Methodology
Cost & Benefits copies dollar limits from primary sources and refuses to fill a blank with a plausible number. Every verified figure has a source URL and a last-checked date. See the source list.
Benefits
Poverty guidelines are the 2026 HHS figures from the January 15, 2026 Federal Register. Alaska and Hawaii use their own tables. A household larger than eight adds the published per-person amount.
SNAP uses the USDA fiscal year 2027 monthly gross (130 percent) and net (100 percent) limits, effective October 1, 2026, not a homemade division of the annual guideline. If income is over the gross limit, the result is unlikely unless someone is 65 or older or disabled, in which case the gross test can be waived and the result is maybe. State broad-based categorical eligibility can raise the gross limit. Those state percentages are not stored.
Medicaid adult expansion is 138 percent of poverty, the statutory 133 percent plus the 5 percent income disregard, and only in states KFF listed as having adopted expansion as of October 2026. Wisconsin and Georgia are marked partial because they have not adopted the full expansion and their substitute limits are not stored. Children at or under 138 percent are treated as inside the federal Medicaid floor. CHIP ceilings above that are unverified, so the result is maybe.
WIC uses 185 percent of poverty plus a categorical test for pregnancy or a child under 5. School meals use 130 percent for free and 185 percent for reduced-price, the factors in the 2026–27 FNA notice. For these guidelines the products are already whole dollars, which matches the notice’s round-up rule.
LIHEAP cannot exclude a household solely for income under 110 percent of poverty, and grantees may go up to the greater of 150 percent of poverty or 60 percent of state median income. Income at or under 110 percent is likely, income inside 150 percent is maybe, and the state median-income option is unverified.
The EITC dollar table, including maximum credits and completed phase-outs, is Revenue Procedure 2025-32 for tax year 2026. The credit estimate phases in to the maximum at the earned-income amount and phases out on a straight line to the completed amount. The childless age rule of 25 through 64 is flagged because the revenue procedure does not restate it. Investment income was not asked. The Child Tax Credit maximum is $2,200 and the refundable amount used in the statute is $1,700. The $200,000 and $400,000 phase-outs were not restated in the revenue procedure, so income above them is maybe or unlikely rather than a firm denial.
Marketplace results use the pre-enhancement range of 100 to 400 percent of poverty because the enhanced credits expired January 1, 2026. Premium dollars are not estimated. In expansion states, income at or under 138 percent is pointed to Medicaid instead.
Lifeline uses 135 percent of the 2026 poverty guidelines, or eligibility through SNAP or Medicaid when those come back likely. The Affordable Connectivity Program has no verified successor in this dataset. SSI uses the 2026 federal benefit rate and the simplified monthly income limits in SSA publication 05-10003. Resources are not collected, and that caveat is shown on every SSI result.
State program cards describe where a state administers SNAP, Medicaid, energy aid, and child care assistance. Their income thresholds are null and marked unverified.
Cities and towns
Town pages use the Census American Community Survey 2024 5-year estimates for places: population, median gross rent, median household income, and median home value. Every place in the 50 states and D.C. with at least 5,000 people is included. If that list is under 15,000, smaller places with a published rent or income are added, largest first.
The county is the 2024 Census county that contains the mapped center of the place. HUD fair-market rent is that county FY 2027 figure. In New England, a town name in the same county uses that town HUD row. If a county has more than one rent and the town does not match, the page says so and does not average the rents.
BEA 2024 regional price parities are used for the metro area when the county is in a metro that BEA publishes. Otherwise the page uses the state index and says so. Electricity is the EIA state residential price for the month named on the page. Gasoline is the EIA annual all-grades price, and that workbook only covers some states.
Town comparison pages are not every pair of towns. Each town is paired with nearby towns and with the largest places in its state, and the 25 largest places are paired with each other. Other pairs can be opened from the search box, but those pages ask search engines not to index them. The alphabetical URL is the permanent one. The reverse order redirects.
Town, affordable, benefit-by-town, and county pages render when someone asks for them, then stay cached. They are not all built up front, so a deploy can finish within a normal build window.
Cost of living
The metro list is the 100 largest metropolitan statistical areas in the Census Bureau Vintage 2024 population estimates, limited to the 50 states and D.C. Puerto Rico is excluded because these benefit rules are not written for it.
Rent dollars are HUD FY 2027 Fair Market Rents from the county file. Each metro is matched to the HUD FMR area whose name contains the metro title or starts with the principal city. When HUD splits a metro, the principal-city area is shown and the rents are not averaged. The area name is printed on the comparison. Census ACS 2024 median gross rent is shown beside it. The two measures are different: FMR is a 40th-percentile gross rent for a standard unit, and the ACS figure is the median rent actually paid.
Place-to-place prices use BEA regional price parities for 2024, released February 19, 2026. All items, goods, housing, utilities, and other services are stored. Goods are the groceries and goods comparison. BLS CPI is a time series and is not used as a place-to-place price level. “Other services” mixes transportation with health care and other services, so it is not labeled as a transport price. Transport is the ACS mean travel time to work, in minutes, for workers who did not work from home.
Childcare prices are not in the verified dataset. The field exists and is null. The pages say so.
The price-level equivalent salary is origin salary times destination all-items RPP divided by origin all-items RPP. The rent equivalent uses the same formula with 2-bedroom FMRs. The after-tax equivalent finds a destination gross salary whose after-tax pay, divided by the destination price parity, matches the origin. Federal tax uses the 2026 brackets and standard deduction. Employee FICA is 6.2 percent of wages up to the 2026 Social Security base of $184,500, plus 1.45 percent Medicare, plus 0.9 percent Additional Medicare Tax over the unindexed $200,000 or $250,000 thresholds.
State income tax uses Tax Foundation’s brackets as of January 1, 2026. A published standard deduction or personal exemption is subtracted when it is a deduction, or taken off the tax when it is a credit. Washington’s capital-gains tax is not applied to wages. Phase-outs, tax-benefit recapture, deductibility of federal tax, and local income taxes beyond the 2023 average add-on rates in the Tax Foundation footnote are not fully modeled. Sales tax is the combined state and population-weighted local rate as of January 1, 2026, and it does not model grocery exemptions. Property tax is median real estate taxes divided by median home value from ACS 2024, which is an effective rate, not a millage.
Privacy
The benefit form keeps a copy in the browser so a refresh does not wipe it. When results appear, we also save an anonymous summary: state, household size, pay range, age ranges, and which programs look likely, maybe, or unlikely. We do not save a name, exact pay, or medical details. Comparison pages save the two places you opened. Read the privacy page. Illustrative households on the public pages are fixed examples. The newsletter signup is a placeholder and stores nothing.