomes

What is real data, and what is a what-if

Real data. Every year from 2000 to 2024 uses that year’s published BLS Consumer Expenditure Survey table, which ranks U.S. households into five income groups (quintiles). From it come income before and after taxes, the income cut-offs between groups, household size, spending on six categories (including education), and benefit income. For the five quintiles, the model reproduces BLS’s averages exactly.

What-ifs. Anything finer than quintiles (deciles, the top 1%, custom splits) is estimated from a smooth income curve inside each quintile. Childcare, benefit amounts other than 100%, expense edits, the minimum tax, investing and the wealth taxes in Fig. 3 are scenarios you control. Fig. 3’s net worth itself is measured data from the Federal Reserve.

Breaks in the dataBefore 2004, BLS income figures cover only households that fully reported their income; from 2004 BLS fills in missing income. From 2013, BLS estimates taxes with a tax model instead of asking households, which is why BLS after-tax income for the top group dips that year. By default this site uses CBO’s consistent federal income tax rates instead, so the dip disappears. To compare, turn on “After tax” and choose a method under Advanced. BLS doesn’t publish the top group’s 2023 public assistance figure, so 2022’s is used. BLS also published no after-tax income for 2024, because it didn’t update its tax model that year; the BLS-based tax methods use each group’s 2023 tax rate for 2024.

Projections · 2025–2026BLS hasn’t published household data for these years yet. They start from 2024 and grow every income and cost with consumer prices (CPI-U), so they show what happens if everything simply kept pace with inflation. The 2025 price average covers 11 months because BLS published no October 2025 index; 2026 covers January–August. The 2026 S&P 500 return is year to date through September 14. Projected years are shaded and dashed.

An outside checkThe Federal Reserve’s survey of household economic well-being found that 63% of U.S. adults would cover a $400 emergency expense with cash or its equivalent in 2024 (the same as in 2023), leaving more than a third who couldn’t. That fits this site’s finding that lower-income groups have little or nothing left after essentials, though the survey measures adults, not households, and savings as well as income.

Definitions

• Left after essentials (the “surplus”). Income under your settings minus what households in each group actually spend on the categories you select, including childcare and any custom expense. It is not a minimum budget: higher-income groups spend more on the same categories. “Insurance & pensions” includes Social Security payroll tax and retirement contributions, so part of it is saving (about $27,300 a year for the highest 20% in 2024, versus about $700 for the lowest). Use “Edit expense amounts” to change the BLS figures.
• Benefits & transfers. Social Security and pensions; public assistance, SSI and SNAP; unemployment, workers’ compensation and veterans’ benefits; regular support payments such as child support.
• Education. BLS’s education spending: tuition, fees, textbooks, supplies and equipment for schools and colleges. It excludes student-loan payments and doesn’t count tax-funded public schooling. Off by default.
• Happiness plateau line. $75,000 of yearly household income in 2008–09 dollars (Kahneman & Deaton, 2010), adjusted with CPI-U for each year: about $110,000 in 2024. A 2023 reanalysis found this plateau only for the least happy 15–20% of people, at about $100,000; for most people happiness keeps rising with income. It is a rough reference, not a threshold for any individual.
• Childcare. Children × price per child, adjusted to each year with the day care and preschool price index.
• Per person. The group’s average amount divided by its average household size, plus any children added.
• Minimum tax. A tax on income, not wealth (under Advanced). “Simple” is a what-if rate above an exemption. “U.S.-style AMT” uses each year’s married-filing-jointly AMT exemption, phase-out and 26%/28% rates, with no deductions or filing status, so it is a rough group-level estimate.
• Where the wealth is (Fig. 3). Household net worth from the Federal Reserve’s Distributional Financial Accounts, by wealth or by income group, 2000–2025. Wealth taxes there are what-ifs: flat charges the rate on net worth above a threshold; the Zucman-style minimum requires covered households to pay at least the rate × total net worth, counting taxes already paid. They are applied to each group’s average, ignore avoidance, and use no official revenue score, so treat them as rough orders of magnitude.
• Investing. Returns accrue on the prior balance; the year’s contribution is added at year-end. A shortfall contributes $0. No taxes, fees or withdrawals: it’s a portfolio built only from this surplus, not a measure of anyone’s actual wealth.