The Average American, 2008–2026

Doing fine on paper, with no cushion underneath.

Who they are

Who they are. The statistical American is 39, lives in a suburb in a household of about 2.5 people with $87k of income, has some college but no four-year degree, and works a service job. Median household income has grown about 27% after inflation since 2012, most of it in 2013–2019; it then stalled for five years before rising 2.6% in 2025 to $87,460, the highest in the series. Roughly 58% are non-Hispanic white, 20% Hispanic, 13–14% Black and 6–7% Asian; about 6 in 10 identify as Christian and 3 in 10 as unaffiliated. Married-couple households have fallen from 71% of all households in 1970 to 47%, and 29% of households are one person. Every panel below should be read against a wide spread around that median: the top and bottom thirds look like different countries.

Cash cushion

Reading the cash trend. All three series improved steadily from 2012 through 2019–2021, then flattened. The Fed's $400 measure hit its best-ever reading in 2021 (pandemic stimulus and forced saving), gave most of it back in 2022 as inflation hit, and has sat at 37% for four straight years. Unbanked households roughly halved over the period. Bankrate's $1,000 question dipped to a low in 2021, drifted in the low 40s, and reached a series high of 47% in the 2026 report.

Do they invest?

Investing. Roughly 6 in 10 adults now own stock in some form, most of it through 401(k)s and IRAs rather than brokerage accounts. Gallup's series bottomed at 52% in 2013 and 2016 after the financial crisis and climbed to 62% by 2024–2025. The Fed's SCF shows 54% of families with a retirement account in 2022 (the highest since 2010) and a record jump in direct stock ownership to 21%, driven by the 2020–2021 retail-trading boom. Ownership is still lopsided: among families that have a retirement account, the median balance was about $87,000 in 2022, and the bottom half of stock-owning households holds only a small fraction of total equity wealth.

Own or rent?

Housing. About two-thirds of US households own their home and one-third rent, and that split has barely moved in 15 years: 65.5% owned in 2012, the rate bottomed at 63.4% in 2016, briefly spiked to 66.6% in the 2020 pandemic buying rush, and has drifted back down to 65.0% by mid-2026 under high mortgage rates. The headline hides a steep age gradient: under-35 householders own at roughly 37%, 35–44 at about 61%, and 65+ at nearly 80%. Younger buyers gained ground in 2016–2022 (34.5% → 39%) and have been losing it since.

Eat out, order in, or cook?

Food. By dollars, Americans now spend more on restaurants and takeout than on groceries: away-from-home food rose from 51% of food spending in 2012 to 57% in 2019, crashed to 52% in 2020 when dining rooms closed, and has climbed back to 56% in 2025. But dollars overstate the shift, because a restaurant meal costs several times a home-cooked one once labor, taxes and tips are included. By frequency, cooking still dominates: 76% of adults cook at home at least a few times a week, versus 34% getting takeout weekly, 27% eating at a sit-down restaurant weekly, and only 17% ordering delivery weekly. Nearly half (48%) never order delivery at all; it is concentrated among affluent young urbanites, 68% of whom order weekly.

Do they travel, and how?

Travel. Most Americans travel, but the car is the default and flying is a roughly half-the-population activity. About 45% of adults flew at least once in 2025, essentially flat since 2019 after the 22% pandemic trough, with 87% having flown at some point in their lives; personal trips make up about three-quarters of air travel. The last nationwide count of long-distance trips (50+ miles, BTS) put 90% in a personal vehicle, 7% by air, 2% by bus and under 1% by train, and among people who travel, 55% say most of their 2025 trips were by car versus 43% by plane. Vacation-taking is sensitive to money: Deloitte found 48% of Americans planned a summer trip with paid lodging in 2024, 53% in 2025, and 45% in 2026, a six-year low, with households earning $100k+ now 55% of summer travelers.

Gas car or EV?

Cars. Overwhelmingly gas. Plug-in EVs are about 2.5% of the roughly 290 million light vehicles on US roads (7.5 million cars) in early 2026, so 39 of every 40 cars still burn gasoline or diesel, hybrids included. New-car buying shifted faster: plug-ins went from 0.4% of sales in 2012 to 2% by 2018 and 9–10% in 2023–2025, but growth stalled and 2025 was the first year battery-EV sales fell, sliding to 6% of sales in the first half of 2026 after the federal tax credit expired in September 2025. The fastest-growing choice is the conventional hybrid, at 13% of 2025 sales and 16% by mid-2026. Roughly 8% of households own no vehicle at all, concentrated in big cities.

Health insurance

Health insurance. Yes, about 92% are covered, and most of that is through work. The uninsured rate fell from 14.8% in 2012 to 9.4% in 2015 as the ACA's marketplaces and Medicaid expansion took effect, drifted to 9.2% in 2019, then fell to a record-low 7.9% in 2023 on pandemic-era Medicaid rules and enhanced marketplace subsidies. It ticked up to 8.2% in 2024 as states resumed Medicaid eligibility checks. The Census Bureau's count for 2025, published in September 2026, put 7.9% of people (26.7 million) uninsured for the whole year, statistically unchanged from 2024. About 54% of people get coverage through an employer, 20% through Medicare, 17% through Medicaid, and 10% buy it themselves, mostly on the ACA marketplaces. The 2026 picture is likely worse: enhanced ACA subsidies expired at the end of 2025, premiums for subsidized enrollees roughly doubled, and marketplace sign-ups fell by over a million, so the uninsured share is expected to rise again when 2026 data arrive. Being insured is also not the same as being able to afford care: high deductibles mean many insured households still face the $400-emergency problem shown above.

Why spending holds up despite higher prices

Why the consumer keeps spending. Prices are up about 44% since 2012 (CPI), yet inflation-adjusted consumer spending is up 46% and has grown in every year except 2020. Five things explain the resilience. Jobs: unemployment has been at or below 4.5% for a decade outside the pandemic, so most households have a paycheck. Real pay: median full-time earnings have outpaced inflation by about 13% since 2012, with the biggest gains at the bottom of the wage scale in 2021–2023. Wealth: household net worth has nearly tripled to $189 trillion, driven by home equity and stocks, and the roughly two-thirds who own homes or retirement accounts spend more when they feel richer (the "wealth effect"). Cheap locked-in debt: 2010s-era fixed mortgages mean debt payments take only 11% of disposable income, below every pre-2008 reading. And a drawdown of buffers: the saving rate has fallen from 7–8% to under 5% (4.1% in August 2026), and revolving credit has reached a record $1.35 trillion, so part of the strength is households spending their cushion. The same data flags the limits: card delinquencies (2.9%) are back to 2011 levels, the bottom third has no savings to draw on, and the $100k+ households that hold the wealth now account for the majority of discretionary spending, so aggregate strength coexists with a stretched median.

Prices vs pay: CPI, ECI and the Big Mac

Prices vs pay. The Consumer Price Index (CPI) tracks what a fixed basket of goods and services costs; the Employment Cost Index (ECI) tracks what employers pay for a fixed mix of jobs, wages plus benefits, so it is the cleanest measure of pay growth because it is not distorted by who is working. From 2012 to 2020 pay grew faster than prices every year (ECI 2–3% vs CPI 1–2%), which is where the decade's real-income gains came from. In 2021–2022 prices ran ahead, 8% CPI against 4.9% ECI in 2022, and households lost roughly 5% of purchasing power in two years. Pay has outpaced prices again since 2023, and by 2026 cumulative compensation is up 51% since 2012 against a 44% rise in prices, so the average worker is ahead of where they started, but only just and only recently. The Big Mac is the everyday version of the same comparison: The Economist's index puts the US price at $6.17 in 2026, up 92% from $3.21 in 2008, well ahead of CPI (+54% over the same span) because fast-food prices carry labor costs that rose faster than goods. Measured in work rather than dollars, though, the burger has barely moved: a median full-time worker needed 10.7 minutes of pay for one in 2008, 12.9 at the 2014 peak, and 11.9 today, because wages rose almost as fast as the menu. The mood panel below makes more sense in that light: people remember the 2022 loss and the 44% higher price level far more than the slow catch-up.

Below the average: mood, wages and stress by group

Below the average. These six panels split the aggregates by who is affected. Consumer sentiment tells the mood story: the Michigan index averaged 96–98 in 2017–2019, collapsed to 59 in the 2022 inflation shock, and at 52 in 2026 is lower than at any point in the 2008–09 recession, even though spending keeps rising, a gap economists call the "vibecession." Wages explain part of why spending held: the lowest-paid quarter of workers saw 7% raises in 2022 and out-earned inflation for three years, the fastest low-wage growth on record, but that advantage has now reversed, with the bottom quarter at 3.6% and trailing the top quarter for the first time since 2014. Credit stress shows where the buffer ran out: the share of card balances 90+ days late has climbed from 7.3% in 2016 to 13% in 2026, close to the 13.5% peak of 2010 when the financial crisis was still washing through, and new serious delinquency among under-30 borrowers runs at about 10% of balances a year, double the rate for people in their sixties. The pattern across all six is consistent with the rest of the page: the aggregate consumer is fine, the young and low-paid are not.

Online or in-store?

Shopping. Nearly everyone shops online sometimes, but by dollars most retail is still physical: e-commerce was 5% of US retail sales in 2012, 11% in 2019, jumped to 14.5% in 2020, and has crept up to 17% in 2026, meaning about $5 of every $6 is still spent in a store, gas station or car dealership. Stated preference matches: 45% of consumers prefer in-store, 28% online, 27% mix both, and 65% visit a store at least weekly. The split is really by category: services and small, standardized goods (internet plans, small electronics, books, phones) are bought mostly online, while groceries, cars, furniture and appliances stay overwhelmingly in-store. The two channels blur, since 86% of shoppers research online before buying in person.

How they spend their day

How they spend their day. Averaged over all seven days and everyone aged 15 and over, the American day in 2025 was 9.0 hours of sleep, 5.2 of leisure, 3.3 of work and 2.0 of household chores. Work looks small because the average includes weekends, students and retirees; full-time workers put in 8.1 hours on the days they work. Television, streaming included, is still half of all leisure at 2.6 hours a day, and 74% of people watch on a given day, but it has slipped from 2.84 hours in 2018. The time has moved to screens people control: games and leisure computer use rose from 20 minutes a day in 2008 to 37, while face-to-face socializing fell from 43 minutes to 35 and reading from 20 to 16. Age splits the audience. Teenagers now spend slightly more time gaming and on computers (1.5 hours) than watching TV, while people 75 and over watch 4.4 hours; adults 35–44 have the least leisure of anyone, 3.9 hours. Two shifts matter for anyone selling to households. Time spent on food preparation and cleanup has climbed from 31 minutes a day to 43, the highest in the series, which fits the cook-at-home pattern above. And 34.5% of workers did some or all of their work at home on days they worked, down from the 38% peak in 2021 but far above the 24% of 2019. Shopping takes 41 minutes a day on average; 35% of people buy consumer goods on a given day and spend about 52 minutes doing it. Women average 2.4 hours a day on household activities against 1.6 for men, and men 5.6 hours of leisure against 4.8.

Who funds the government?

Who funds the government. In the US it is people, directly. Of $5.2 trillion in federal revenue in fiscal 2025, half came from the individual income tax and another third from the payroll taxes that fund Social Security and Medicare, so 84% of federal money comes off paychecks and personal tax returns. Corporate income tax supplied about 9%, down from 20–25% in the 1950s and 1960s, and even a record tariff year raised under 4%. China is the reverse: its largest tax is the value-added tax (39% of 2025 tax receipts), then corporate income tax (23%), with individual income tax at 9%. Count China's social insurance contributions, which its budget keeps separate, and they become the largest single line (about 32% in the OECD's 2023 accounting), with employers typically contributing 20–28% of payroll. So the average American sees tax on a pay stub and files a return every April, while the average Chinese worker meets the state mostly inside prices and payroll deductions. Two cautions. Who remits a tax is not who bears it: VAT is collected from firms but built into consumer prices, and research finds employers pass their payroll-tax share on as lower wages. And many US businesses are pass-throughs whose profits are taxed on their owners' personal returns, which inflates the individual share and shrinks the corporate one. The US panel is federal only; states and cities rely mainly on property and sales taxes.

How the states fund themselves. States tax spending about as much as income, which makes them very different from Washington. Across the 50 state governments in fiscal 2025, personal income tax supplied 34% of tax revenue and general sales tax 31%, with selective sales taxes (fuel, tobacco, alcohol, insurance premiums) adding 14% and corporate income tax 11%. The average hides how differently states do it. Oregon raises 64% from personal income tax and has no general sales tax; Florida, Texas and South Dakota raise more than 60% from general sales tax and levy no personal income tax at all. Resource states tax what comes out of the ground: severance taxes are 47% of North Dakota's tax revenue, 44% of Alaska's and 31% of New Mexico's. New Hampshire leans on corporate income tax (33%), Delaware on license taxes, chiefly its corporate franchise tax (39%), and Vermont on a statewide property tax for schools (29%). These are state-government taxes only, so property tax, the main local tax, mostly does not appear.

Caveats and data

Fed figures are from the Survey of Household Economics and Decisionmaking (SHED), asked each fall about the year shown. FDIC surveys run every two years. Bankrate figures are dated by report year (survey fielded the prior December); values before 2021 are taken from Bankrate's earlier annual reports and should be treated as approximate. The LendingClub/PYMNTS "living paycheck to paycheck" self-report series only began in 2021 (52% at its April 2021 low → 61% mid-2022 → about 62% through 2026), so it is not charted here. The Survey of Consumer Finances runs every three years (2025 results are due in late 2026); the 2013 and 2016 SCF points are taken from the Fed's earlier bulletins. The BTS long-distance mode split dates from the 2001 National Household Travel Survey, the most recent nationwide long-distance count; the Talker Research mode figures sample people who travel, not all adults. Plug-in sales shares for 2012–2019 are computed from AFDC annual plug-in sales over total light-duty sales; 2020–2023 are Argonne/industry estimates; 2024–2025 are Argonne's published shares (9.9%, 9.1%). The 2020 spike in real median earnings is a composition effect (low-wage workers lost jobs first), not a raise. ACS uninsured figures for 2012–2022 are from the Census Bureau's published 1-year estimates (no 2020 1-year release); 2023–2024 from ACS table S2701. Series that begin after 2008 (Fed SHED 2013, Bankrate 2015, A4A flying survey 2015, Census under-35 homeownership 2016) simply start later. Atlanta Fed wage-quartile series are 12-month moving averages of median hourly wage growth from matched CPS respondents; NY Fed delinquency series are from the Consumer Credit Panel (a 5% sample of Equifax credit files) and the 2026 values cover Q1–Q2. The Big Mac series averages The Economist's January and July readings (one reading for 2008, 2009 and 2011); "minutes of work" divides it by BLS median usual weekly earnings (LES1252881500Q) over a 40-hour week. Federal revenue shares are final FY2025 receipts as published by USAFacts and CBPP; customs is $195 billion of $5.2 trillion and the last bar is the remainder (excise about 2%, estate 0.6%, other fees). China's 2025 shares are computed from the Ministry of Finance annual fiscal report (tax revenue ¥17.64 trillion); "all other" nets import VAT, import consumption tax and customs (¥2.06 trillion) against export rebates (¥2.13 trillion). The 2023 China panel uses the OECD's accounting, which counts social security contributions as tax and covers all levels of government, so it is not strictly comparable with the federal-only US panel. State tax shares are from the Census Bureau's 2025 Annual Survey of State Government Tax Collections as compiled by Pew's Fiscal 50, rounded to one decimal, with licenses and other taxes combined. Census classifies Oregon's corporate activity tax as a general sales tax, and Washington's capital gains tax and New Hampshire's interest-and-dividends tax as personal income tax; South Dakota has no corporate income tax, and the 2% Census reports there is its bank franchise tax. Time-use figures are from the BLS American Time Use Survey (2025 results, released June 2026; trend series from the BLS database). They count only each person's main activity at any moment, so watching TV while eating is recorded once and phone use alongside another activity is mostly missed; averages cover all days of the week and all civilians 15 and over; minute values are converted from hours and rounded. BLS published no 2020 annual estimates because collection was suspended that spring, so the lines run straight from 2019 to 2021. Refreshed October 2, 2026: household income, poverty and coverage figures are from the Census Bureau's September 2026 release for 2025, and the income series is restated in 2025 dollars, so earlier years differ from the 2024-dollar figures shown before. The saving rate and household net worth were re-pulled from FRED and now run higher for 2021–2026 (saving rate) and slightly lower in most years (net worth) than the vintage shown before. The Census ACS 2025 uninsured estimate has been delayed, so that panel still ends at 2024. FRED series used: LES1252881500Q, CPIAUCSL, ECIALLCIV, UMCSENT, PCEC96, TNWBSHNO, LES1252881600Q, UNRATE, TDSP, PSAVERT, MEHOINUSA672N, ECOMPCTSA.

This page is a port of the original dashboard. Its companion is The Average Chinese, 2008–2026.