From Here, Anywhere

How the numbers work

Money to bring: how long the money lasts, the bonds that hold it, and how the plan is tested.

What the number means

The money to bring is the amount that, held in bonds, pays your monthly spending for the years you choose and reaches zero at the end. If you choose the rest of your life, the plan shows the amount whose interest alone pays for your spending, so the money itself is never spent. Prices rise until the year you move, at 3.4% a year, so moving later needs more money on the first day.

Which bonds hold the money

The plan holds US inflation-linked Treasury bonds (TIPS), at 2.91% a year above inflation (the 10-year rate on October 6, 2026), with a 3% cost each time money is exchanged into local currency. Where we can verify a country's own government bond yield against the government's own figures, and a buyer from abroad can purchase that bond, the plan uses that bond instead and there is no one-time exchange cost. Today that is Brazil, where a buyer from abroad needs a representative in the country, a registration and a bank that keeps the bonds. For every other country with its own currency, either we could not confirm a current yield or we could not show that a newcomer can buy the bond before moving, so the plan uses US bonds. We check the other countries one at a time.

Countries that use the US dollar

For countries that use the US dollar, such as Panama and Ecuador, you buy US inflation-linked Treasury bonds at the same 2.91% above US inflation. Prices in these countries rise at the country's own rate of inflation, which is lower than in the United States, so spending grows more slowly than the bonds' inflation protection. We count that gap in your favor. There is no exchange cost, because the money never changes currency. Interest is taxed at the US rate.

Monthly spending and household size

Spending is for the household in a typical month, rent included, in today's prices, scaled for household size by the number of people raised to the power of 0.7. How each level's spending is set is on How levels are worked out. Tax is on Taxes and data sources.

How we test a plan against history

On Will the money last? we replay the plan starting in every year of US history we have, from 1928 on. Each starting year lives through the real stock returns and price rises that followed, and we count how many starting years made it. The bond part earns today's yield on TIPS. Stocks earn what a broad US stock index actually earned. The mix is reset to your chosen split each year. "For life" spends only the income, so the money itself is kept. Dividend choices are tested on the dividends of the whole S&P 500 since 1928, because the lists of long-time dividend raisers only go back a few decades.

With the local currency switched on, each starting year also lives through that country's real exchange rate and price changes since 1960, from World Bank data. Changing dollars into local money costs 3% each time, except in countries that use the US dollar. Some of these countries were much poorer in 1960. As they grew, their prices rose toward US prices, so a dollar bought less there each decade.

Local government bonds are not tested against history, because they depend on that government paying. For every country except Brazil the plan holds US TIPS, so there is no local-bond choice on that page.

Social Security and other income

On Will the money last? you can add Social Security and other income. Social Security figures use the 2026 benefit formula and rise each year with US prices, like TIPS. Without a payout you enter, the Social Security estimate comes from the 2026 formula for 35 years at a stated yearly pay in today's dollars: about US$100,000 for the top 20% of earners and US$150,000 for the top 10%. The top 5% and top 1% groups are high earners whose pay is at or above the earnings cap of US$184,500 a year, so they are assumed to receive the maximum benefit for the age they start (US$4,152 a month at 67), even though not every person in those groups has the maximum. A spouse who receives half gets half of your full-age amount, less the reduction for claiming before 67 (32.5% of your full-age amount at 62, rising to 50% at 67), and nothing extra for waiting past 67. The cut choice starts at 22% from 2033. Other income can be a fixed amount, which buys a little less each year, or one that rises with prices. Income is counted in US dollars, so with the local currency switched on it carries the same exchange rate effect as TIPS.

How the Social Security claim page was checked

The page that shows your check at each claim age uses the 2026 Social Security formula (Social Security Administration, 2026) and the June 2026 Trustees Report for the possible cut. Its answers were compared with the matching spreadsheet on 112 sets of numbers, recalculated in LibreOffice, and with a second, separate calculation written in Python. Every check, total and break-even age matched to the cent. The 2026 maximum of US$4,152 a month at full age is from the Social Security Administration's 2026 fact sheet. Dates: October 8, 2026.

How the investment history page was checked

The page that shows what stocks, bonds, cash or a mix did since 1928 uses Aswath Damodaran's yearly returns (S&P 500 with dividends, the 10-year Treasury, 3-month Treasury bills) and the consumer price index. Its answers were compared with the matching spreadsheet on 113 sets of numbers (first and last years, period length, stock share and amount, including one-year spans, the last two years of history, a period longer than the years chosen, zero and very large amounts), recalculated in LibreOffice, and with a second, separate calculation written in Python. Every growth rate, strongest and weakest year, fall from a high and ending value matched. Dates: October 9, 2026.

How the other calculators were checked

Each calculator below was compared with its matching spreadsheet on 120 or more sets of numbers (every boundary of an age or income group, zero, the lowest and highest values the page accepts), recalculated in LibreOffice, and with a second, separate calculation written in Python. Where a page differs from its spreadsheet on purpose, the difference is named in its entry here and in the dated notes on How it works.

  • What do people my age and income spend? The page that shows what households of your age and income spend uses the Bureau of Labor Statistics Consumer Expenditure Survey for 2024, by age of the household head and by income before taxes, multiplied by 1.057 to put 2024 dollars into today's, and the survey's yearly results from 1984 to 2024. Its answers were compared with the matching spreadsheet on 120 sets of numbers (ages and incomes at every group boundary, the lowest and highest values the page accepts, and with and without the household's own spending), recalculated in LibreOffice, and with a second, separate calculation written in Python. Every line of spending, share, group total, comparison and trend figure matched. The figures are averages for a group, not what one household spends. Spending does not include income tax, and "personal insurance and pensions" includes Social Security tax and pension savings, which is why it falls after 65. Age and income are shown separately because the survey does not publish them crossed. Incomes are self-reported and high incomes are under-reported, so spending as a share of income runs high at the top. The bars and tables compare different ages at one time; they do not follow one household as it ages. Dates: October 9, 2026.
  • Am I saving enough? The saving page projects your retirement accounts with one steady growth rate above inflation, pay level after inflation, saving added at the end of each year, no taxes or fees, and a share of savings drawn each year, then adds Social Security by the 2026 formula (on your average pay, not your earnings record). It finds the lowest total saving rate (you plus your employer) from 0% to 30% that reaches the income you want, at growth rates of 1% to 5% and at yours. Savings rates of other people are Vanguard's How America Saves 2026 (plan participants, who save more than people with no plan; its income bands do not line up with the Federal Reserve ranges, so the closest band is used). Ranks of retirement accounts use the Federal Reserve survey of 2022 in 2026 dollars (the age and income range only when it holds at least 75 surveyed households). Its answers were compared with the matching spreadsheet on 126 sets of numbers (every age and income boundary, zero pay, zero accounts, stop-work ages already reached, all Social Security choices), the whole Savings table included, recalculated in LibreOffice, and with a second calculation written in Python; the two agreed to 14 digits. The projection is a steady-growth illustration, not a test against past markets; the page 'Retiring into a bad market' does that. Dates: October 9, 2026. The 22% Social Security cut counts only when your Social Security starts in 2033 or later; the spreadsheet applies it to every figure, so the comparison gave the spreadsheet no cut for a start before 2033.
  • Savings check. The Savings check follows the matching spreadsheet step by step: the Fed's 2022 numbers for your level and age (put in today's dollars with a factor of 1.13), each account grown at a planning return (stocks 5% above price rises, inflation-linked bonds 2.91%) with your monthly saving, the plan year by year after you stop (spending after tax, less Social Security and other income, taken from savings), and every past starting year since 1928 turned into the money needed on your last working day. The money needed at your sureness is the matching percentile of those years. The spreadsheet was recalculated in LibreOffice for 126 input sets (its yearly home costs set to 2.9%, its stock history to full precision and its country table refreshed from the site's engine) and agreed with an independent calculation written in Python to 9 digits; the page's method, run in the same Python calculation, gives the page's figures on the same 126 sets. Deliberate differences from the spreadsheet: tax on each retirement year comes from the 2026 federal rules (taxable part of Social Security, standard and age-65 deductions, 0%, 15% and 20% on gains, the 3.8% investment income tax) and the state's rules from the site's tax file, instead of flat rates of 18%, 8% and 15%; the Fed's retirement balance is counted as tax-deferred and Roth is what you type; housing is the level's share of spending (Consumer Expenditure Survey 2024); home costs are the state's median property tax plus 0.5% insurance and 1.5% upkeep, selling costs 6%, and the tax on the gain when you give what you paid (the Section 121 amount, then the 2026 rates); the 22% Social Security cut counts only for benefit years from 2033; full retirement age comes from your birth year. Not counted: a penalty on withdrawals before age 59 and a half, required withdrawals, and the tax saving from deducting mortgage interest. Market examples carry only figures with a dated, named source; municipal and company bonds, the floating rate note yield, money fund managers ranked 6 to 10, bond funds ranked 3 to 11 and their fees, and the 2075 target-date fund are not shown. Dates: October 9, 2026.
  • Rent or buy, year by year and The home you want: buy or rent. The two rent-and-buy pages use national figures: a typical home price (National Association of Realtors median for existing homes, carried back with the Case-Shiller index), typical rent (Zillow, carried back with the Bureau of Labor Statistics rent index, which follows mostly apartments and long-time tenants), and the Freddie Mac 30-year mortgage rate. Their answers were compared with the matching spreadsheets on 126 sets of inputs each, recalculated in LibreOffice, and with a second, separate calculation written in Python; every figure matched. The option to itemize deductions has no spreadsheet; it was checked against the separate calculation only. Insurance, upkeep and selling costs are steady percentages, growth and returns are assumptions you can change, and refinancing is left out. Dates: October 9, 2026.
  • How often you move. How often you move uses national figures: the typical existing home (National Association of Realtors, January 2026), typical rent (Zillow, July 2026), the Freddie Mac 30-year mortgage rate, and a typical stay of 11 years (National Association of Realtors, 2025). The shorter and longer stays (2, 3, 27 and 44 years) come from how long owners have lived in their homes (Census American Community Survey 2024, table B25038), scaled to the typical stay. Each move assumes a new home of the same value, which is the same market; moving to a bigger home or a less expensive place is not covered here. Its answers were compared with the matching spreadsheet on 126 sets of inputs, recalculated in LibreOffice, and with a second, separate calculation written in Python; every figure matched. The option to itemize deductions has no spreadsheet; it was checked against the separate calculation only. Dates: October 9, 2026.
  • Cook at home or eat out? The meal page takes the USDA Economic Research Service food spending per person, at home and away from home, divides by 365 days, the meals a day you choose and the share of meals at home or away (the share of calories: 17.7% in 1977 and earlier, 31.6% in 2007 and later, a straight line between) to get the money cost of a meal. The time cost multiplies the hours of cooking, cleanup and shopping (American Time Use Survey minutes, straight lines between 1965, 2007 and 2024) by the after-tax hourly pay of a factory worker and by the share of that wage you choose; restaurant time works the same with the minutes you give. The value of time is your choice, and 3 meals a day and 15 minutes at a restaurant are assumptions. Its answers were compared with the matching spreadsheet on 126 input sets, recalculated in LibreOffice, and with a second calculation written in Python; the two agreed to 14 digits. Years run 1960 to 2025. Dates: October 9, 2026.
  • Are groceries more expensive than ever? The page that compares the price of the same 11 foods in two years uses average retail prices from the Census Bureau (to 1970), the Statistical Abstract (1971 to 1977) and the Bureau of Labor Statistics (from 1980), the consumer price index, and factory pay for production and nonsupervisory workers (Census to 1970, Bureau of Labor Statistics from 1971; the two series differ by about 4% where they join). Its answers were compared with the matching spreadsheet on 126 sets of years, weekly amounts and birth years, recalculated in LibreOffice, and with a second, separate calculation written in Python; every price, real change, share of pay, hours of work, rank and birth-year figure matched. The basket is one unit of each food, not what a household buys. Milk changes from half a gallon to a gallon in 1998. Some prices are estimates or doubtful Census values (butter 2013 to 2017, coffee 2008, 1952 potatoes, 1956 bacon, each under 1% of the basket). After-tax pay is an estimate for one single worker with federal tax only. No food prices exist for 1919, 1930, 1978 or 1979, and there is no pay data for 1917 and 1918. Figures for 2026 are January to August 2026. Dates: October 9, 2026.
  • Which foods got less expensive, and which cost more?, How many hours of work does it take to afford groceries? and Minimum wage: did it keep up with prices? The page uses the shared historical price-and-pay series (food prices from the Census Bureau, the Statistical Abstract and the Bureau of Labor Statistics; the consumer price index; factory pay for production and nonsupervisory workers, Census to 1970 and Bureau of Labor Statistics from 1971, which differ by about 4% where they join; median full-time earnings from 1979; the federal minimum wage from the Department of Labor). Its answers were compared with the matching spreadsheet on 126 sets of inputs, recalculated in LibreOffice, and with a second, separate calculation written in Python; every figure matched. After-tax pay is an estimate for one worker with federal tax only. The 11-food basket is one unit of each food, not what a household buys; some prices are estimates or doubtful Census values. Figures for 2026 are January to August 2026. Dates: October 9, 2026.
  • What did a new car cost? The new-car page reads the shared file of yearly prices and pay. Weeks of pay divide the price by factory weekly pay after tax; hours divide it by hourly pay after tax (one single worker, federal tax estimate); months of income divide it by the median household income over twelve; gallons divide it by the price of gasoline. In today's dollars multiplies by the 2026 consumer price index and divides by that of the year. The price is the government's average for passenger cars from 1967; the entry-level Ford is held for 1913 to 1927 and 1956 and is a different car. Kelley Blue Book, a company, reports about US$50,000 for all new vehicles; that is the company's own figure. Its answers were compared with the matching spreadsheet on 126 pairs of years, recalculated in LibreOffice, and with a second calculation written in Python; the two agreed to 14 digits. 2026 means January to August. Dates: October 9, 2026.
  • What age can I retire? What age can I retire tries each stop-work age from 50 to 75 on every start year since 1928 with enough history. For each, it works out the savings you would have then, the money needed on day one to pay your spending minus other income and Social Security, and counts the start years in which your savings were enough. Returns are the S&P 500 with dividends and the 10-year Treasury (Aswath Damodaran), after inflation, with the mix reset every year. Social Security uses the 2026 formula and the possible 22% cut from 2033. The page was compared with the matching spreadsheet on 126 sets of choices, every stop-work age compared, recalculated in LibreOffice, and with a second, separate calculation written in Python. Taxes, fees, care costs and changes in spending with age are not included. Dates: October 9, 2026.
  • Retiring into a bad market. Retiring into a bad market replays every start year since 1928 that has enough history for the years you choose. Stock returns are the S&P 500 with dividends and bond returns the 10-year Treasury (Aswath Damodaran, NYU Stern); prices come from the consumer price index. The mix is reset to your stock share every year and spending comes out at the start of each year. The page was compared with the matching spreadsheet on 126 sets of choices, every start year compared, recalculated in LibreOffice, and with a second, separate calculation written in Python. Start years overlap, so the number of start years is not the number of separate tests. Taxes and fees are not included. Dates: October 9, 2026.
  • What will US$1 million be worth in 30 years? The million-dollar page takes Damodaran's yearly table (S&P 500 with dividends, 10-year Treasury bond, December-to-December inflation, 1928 to 2025), resets the stock and bond mix to the share you choose every year, and looks at every past period of the chosen length that ended before the start year. The price multiple gives what the money buys as cash; the mix value multiple gives what it grows to, before and after inflation; the low, middle and high cases are the 10th percentile, median and 90th percentile (the same method as the spreadsheet). A past start year also shows what really happened, if the period ended by 2025. It is a range from past periods, not a forecast: periods overlap, results are before fees and tax, and the United States had one of the strongest stock results of the 20th century. Its answers were compared with the matching spreadsheet on 126 input sets, recalculated in LibreOffice with full-precision table values, and with a second calculation written in Python; the two agreed to 14 digits. Dates: October 9, 2026.
  • What will groceries cost in 10 years? The grocery range page takes every past period of the chosen length (1 to 30 years) that ended by the start year, in which the 11-food basket has a price at both ends, and works out its yearly price growth. The low case is the 10th percentile, the middle case the median and the high case the 90th percentile of those rates (the same method as the spreadsheet), applied to your weekly grocery money; the page also shows the share of a factory worker's after-tax pay and, for a past start year, what really happened. It is a range from past decades, not a forecast: the periods overlap, so the range is narrower than it looks. The basket is one unit of each of 11 foods, with no price for 1919, 1930, 1978 and 1979. Start years run 1931 to 2026. Its answers were compared with the matching spreadsheet on 126 input sets, recalculated in LibreOffice, and with a second calculation written in Python; the two agreed to 14 digits. Dates: October 9, 2026.
  • Retirement: what to do with the home. The home page compares six ways of handling a home in retirement year by year from your age to 100: your spending, housing costs and income are taken from savings at the start of each year, mortgage payments and reverse mortgage payments come monthly through the year, and what is left earns one steady return you choose (6% a year before inflation to start) while prices rise at the site's 3.4%. Selling costs 6%, buying costs and moving costs are shown with the numbers; the tax on a sale follows the 2026 federal and state rules used on the Same market or a new one page (the Section 121 tax-free amount of US$250,000 single or US$500,000 joint, then 0%, 15% or 20%, the 3.8% investment income tax and the state's tax); with no sale there is no tax and the home passes to heirs. The reverse mortgage amount is HUD's principal limit factor for the youngest borrower's age and the expected rate (HUD tables for case numbers from October 2, 2017), less 2% upfront insurance, the lender's fee, other closing costs and the old mortgage; a lender's quote may differ. Property tax is the state's median rate (Census, 2024), 0.9% with no state. Income tax on withdrawals and rent received is not counted, except the optional line for owners who itemize. The workings were compared with the matching spreadsheet on 126 input sets recalculated in LibreOffice (its reverse mortgage factor and the tax on the sale put in), which agreed with an independent calculation written in Python to 9 digits; the same Python calculation, run one month at a time, gives the page's figures, which the page matched on 163 input sets. Dates: October 9, 2026.
  • History test: buy or rent. History test, buy versus rent follows every start year from 1971, when the Freddie Mac mortgage survey begins. Each start year uses that year's average 30-year rate, the typical home price (Case-Shiller national index as published by Aswath Damodaran, carried back from the National Association of Realtors figure for January 2026), the typical rent (Zillow, carried back with the Bureau of Labor Statistics rent index, which follows mostly apartments), and the actual returns of the stocks, bonds and bills you choose. Results are in 2026 dollars using consumer prices. Local markets differ a great deal from the national figures, and income tax, refinancing and the mortgage interest deduction are left out. Its answers were compared with the matching spreadsheet on 126 sets of inputs, every start year included, recalculated in LibreOffice, and with a second, separate calculation written in Python; every figure matched. Dates: October 9, 2026.
  • How long would my savings last, country by country? The runway page ranks the places by how many years your money lasts at your level, household size and stop-work date. It uses the same money engine as the main calculator: for each place it searches for the plan length whose money to bring equals your money, so the result at 20, 25, 30, 35 or 40 years, or for life, is exactly the main calculator's figure. It was compared with the matching spreadsheet on 126 input sets, recalculated in LibreOffice with its table refreshed from a second, independent calculation of the engine written in Python (the spreadsheet's own table is the figures of October 6, 2026, which the site has since changed by up to 3.3% in some places); every result of the three kinds (years, more than 60 years, interest alone covers it) agreed, and the years agreed within 0.2 of a year (the spreadsheet draws a straight line between 18 points). It was also compared with the main calculator's own page on 120 sets of choices, tile by tile. Social Security, pensions and rental income are not counted. Dates: October 9, 2026.
  • Same market or a new one? The home-sale page works out the cash freed by selling your home and buying the same kind of home, a smaller one, a less expensive US place, or buying or renting abroad. The tax on the gain uses the 2026 rules: the gain above the Section 121 tax-free amount (US$250,000 single, US$500,000 joint) is stacked on your other income after the standard deduction and taxed at 0%, 15% or 20% (breakpoints from IRS Revenue Procedure 2025-32), plus the 3.8% investment income tax over US$200,000 single or US$250,000 joint, plus the state's income tax on the gain (Washington's capital gains tax leaves out real estate). Property tax is the state's median rate (Census, 2024), 0.9% with no state; insurance 0.5%, upkeep 1.5%, selling costs 6%, buying costs 3% (8% abroad), moving 1% of the sale price (3% abroad). The spreadsheet's flat 18.8% rate is not used. Its answers were compared with the matching spreadsheet on 126 input sets, recalculated in LibreOffice with the tax of a second calculation written in Python put in; the two agreed to 14 digits. Dates: October 9, 2026.
  • Is it harder to buy a home than in 1970? The home page compares the Census median price of a new home with median household income (1967 to 2025) and factory pay, then the monthly payment on the chosen loan, the share of after-tax income it takes and the years of saving for the down payment. The mortgage rate is the Freddie Mac Primary Mortgage Market Survey yearly average of the 30-year fixed rate; it starts in April 1971, so 1963 to 1970 use the 1971 rate and the page says so; 2026 has no yearly average yet and so no payment. The payment uses the same loan function as the rent or buy pages. The spreadsheet's rate column was wrong in 11 of 56 years and was replaced before the comparison; its verdict showed the letters IF when buying was not harder, which the page replaces with harder, easier and about the same (5% either way). Its answers were compared with the matching spreadsheet (corrected rates) on 126 input sets, recalculated in LibreOffice, and with a second calculation written in Python; the two agreed to 14 digits. Dates: October 9, 2026.
  • What do people like me do? The page that shows what people like you do uses the Federal Reserve Survey of Consumer Finances 2022 by age group and income range (dollars multiplied by 1.13 to reach 2026 dollars): how many households own retirement accounts, pensions, homes, stocks, bonds and money market accounts, what they hold, whether they spend less than, about the same as, or more than they earn (the survey does not record an amount saved), their debts and work. It adds the average share of pay saved by Vanguard plan participants (How America Saves 2026, 2025 data) and Bureau of Labor Statistics 2024 spending and income by age; Vanguard and the Bureau publish their own age and income bands, so each card names the band used (Under 35 uses the 25 to 34 band; 65 to 74 and 75 and older use Vanguard's 65 and older). Groups with fewer than 75 surveyed households fall back to the age group alone. Its answers were compared with the matching spreadsheet on 126 sets of numbers, recalculated in LibreOffice, and with a second calculation written in Python that reads the spreadsheet's Peers sheet; the spreadsheet's own comparison wording was matched too. Dates: October 9, 2026.
  • Where do I rank? The page that shows where a household ranks uses the Federal Reserve Survey of Consumer Finances 2022 (about 4,600 households standing for 131 million): for each age group and income range, the weighted percentile points 1 to 99 of net worth, financial assets, retirement accounts and home equity, with dollars multiplied by 1.13 to reach 2026 dollars. A rank is the number of the 99 points strictly below your amount, so a household with nothing in retirement accounts shows 0%, and 99 is written "at least 99%"; the age and income range is used only when it holds at least 75 surveyed households. The income level (Comfortable, Refined, Exclusive, World-class luxury) is where your income sits against the 80th, 90th, 95th and 99th points of all households. Its answers were compared with the matching spreadsheet on 126 sets of numbers (every age and income group boundary, zero and negative net worth, very large amounts), recalculated in LibreOffice, and with a second calculation written in Python that reads the spreadsheet's curves; the curves themselves were recomputed from the Fed's file in the October 8 audit within 1 dollar. The survey leaves out the very richest 400 families and samples the top less often, so ranks above the 95th point are rough. Dates: October 9, 2026.
  • If one of us dies first. The page that shows what happens to a couple's income when one of them dies first uses the Social Security survivor rules (the survivor gets the larger of the survivor benefit and his or her own benefit; the survivor benefit depends on whether the spouse who died had claimed, with at least 82.5% of the full-age amount if claimed early, nothing before 60, 71.5% at 60 rising to 100% at 67), the pension share you enter, and the Bureau of Labor Statistics finding that one-person households spend 60% of what two-person households spend (61% for households with no earner). Its answers were compared with the matching spreadsheet on 126 sets of numbers (death before and after the claim age, a survivor under 60 and aged 60 to 66, every claim age, spouse or you first, pensions with every survivor share), recalculated in LibreOffice, and with a second, separate calculation written in Python. Every Social Security, income, spending, shortfall and pension figure matched. The 22% cut from 2033, which the spreadsheet does not have, was checked with the second calculation only. Income tax, growth on savings, the one-time death payment of US$255 and a survivor's own benefit that starts later are not included. Dates: October 9, 2026.
  • Have wages kept up with prices? The page that compares pay with prices reads one shared file of yearly figures: the consumer price index of the Bureau of Labor Statistics, factory pay (Census to 1970, Bureau of Labor Statistics from 1971), the federal minimum wage, median weekly earnings (from 1979), the Census median household and family incomes (to 2025), and prices for 11 foods, rent, restaurant meals, new homes, new passenger cars and gasoline. After-tax pay is a federal estimate for one single worker or one married couple (income tax, Social Security and Medicare, no state tax, no credits). A pay measure is ahead of prices when it grew more than 2% faster, behind when more than 2% slower. Its answers were compared with the matching spreadsheet on 126 pairs of years (every first year with no data, reversed years, the same year twice, 2026 with no income figures), recalculated in LibreOffice, and with a second calculation written in Python; the two agreed to 15 digits. Factory pay joins two series in 1970 (a step of about 4%) and covers production workers in manufacturing, not all workers; the car price covers passenger cars only; 2026 means January to August. Dates: October 9, 2026.
  • What did it cost the year you were born? The birth-year page reads the same shared file of yearly prices and pay as the pay-against-prices page. In today's dollars means the birth-year price times the consumer price index of the comparison year divided by that of the birth year. Hours of work divide a price by factory pay after tax per hour (one single worker, federal tax estimate). The basket is one unit of each of 11 foods, not a household's shopping list. No basket exists for 1913 to 1916, 1919, 1930, 1978 and 1979, and the page says so in words. Its answers were compared with the matching spreadsheet on 126 pairs of years, recalculated in LibreOffice, and with a second calculation written in Python; the two agreed to 15 digits. The car price is the government's average for passenger cars only; the entry-level Ford is held for 18 selected years; 2026 means January to August. Dates: October 9, 2026.

Estimates

Every figure is an estimate for planning, based on public data as of October 2026. The full list of what each page assumes is on Disclaimers and what this is not, and the dated changes to the method are on How it works.

Back to the page you came from