Metrics
| Metric | Definition | Source |
|---|---|---|
| Median gross rent | Monthly rent plus utilities, renter-occupied units. | ACS 5-year B250641 |
| Rent to income | Median rent x 12 / median household income. Above 30% is the usual stress line. | Derived from ACS 5-year B25064, B190131 |
| Property tax rate | Median real estate taxes paid by owners as a share of median home value. An effective rate, not the legal millage. | Derived from ACS 5-year B25103, B250771 |
| Property tax paid (median) | Median annual real estate taxes paid by owner-occupied households. Top-coded at $10,000. | ACS 5-year B251031 |
| Home price to income | Median home value / median household income. | Derived from ACS 5-year B25077, B190131 |
| Prices vs US average | Price level for all goods and services; the US average is 100. | BEA Regional Price Parities2 |
| Fair market rent, 2 bedrooms | HUD's 40th-percentile rent for a standard two-bedroom unit. | HUD Fair Market Rents3 |
| HUD area median income | Median family income HUD uses to set housing assistance limits. | HUD Income Limits4 |
| Median home value | Owner-estimated value, owner-occupied units. | ACS 5-year B250771 |
| Home prices, 1-year change | Change in the repeat-sales house price index over the latest year. | FHFA House Price Index5 |
| Home prices, 5-year change | Change in the repeat-sales house price index over five years. | FHFA House Price Index5 |
| Homeownership | Owner-occupied share of occupied homes. | ACS 5-year B250031 |
| Renters paying 30%+ | Share of renter households spending 30% or more of income on rent. | ACS 5-year B250701 |
| Owner costs with a mortgage | Median monthly owner costs for homes with a mortgage. | ACS 5-year B250881 |
| Vacant homes | Vacant share of all housing units. | ACS 5-year B250021 |
| Median year built | Half of homes were built before this year. | ACS 5-year B250351 |
| Median household income | Half of households earn more, half less. | ACS 5-year B190131 |
| Income per person | Total income divided by residents. | ACS 5-year B193011 |
| Unemployment | Unemployed / civilian labor force, five-year average. | ACS 5-year B230251 |
| Unemployment, latest month | Monthly unemployment rate, not seasonally adjusted. | BLS Local Area Unemployment Statistics6 |
| Average annual pay | Average pay per job covered by unemployment insurance. | BLS QCEW7 |
| Pay growth, 1 year | Change in average annual pay over the latest year. | BLS QCEW7 |
| Jobs | Average employment covered by unemployment insurance. | BLS QCEW7 |
| Poverty rate | Residents below the federal poverty line. | ACS 5-year B170011 |
| Mean commute | Average one-way travel time to work. | ACS 5-year B080131 |
| Works from home | Workers who mostly work at home. | ACS 5-year B083011 |
| Population | Total residents. | ACS 5-year B010031 |
| Population change since 2020 | Change from the 2020 estimates base to the latest annual estimate. | Census Population Estimates8 |
| Median age | Half of residents are older, half younger. | ACS 5-year B010021 |
| Bachelor's or higher | Adults 25+ with a bachelor's degree or more. | ACS 5-year B150031 |
| High school or higher | Adults 25+ with a high school diploma or more. | ACS 5-year B150031 |
| Household size | Average people per occupied home. | ACS 5-year B250101 |
| Born abroad | Residents born outside the United States. | ACS 5-year B050021 |
| No health insurance | Civilian residents without health coverage. | ACS 5-year S27011 |
| Homes with broadband | Households with a broadband subscription of any type. | ACS 5-year B280021 |
| Commutes by transit | Workers who mostly take public transportation to work. | ACS 5-year B083011 |
| Moved here in the past year | Residents who lived in another county, another state or abroad a year earlier. | ACS 5-year B070031 |
| Moved from another state | Residents who lived in another state a year earlier. | ACS 5-year B070031 |
| Net movers per 1,000 people | People moving in minus people moving out over one tax year, per 1,000 residents who filed. County-level. | IRS SOI migration data9 |
| Income of households moving in | Average adjusted gross income per tax return of households that moved into the county. | IRS SOI migration data9 |
| Income of households moving out | Average adjusted gross income per tax return of households that left the county. | IRS SOI migration data9 |
| Restaurants and bars per 10k | Food service and drinking places with paid staff, per 10,000 residents (NAICS 722). | Census County Business Patterns10 |
| Shops per 10k | Retail stores with paid staff, per 10,000 residents (NAICS 44-45). | Census County Business Patterns10 |
| Arts, recreation and fitness per 10k | Venues, gyms, golf courses, museums and similar businesses per 10,000 residents (NAICS 71). | Census County Business Patterns10 |
| Doctors' offices and clinics per 10k | Outpatient health care establishments per 10,000 residents (NAICS 621). | Census County Business Patterns10 |
| Child care centers per 10k | Child day care services with paid staff per 10,000 residents (NAICS 6244). | Census County Business Patterns10 |
| Parks per 10k | Parks mapped in OpenStreetMap within the city's footprint, per 10,000 residents. Mapping completeness varies by place. | OpenStreetMap11 |
| Playgrounds per 10k | Playgrounds mapped in OpenStreetMap, per 10,000 residents. | OpenStreetMap11 |
| Grocery stores per 10k | Supermarkets and grocers mapped in OpenStreetMap, per 10,000 residents. | OpenStreetMap11 |
| Restaurants, cafes and bars per 10k | Restaurants, cafes, bars and pubs mapped in OpenStreetMap, per 10,000 residents. | OpenStreetMap11 |
| Libraries per 10k | Public and other libraries mapped in OpenStreetMap, per 10,000 residents. | OpenStreetMap11 |
| Named paths and bike trails | Length in kilometres of named footpaths, cycleways and bridleways mapped within the city's footprint. | OpenStreetMap11 |
| Transit stops per sq mi | Bus, tram and rail stops mapped in OpenStreetMap, per square mile of the city's footprint. | OpenStreetMap11 |
| Nearest National Park Service site | Straight-line miles to the nearest park, monument, historic site or recreation area run by the National Park Service. | National Park Service sites via Wikidata12 |
| National Park Service sites within 60 miles | Parks, monuments, historic sites and recreation areas run by the National Park Service within 60 miles. | National Park Service sites via Wikidata12 |
| National Register listings | Buildings, districts and sites listed on the National Register of Historic Places. | NPS National Register of Historic Places13 |
| Founded or incorporated | Earliest founding or incorporation date recorded in Wikidata. | Wikidata14 |
| Violent crime per 100k | Violent offenses the city police department reported to the FBI, per 100,000 residents. From the FBI's city table (Crime in the United States, Table 8) when the department reported all year; otherwise from its monthly reports, scaled to a full year. | FBI Crime Data Explorer15 |
| Property crime per 100k | Property offenses (burglary, theft, vehicle theft, arson) reported to the FBI, per 100,000 residents. | FBI Crime Data Explorer15 |
| Public schools | Open public schools with an address here. | NCES Common Core of Data16 |
| Public school students | Students enrolled in those schools. | NCES Common Core of Data16 |
| Students per teacher | Enrolled students per full-time-equivalent teacher. | NCES Common Core of Data16 |
| July average high | Average daily maximum temperature in July. | NOAA 1991-2020 normals17 |
| January average low | Average daily minimum temperature in January. | NOAA 1991-2020 normals17 |
| Average temperature | Average of daily mean temperatures over the year. | NOAA 1991-2020 normals17 |
| Annual precipitation | Average yearly rain and melted snow. | NOAA 1991-2020 normals17 |
| Natural hazard risk | Composite of expected losses, social vulnerability and resilience; 0 to 100 against all US counties. | FEMA National Risk Index18 |
| Hurricane risk | Hurricane risk score, 0 to 100 against all US counties. | FEMA National Risk Index18 |
| Tornado risk | Tornado risk score, 0 to 100 against all US counties. | FEMA National Risk Index18 |
| Wildfire risk | Wildfire risk score, 0 to 100 against all US counties. | FEMA National Risk Index18 |
| River flood risk | Riverine flooding risk score, 0 to 100 against all US counties. | FEMA National Risk Index18 |
| Heat wave risk | Heat wave risk score, 0 to 100 against all US counties. | FEMA National Risk Index18 |
| Median air quality index | Median daily AQI over the latest full year; under 50 is good. | EPA Air Quality System19 |
| Good air days | Share of monitored days rated good. | EPA Air Quality System19 |
Citytistics measures
These combine public figures in ways the publishers do not. Each is computed the same way for every place, from figures measured for that place, and ranked against all places that pass the data gate.
| Measure | Formula | Why it matters |
|---|---|---|
| Paycheck stretch | (median household income − 12 × median gross rent) × 100 ÷ price level of everything except housing. That price level is BEA's all-items Regional Price Parity with housing taken out at a 17% basket weight: (RPP all items − 0.17 × RPP housing) ÷ 0.83. | Dollars a typical household has left after rent, in US-average prices. Income and rent alone miss that groceries and services cost more in some places. |
| Pay vs rent | Compound yearly growth of median household income minus that of median gross rent, from the oldest to the newest ACS 5-year release we hold (2015-19 to 2020-24). | Shows whether housing is getting harder or easier to afford, not just where it stands today. |
| Own vs rent gap | Median monthly owner costs for homes with a mortgage minus median gross rent. | What buying costs each month over renting, at today's typical mortgage. |
| Commute hours | Mean one-way commute × 2 × 230 working days ÷ 60 × (1 − share working from home). An estimate: it assumes a full-time year and counts remote workers as zero. | Turns minutes into a yearly cost of time that accounts for remote work. |
| Family fit | Average percentile of six figures, each turned so 100 is better: rent to income (lower), home price to income (lower), students per teacher (lower), homeownership (higher), household size (higher), poverty rate (lower). Needs at least five, all measured for the place. | Cost and space for a household with children. It leaves out school test scores and crime, which we do not hold yet. |
| Growth | Average percentile of population change since 2020 and yearly income growth. Places whose population figure is borrowed from their county have no score. | Where people and pay are both moving up. |
| Places like this, but cheaper | Nearest neighbours on income, rent, home value, education and age, within a quarter to four times the population, whose rent takes at least 3 points less of income. | Answers "where else could I live like this for less". |
2030 projections
Every place page carries a 2030 outlook for median rent, median household income, median home value and population, and two rankings are built on it. These are Citytistics projections of recent trends. They are not predictions of the economy and not advice to buy or sell.
- Trend. A straight line through the logarithm of the place's own series gives its yearly growth rate. Rent, income and home value use the ACS 5-year releases we hold (one point per release); population uses the Census Bureau's yearly estimates.
- Blend. That local rate is mixed with the typical (median) rate across every place we rate. Short survey series are noisy, so the local share is set by testing, not by judgment: we re-ran the method on the first three releases, projected the next three, and kept the mix that missed least.
- Temper. The releases we hold span the fastest rent and price inflation in forty years, so the blended rate eases toward the long-run US rate, closing a quarter of the gap each year. Long-run rates, 1991 to 2024: rent 3.3% a year (BLS consumer price index, rent of primary residence), median household income 3.1% (Census Bureau), home prices 4.4% (FHFA purchase-only house price index). Population is not tempered.
- Project. The latest value is carried forward at the tempered rate to 2030.
- Range. The likely range covers 80% of the misses in that test, one to three years out, widened with the square root of the horizon beyond three years.
| Projection | Own-trend share | Typical yearly change | Places tested | Median miss, 1 / 3 years out | 80% of misses within, 3 years out |
|---|---|---|---|---|---|
| Median rent | 20% | 5.6% | 6,402 | 5.2% / 8.0% | 15.0% |
| Median household income | 20% | 5.4% | 6,507 | 4.7% / 6.0% | 12.6% |
| Median home value | 40% | 8.8% | 6,499 | 10.4% / 16.3% | 25.4% |
| Population | 50% | 0.3% | 4,800 | 0.5% / 1.4% | 3.3% |
Read the test honestly: it covers one stretch of history, 2019 to 2024, which included the fastest rent and price inflation in forty years, so the method, tempering included, under-projected in that window and the ranges are wide for that reason. ACS 5-year medians also lag the market, because each release averages five years of responses. Derived figures on place pages, such as 2030 rent to income, divide one projection by another.
The Citytistics Index (CTX)
One number for how a place does on the things most people weigh when choosing where to live. It runs on a 0 to 100 scale and reads as a share: a CTX of 74 means the place scores better than 74% of the 6,612 places we rate, so the top 1% all read 99. Those places show their national rank instead (#1 is best), so the leaders are told apart. It is built in three steps, the same way for every place, from published figures only.
- Category ratings. Each input becomes a percentile among every place that passes the data test, turned so that 100 is best. A category averages its inputs. Missing inputs are never filled in.
- Weighted average. The category ratings are averaged with the weights below. When a category has no data, its weight is spread over the others in proportion. A place needs Affordability and Economy and at least 5 of the 7 categories, or it gets no CTX.
- Percentile. The weighted average is ranked against every rated place, so the scale stays even from top to bottom and one extreme place cannot squash the rest.
| Category | Weight | Inputs | Why it counts |
|---|---|---|---|
| Affordability | 20% | Rent to income and home price to income (lower is better). | Whether a typical household can afford to live here at all. |
| Economy | 20% | Median household income (higher) and unemployment (lower). | Pay and the chance of finding work. |
| Safety | 20% | Reported violent crime per 100k (lower), averaged with property crime per 100k (lower) where the city reports it. Agency coverage shown separately. | The first question most people ask about a place. |
| Amenities | 15% | Average percentile of parks, grocery stores, restaurants and cafes, and libraries per 10k residents, trail length per 10k residents, and transit stops per square mile, all mapped in OpenStreetMap inside the city limits. Needs at least four. | Parks, groceries, places to eat, libraries, trails and transit within the city limits. |
| Education | 10% | Share of adults 25+ with a bachelor's degree or more. | A proxy for the local job mix and schools; we hold no test scores yet. |
| Commute | 10% | Mean one-way commute time (lower). | Time lost every working day. |
| Growth | 5% | Average percentile of population growth since 2020 and yearly income growth (Citytistics measure). | Whether people and pay are moving in; a tiebreaker, not a driver. |
The bands on each page: 90 and up, Exceptional; 75 to 89, Strong; 55 to 74, Above average; 45 to 54, Average; 25 to 44, Below average; under 25, Weak.
What it leaves out. School test scores, air quality, weather and natural hazard risk are not in the CTX yet, because we do not hold them for every place or they are county figures shared by every place in the county. Crime counts only offenses reported by the city police department, and about a third of places have no full-year city figure; those are rated without Safety, and their page says so. Map-based counts depend on how thoroughly volunteers have mapped a place. The CTX is a summary, not a verdict: the category ratings and every figure behind them sit on the same page.
Figures refresh as publishers release new data, and the CTX is recomputed with them. Weights change only with a dated note on this page.
What stands out
Each place page lists the facts that make it unlike places of a similar size. For about 20 figures we work out where the place sits among places in its size band (under 10,000; 10,000 to 25,000; 25,000 to 50,000; 50,000 to 100,000; 100,000 to 250,000; 250,000 or more) and, when it is in the top or bottom three, in its state. We keep the most extreme fact per topic and show up to five. The sentences are filled from the numbers; no text is written by a person or a language model. County and metro figures never count, because they are shared by every place in the county and cannot tell one page from another.
Some readings only appear when two figures sit side by side, such as high incomes alongside many cost-burdened renters, or cheap rent in a county with high natural hazard risk. Those are rule-based too, with the thresholds printed in our source code.
Each factor is scored 0 to 100 as a percentile: the share of indexed places it does better than (100 is best), so one extreme place cannot squash everyone else. A factor built from several metrics averages them. Missing inputs are never filled in: the place is ranked on the factors it has.
- Affordability: Rent to income and home price to income (lower is better).
- Economy: Median household income (higher) and unemployment (lower).
- Education: Share of adults 25+ with a bachelor's degree or more.
- Safety: Reported violent crime per 100k (lower), averaged with property crime per 100k (lower) where the city reports it. Agency coverage shown separately.
- Commute: Mean one-way commute time (lower).
- Income growth: Change in median household income over the last 5 ACS years.
- Amenities: Average percentile of parks, grocery stores, restaurants and cafes, and libraries per 10k residents, trail length per 10k residents, and transit stops per square mile, all mapped in OpenStreetMap inside the city limits. Needs at least four.
- Paycheck stretch: Dollars the median household has left after the median rent, with other prices set to the US average (Citytistics measure).
- Pay vs rent: Yearly income growth minus yearly rent growth across the last five ACS releases (Citytistics measure).
- Family fit: Average percentile of rent to income, price to income, students per teacher, homeownership, household size and poverty (Citytistics measure).
- Growth: Average percentile of population growth since 2020 and yearly income growth (Citytistics measure).
- 2030 affordability: Projected 2030 rent to income and home value to income, from each place's own trend blended with the typical trend (lower is better; Citytistics projection).
- Home value outlook: Projected yearly growth in the median home value to 2030 (Citytistics projection).
Rankings are weighted averages of these factors. Each ranking page prints its weights, and you can change them.
How we find what stands out
Every week, after the data refresh, the City Insight Engine works through every place that passes the data test. It runs once, ahead of time; nothing is calculated or written while you wait, and no language model writes any sentence. Each sentence is a template filled with a computed figure, and it appears only when the figure clears a stated threshold. A place that is ordinary on a topic gets no sentence about it.
- Size-band outliers. A figure is mentioned when it is in the top or bottom quarter of places of similar size (at least 40 places per band), with its rank in the state when it is in the top three.
- Statistical peers. The 25 places closest on size, age, education, homeownership, household size, housing age, share born abroad and remote work, with a penalty for being in another Census division. Cost and outcomes are left out of the match, so we can ask how a place does against places built like it. A gap is mentioned when it is at least one peer standard deviation.
- Affordability trajectory. Rent as a share of income, and home value in years of income, in the 2015 to 2019 survey against the 2020 to 2024 survey. The two do not share any survey years. A place is called more or less affordable only when both measures moved together by at least 1.5 points and 0.4 years of income.
- Rent and home value against fundamentals Experimental. A regression of log rent and log home value on income, education, housing age, homeownership, household size, size, remote work and age, within each metro area (or state outside metros). The gap is how much more or less housing costs than places with the same traits in the same market. Out of sample, it explains 52% of the variation in rent within markets and 77% for home values (5-fold cross-validation). It cannot see crime, school quality or location within the metro, so a low figure can mean good value or a drawback the census does not measure. We mention gaps of 20% or more.
- Growth sustainability Experimental. For places growing faster than three quarters of places their size, four signals: rent outrunning pay, home prices pulling away from incomes, very few empty homes, and a high share of cost-burdened renters. Three or more is called strain; none or one, solid footing. It is a checklist, not a forecast.
- Change detector. The latest year of population estimates against the average of the years before, and home value growth against income growth.
Each finding is scored on size of effect (45%), originality (25%), reliability (15%) and how often the topic matters to people deciding where to live (15%). A page shows the strongest six, at most two per theme, with contradictions removed. The theme of the strongest finding sets the order of the page.
Which pages we publish
Search engines see a place page only when it passes the data test below and earns a Page Value Score of 80 or more out of 100, or has 50,000 or more residents. The score weighs original analysis (35%), whether the core figures and their history are present (15%), statistical reliability (15%), how different the page's findings are from its five nearest neighbours' (15%), usefulness of the tools on it (10%), freshness (5%) and transparency (5%). In the current data 4,248 of 6,645 place pages pass. The rest keep a page for readers, marked noindex.
Before the score existed, the page must say something specific: the place has 50,000 or more residents, or it sits in the top or bottom 10% of places its size on at least 3 topics, or in the top or bottom 20% on at least 5. Places that pass the first test but not the second keep a page for readers, marked noindex. A size band needs at least 40 places before we compare against it. Comparison pages are open to search engines only for pairs people plausibly look up: big-city moves between the main cities of the 30 largest metros (each with its nearest and most alike), and neighbouring places of 50,000+ in the same metro, within 25 miles.
A place page is open to search engines only when the place has at least 5,000 residents, all core metrics (population, income, rent, home value) and at least 15 reliable measures. Smaller or thinner places still have a page for people who search for them, marked "limited data" and kept out of search indexes. Comparison pages are indexed only for pairs people actually look for.
Crime data
Rates use offenses reported to the FBI. Not every agency reports every year, so each figure carries its agency coverage. Below 80% coverage we flag the figure and down-weight it in rankings.
Survey estimates
ACS figures are survey estimates with margins of error, which are larger for small places. We show 5-year estimates for stability, and we hide any figure whose margin of error is too wide to rely on (a coefficient of variation above 40%). Where a place has no reliable figure of its own, its page may show the county or metro figure, labeled as such.
Sources
Sources include US-government public-domain publications as well as third-party datasets. OpenStreetMap contributors provide mapped amenities under the Open Database License (ODbL), and Wikidata is CC0. Attribution and applicable licenses are listed below. Each source is checked for newer releases on its own schedule.
- American Community Survey 5-year estimates · US Census Bureau · 2020 to 2024 estimatesTables B01002, B01003, B05002, B07003, B08013, B08301, B15003, B17001, B19013, B19301, B23025, B25002, B25003, B25010, B25035, B25064, B25070, B25077, B25088, B25103, B28002, S2701
- Regional Price Parities · US Bureau of Economic Analysis · Latest release, retrieved 2026-11-14
- Fair Market Rents · US Department of Housing and Urban Development · Fiscal year 2027
- Income Limits · US Department of Housing and Urban Development · Fiscal year 2026
- House Price Index · Federal Housing Finance Agency · Latest release, retrieved 2026-11-14
- Local Area Unemployment Statistics · US Bureau of Labor Statistics · Not in this build
- Quarterly Census of Employment and Wages · US Bureau of Labor Statistics · 2025 annual averages
- Population Estimates · US Census Bureau · Vintage 2025
- SOI Tax Stats, county-to-county migration · Internal Revenue Service · Tax years 2022 to 2023Counts tax returns and the people on them; flows under 20 returns are suppressed.
- County Business Patterns · US Census Bureau · 2023 data
- OpenStreetMap · © OpenStreetMap contributors, available under the Open Database License (ODbL) · Map data as of 2026-11
- National Park Service sites · Wikidata (CC0), units operated by the National Park Service · Retrieved 2026-11
- National Register of Historic Places · National Park Service · Listings through 2026-05-22
- Wikidata · Wikimedia Foundation (CC0) · Retrieved 2026-11
- Crime Data Explorer · Federal Bureau of Investigation · 2025 data
- Common Core of Data · National Center for Education Statistics · 2024 release
- 1991-2020 climate normals · NOAA National Centers for Environmental Information · Series updated 2026-11-05Averages for 1991 to 2020 computed from NOAA's county climate series (nClimDiv).nClimDiv county series
- National Risk Index · Federal Emergency Management Agency · Latest release, retrieved 2026-11-14NRI data and maps
- Air Quality System, annual AQI by county · US Environmental Protection Agency · 2025 dataAirData download files
Derived by Citytistics from the tables above: rent to income (B25064, B19013); property tax rate (B25103, B25077); home price to income (B25077, B19013).
Trend lines also use the American Community Survey 1-year estimates for areas of 65,000 or more, and per capita personal income from the BEA county and metro accounts.
This product uses the Census Bureau Data API but is not endorsed or certified by the Census Bureau. Data as of 2026-11-14.
Reuse
Quote and cite any figure with a link to its page. Bulk or programmatic use needs an API plan.
Ads and independence
Advertisers never change data, scores or rankings. Ads are labeled and never appear inside an answer. How ads work here.