Indexes
On this page
An Index combines several metrics into one rank, so you can find the places that do well on all of them at once. Birdseye provides 11 ready-made Indexes that answer common questions, such as how affordable, safe or fast-growing a place is, and they are free to view. With Pro you can also build your own, choosing the metrics, their weights and which end of each is better. This page explains how to read an Index, how it is calculated, and what is inside each Birdseye Index.
How to read an Index#
An Index value is a rank from 0 to 100. It shows where a place stands against every other place of the same type in the country: tracts against tracts, ZIPs against ZIPs, counties against counties, metros against metros. A tract at 80 ranks above about 80% of the tracts that have a value for that Index. A place at 50 is in the middle.
Higher is always better on the Index's own terms. The name says what it rewards: Affordability rewards low costs, and Safety rewards low crime.
Read the rank as a signal of where a place stands, not as an exact measurement. Much of the underlying data comes from surveys and estimates, so a few points between two places, such as 64 and 61, should not be read as a real difference. Look at the broad position, then at the metrics behind it.
The rank is relative, not a grade. A high rank means better than most places of the same type, not good by a fixed standard. A place's rank can also move when other places change, even if its own numbers stay the same.
The ranks are national, so places inside one metro can bunch together. In an expensive metro, most tracts may rank low on Affordability. To compare places within one area, read the metrics behind the Index.
The map colors follow the rank, so one color means the same rank at every geography.
Birdseye Indexes#
Each Birdseye Index has a fixed list of metrics with fixed shares, the same for every user.
| Index | What it measures |
|---|---|
| Affordability | How low home prices and ownership costs are, with one measure against local incomes |
| Safety | Reported crime and housing vacancy |
| Schools | How the local public schools perform |
| Economy | Jobs, incomes and work within reach |
| Appreciation | How much home values have risen here over five and ten years |
| Accessibility | How easily you can reach work and get around from here |
| Rental Yield | What a rental here earns relative to what it costs |
| Growth | Whether the people, jobs and incomes here have been increasing over the past five years |
| Buyer's Leverage | How much room a buyer has here right now, against this market's own normal |
| Stability | How steady this market and its job base have been |
| Lending Health | How readily, and how cheaply, mortgage credit is reaching buyers here |
Index details lists the metrics in each one.
Custom Indexes#
You build a custom Index from your own choice of metrics. It has no geography of its own: it colors the geography the map shows and follows the map when you switch. You can use any metric at any of the four geographies. Where a metric is not measured at that geography, a value from a related geography stands in, and a label under the metric says which. See Geographies.
Weights. Each metric has a weight from 0 to 100. Only the ratios count: 50 and 50 count the same as 100 and 100. The panel shows each metric's share of the total. A metric at 0 is left out.
Direction. Each metric starts on its usual reading, higher or lower is better, and you can flip it. A renter and a landlord want opposite ends of the same rent figure.
Calculation#
An Index is built in five steps:
- One scale. Each metric is compared with the same metric for every place of the same type in the country. That puts rents, crime rates and commute times on one scale, so they can be combined. A few extreme places cannot stretch the scale for everyone else.
- Direction. Each metric is turned so that its better end counts as higher.
- Weights. Each metric counts in proportion to its share.
- Blend. The weighted metrics are combined into one number.
- Rank. That number is ranked against every place of the same type.
The blend in step 4 differs between Birdseye and custom Indexes:
- Birdseye Indexes use a plain weighted average, so a strong metric can make up for a weak one.
- Custom Indexes use a balanced blend that sits between a plain average and a geometric mean. A place that is good on every metric beats a place that is excellent on one and poor on another.
An example. Two metrics carry equal weight. Place A is at the 90th percentile on one and the 30th on the other. Place B is at the 60th on both. A plain average puts both at 60. The balanced blend puts Place A at 55 and Place B at 60, so in a custom Index Place B ranks higher.
Geographies#
Every Index exists at four geographies: census tract, ZIP code, county and metro. Metro covers both metropolitan and micropolitan areas. An Index is calculated separately at each geography, from that geography's own values. A county at 70 and a tract at 70 each rank above about 70% of their own type; they are not compared with each other. The same area can rank differently at two geographies.
Where a metric is not measured at a geography, a value from a related geography stands in. A label shows which:
| Label | What it means |
|---|---|
| Native | Measured at this geography. |
| City or county data | The tract takes the figure of the city or county government that covers it. For crime, the label is "City or sheriff data". |
| School district data | The figure comes from the school districts that serve the place, weighted by children. |
| County value on every tract | Every tract in a county gets the county's figure. For ZIPs, the label is "County value on every ZIP". |
| Metro value on every county | Every county in a metro gets the metro's figure. |
| Blend of ZIPs, by housing units | A tract's figure blends the ZIPs it overlaps, weighted by housing units. For ZIPs, the label is "Blend of tracts, by housing units". |
| From member tracts | A county's or metro's figure is built from the tracts inside it. |
| From member counties | A metro's figure is built from its counties. |
| Metro where published, else member counties | The metro's own figure where the source publishes one; otherwise it is built from its counties. |
| No data at this geography | The metric has no value here. |
A value from a larger area is the same for every place inside it, so it cannot tell those places apart. Conditions can differ a lot inside one county; crime is the clearest case.
Missing data#
Not every metric has a value for every place. A source may not publish for an area, may suppress a figure for privacy, or may not cover the area at all.
Places with fewer than 100 households are left out and show no value. Most are industrial land, parks, prisons or dormitories, where a few residents would give an unreliable figure.
A Birdseye Index needs at least 60% of its metrics, rounded up: 2 of 2, 2 of 3, 3 of 4 or 3 of 5. With fewer, the place shows no value and is grey on the map. With enough, the Index is built from the metrics the place has, and their shares are scaled up to fill the gap. A place built from 3 of 5 metrics is ranked alongside places that have all 5.
A custom Index has no minimum. Each place is scored on the metrics it has.
Index details#
Each table lists the Index's metrics, their shares, which end each rewards, where the number comes from and its period. Shares add to 100%, give or take rounding. Coverage counts the places with a value, before the 100-household rule. All values were last rebuilt November 10, 2026.
Affordability
How low home prices and ownership costs are, with one measure against local incomes.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Price-to-income ratio | 25% | Lower | Measured at every geography | 2020–2024 (ACS 5-year) |
| Zillow home value | 25% | Lower | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | August 2026 |
| Median asking price per square foot | 25% | Lower | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | September 2026 |
| Median property tax | 12.5% | Lower | Measured at every geography | 2020–2024 (ACS 5-year) |
| Median monthly owner costs | 12.5% | Lower | Measured at every geography | 2020–2024 (ACS 5-year) |
Coverage: 82,486 of 85,184 tracts, 29,002 of 33,627 ZIPs, 3,137 of 3,144 counties and all 925 metros.
Read with care: Rent is not in this Index. Price level dominates: three metrics are prices, and property tax and owner costs rise with price, so the Index favors places where homes are cheap. Some places are cheap because jobs or people are leaving, so read it next to Economy and Growth. At tract level, half the weight comes from ZIP figures, so neighboring tracts move together.
Safety
Reported crime and housing vacancy.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Crime index | 50% | Lower | Tracts: city or sheriff data. ZIPs: blend of tracts. Counties and metros: from member tracts | 2024 |
| Vacancy rate (Census) | 50% | Lower | Measured at every geography | 2020–2024 (ACS 5-year) |
Coverage: 68,416 of 85,184 tracts, 26,727 of 33,627 ZIPs, 2,717 of 3,144 counties and 824 of 925 metros.
Read with care: A tract takes the crime rate of the police agency that covers it, so every tract in one city shares a rate, and conditions differ a lot inside a city. Places with no reporting agency have no crime figure, which is why coverage is lower. Vacancy counts empty homes of every kind, including seasonal homes, so resort areas can rank low without a crime problem.
Schools
How the local public schools perform.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Math proficiency | 25% | Higher | School district data at every geography | 2020–21 school year |
| ELA proficiency | 25% | Higher | School district data at every geography | 2021–22 school year |
| Graduation rate | 25% | Higher | School district data at every geography | 2020–21 school year |
| Student-teacher ratio | 25% | Lower | School district data at every geography | 2024–25 school year |
Coverage: 79,414 of 85,184 tracts, 32,332 of 33,627 ZIPs, 3,084 of 3,144 counties and 903 of 925 metros.
Read with care: Figures describe the school districts that serve a place, weighted by children, not the nearest school. Proficiency is measured against each state's own standard, so it does not compare cleanly across states. The test and graduation figures are several years older than the rest of the data.
Economy
Jobs, incomes and work within reach.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Median household income (Census) | 25% | Higher | Measured at every geography | 2020–2024 (ACS 5-year) |
| Accessible jobs | 25% | Higher | Measured at tracts. ZIPs: blend of tracts. Counties and metros: from member tracts | 2023 |
| High-wage job share | 25% | Higher | Measured at tracts. ZIPs: blend of tracts. Counties and metros: from member tracts | 2023 |
| Unemployment rate | 25% | Lower | Tracts and ZIPs: county value. Measured at counties. Metros: metro where published, else member counties | August 2026 |
Coverage: 80,807 of 85,184 tracts, 31,964 of 33,627 ZIPs, 3,031 of 3,144 counties and 888 of 925 metros.
Read with care: Income describes the people who live in a place. The two job metrics describe jobs located in or within reach of it, whoever holds them. Unemployment is a county figure at tract and ZIP level, so it moves every tract in a county together.
Appreciation
How much home values have risen here over five and ten years.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Zillow home value, 5-yr change | 50% | Higher | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | August 2021–August 2026 |
| Zillow home value, 10-yr change | 50% | Higher | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | August 2016–August 2026 |
Coverage: 81,563 of 85,184 tracts, 23,325 of 33,627 ZIPs, 2,970 of 3,144 counties and 858 of 925 metros.
Read with care: This is past price growth, not a forecast. Strong past growth can mean a market has already priced in its gains, so read it next to Affordability. Both metrics come from one source, so where it has no figure the place has no value. At tract level, all of it comes from ZIP figures.
Accessibility
How easily you can reach work and get around from here.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Mean commute minutes | 33.3% | Lower | Measured at every geography | 2020–2024 (ACS 5-year) |
| Accessible jobs | 33.3% | Higher | Measured at tracts. ZIPs: blend of tracts. Counties and metros: from member tracts | 2023 |
| Walk commute rate | 13.3% | Higher | Measured at every geography | 2020–2024 (ACS 5-year) |
| Transit commute rate | 13.3% | Higher | Measured at every geography | 2020–2024 (ACS 5-year) |
| Bike commute rate | 6.7% | Higher | Measured at every geography | 2020–2024 (ACS 5-year) |
Coverage: 82,996 of 85,184 tracts, 32,667 of 33,627 ZIPs, all 3,144 counties and all 925 metros.
Read with care: Commute time and mode describe how residents actually travel, not what service exists. A place with good transit that few people use ranks low on transit. Accessible jobs favors dense metro cores, so rural places rank low almost everywhere.
Rental Yield
What a rental here earns relative to what it costs.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Gross rental yield | 66.7% | Higher | Measured at every geography | 2020–2024 (ACS 5-year) |
| Renter share | 16.7% | Higher | Measured at every geography | 2020–2024 (ACS 5-year) |
| Vacancy rate (Census) | 16.7% | Lower | Measured at every geography | 2020–2024 (ACS 5-year) |
Coverage: 82,926 of 85,184 tracts, 32,689 of 33,627 ZIPs, all 3,144 counties and all 925 metros.
Read with care: Gross yield is a year's rent against home value, before taxes, insurance, upkeep and vacancy, so it is not a return. Yield is highest where prices are low, which can mean weak demand, so read it next to Growth and Economy. All three metrics are survey averages over 2020–2024, so they lag today's rents and prices.
Growth
Whether the people, jobs and incomes here have been increasing over the past five years.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Population, 5-yr change | 25% | Higher | Measured at every geography | 2015–19 to 2020–24 (ACS 5-year) |
| Accessible jobs, 5-yr change | 25% | Higher | Measured at tracts. ZIPs: blend of tracts. Counties and metros: from member tracts | 2018–2023 |
| Median household income, 5-yr change | 25% | Higher | Measured at every geography | 2015–19 to 2020–24 (ACS 5-year) |
| Permitted units per 1k population, 3-year pool | 12.5% | Higher | Tracts: city or county data. ZIPs: county value. Measured at counties and metros | 2023–2025 |
| Employment, 5-yr change | 12.5% | Higher | Tracts and ZIPs: county value. Measured at counties and metros | Q3 2020–Q3 2025 |
Coverage: 83,845 of 85,184 tracts, 33,267 of 33,627 ZIPs, 3,137 of 3,144 counties and 923 of 925 metros.
Read with care: Growth looks back; it does not forecast. Permits and employment are city or county figures below the county level, so they cannot separate one neighborhood from another. Employment change starts in late 2020, when jobs were still recovering from the pandemic, so it can overstate growth.
Buyer's Leverage
How much room a buyer has here right now, against this market's own normal.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Listings for sale vs seasonal norm | 28.6% | Higher | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | Jul–Sep 2026 vs. earlier years |
| Days on market vs seasonal norm | 28.6% | Higher | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | Jul–Sep 2026 vs. earlier years |
| Price-cut share vs seasonal norm | 28.6% | Higher | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | Jul–Sep 2026 vs. earlier years |
| Median asking price vs seasonal norm | 14.3% | Lower | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | Jul–Sep 2026 vs. earlier years |
Coverage: 74,411 of 85,184 tracts, 14,818 of 33,627 ZIPs, 2,837 of 3,144 counties and all 925 metros.
Read with care: Each metric compares a market with its own normal for the season, not with other places. A high rank means more room than usual here, not a cheap market. It moves month to month by design. Many ZIPs have only a few listings, so a few homes can swing them, which is also why ZIP coverage is low. At tract level, all of it comes from ZIP figures.
Stability
How steady this market and its job base have been.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Population volatility 5-year | 20% | Lower | Measured at tracts. ZIPs: blend of tracts. Counties and metros: from member tracts | 2020–2024 (ACS 5-year) |
| Job sector diversity | 20% | Higher | Measured at tracts. ZIPs: blend of tracts. Counties and metros: from member tracts | 2023 |
| Home value drawdown from 5-year peak | 20% | Smaller drop | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | August 2021–August 2026 |
| Unemployment max drawdown | 20% | Smaller rise | Tracts and ZIPs: county value. Measured at counties. Metros: metro where published, else member counties | 2010–2026 |
| Home value volatility, 5-year | 20% | Lower | Tracts: blend of ZIPs, by housing units. Measured at ZIPs, counties and metros | August 2021–August 2026 |
Coverage: 83,920 of 85,184 tracts, 33,143 of 33,627 ZIPs, 3,126 of 3,144 counties and 923 of 925 metros.
Read with care: Stability rewards steady, settled markets. It says nothing about whether prices are high or rising. Population volatility in small tracts partly reflects survey noise. Unemployment is a county figure at tract and ZIP level.
Lending Health
How readily, and how cheaply, mortgage credit is reaching buyers here.
| Metric | Share | Better when | Where the number comes from | Period |
|---|---|---|---|---|
| Purchase denial rate | 50% | Lower | Measured at tracts, counties and metros. ZIPs: blend of tracts | 2023–2025 (pooled) |
| Average purchase interest rate | 50% | Lower | Measured at tracts, counties and metros. ZIPs: blend of tracts | 2023–2025 (pooled) |
Coverage: 80,829 of 85,184 tracts, 32,551 of 33,627 ZIPs, 3,107 of 3,144 counties and all 925 metros.
Read with care: The figures come from home-purchase mortgages that lenders report, pooled over up to three years, so places with few sales rest on few loans. Interest rates mostly follow when loans were made, so differences between places are small.
Limitations#
- Made by Birdseye. We made these Indexes to make the data easier to read. No outside body has reviewed or tested them. The metrics and their shares are our judgment. Use an Index to narrow a search, then read the metrics underneath it.
- Data, not judgments about people. Indexes describe published data. They do not judge the desirability of a neighborhood or the people who live there. Do not use them to steer clients or as an appraisal adjustment.
- Mixed dates. An Index can combine a monthly figure with a five-year survey average. Each table above gives each metric's period.
- Past, not future. Trend metrics describe what has happened. They do not forecast.
Birdseye summarizes data from public and commercial sources. Each source’s publisher is the authority on its data and methods. See our Terms of Service.