For journalists and researchers

Neighborhood data you can cite

Saktoo measures every part of Los Angeles and Orange County on rent, pollution exposure, tree canopy, rent burden, commute time, walkability, reported incidents and economic mobility. All of it is free to use, including commercially, as long as you credit us.

The downloads below are aggregated to named neighborhoods and to cities, because that is what stories are written about. Every column names its source, and the limitations are listed as plainly as the figures.

Download

Data as of 2026-08-20 · methodology v1 · 7,644 areas measured

Checking one or two specific places? The comparison tool puts up to four side by side against the whole region and exports just those rows. For the raw hexagon grid or a GeoJSON, the Index Explorer exports both formats for any view, or email us and we will run the cut for you.

How to cite it

Saktoo Neighborhood Index, data as of 2026-08-20, methodology v1. Retrieved from https://saktoo.com/data. Licensed CC BY 4.0; place counts © OpenStreetMap contributors, ODbL.

For a figure taken off the map, the Explorer generates a citation for the exact view you were looking at, including the weighting and any filters, so an editor can reopen it. Every published snapshot stays online at its own address, so a number cited today still resolves in a year.

Licence

Free to reuse, including commercially

This is a collection of separately licensed data rather than one blanket-licensed file, because we are not in a position to offer one.

CC BY 4.0 for everything except the columns noted below. Reuse, republish, remix and build on these figures, including commercially, as long as you credit Saktoo and link back. Full terms.

ODbL 1.0 for the columns counting nearby places. Columns counting nearby places come from OpenStreetMap. Reuse them freely, credit OpenStreetMap contributors, and keep any database you build from them under the same licence. Full terms.

Not included: HOLC redlining grades are published by the Digital Scholarship Lab at the University of Richmond under a non-commercial licence. Saktoo does not redistribute them, and the redlining comparison is not offered under the terms above.

Saktoo Neighborhood Index (saktoo.com/data). Place counts © OpenStreetMap contributors, ODbL.

Data dictionary

What every column means

Identity and coverage

Who the row is about, and how much data stands behind it. Check these before quoting anything else in the row.

neighborhood / city
The place the row describes.
county
LA or OC. Places straddling the line take the county holding most of their area.
hexes
How many hexagons (about 0.46 km² each) were averaged to make the row. More hexes means a steadier figure.
thin_sample
yes when the row rests on fewer than three hexagons.Do not rank on a thin row. It is published so small real places are not missing, not so they can be compared.
name_is_invented
yes when the neighborhood name was generated by a clustering pass rather than taken from a documented source.An invented name is a label for an area, not a name residents use. Do not print it as though it were one.
name_source / boundary_confidence
Where the name and the boundary came from: official, documented, la_times, cluster, or a Census city floor.

Absolute figures

Real-world units, straight from the sources named above. These are the columns to quote: they need no explanation and carry their own provenance.

median_rent_usd
Median gross rent per month across the area's hexagons.
median_home_value_usd
Median owner-reported home value.This is a Census self-reported median, not a valuation or a sale price, and it is top-coded at $2,000,001 so the most expensive areas tie at the ceiling.
tree_canopy_pct
Share of land under tree canopy.
avg_commute_minutes
Mean one-way commute for workers living here.
walkability_0_100 / transit_access_0_100
EPA Smart Location indices, rescaled 0 to 100.
pct_same_home_10_years
Share of households in the same home for a decade or more.
pct_overcrowded
Share of homes above one occupant per room.
pct_vacant / pct_broadband
Share of units vacant; share of households with a broadband subscription.
pct_single_family / pct_large_multifamily
What the housing stock is built of.
median_bedrooms / median_year_built
Typical unit size and typical construction era.

Nearby places

Counts from OpenStreetMap. These are the share-alike columns, and they are averages rather than totals.

restaurants_nearby_avg and the rest
Average number of that kind of place within a hexagon and its six neighbours, about 1.2 km across.Neighbouring hexagons overlap, so these are NOT neighborhood totals and must not be summed or reported as a count of businesses.

Saktoo percentile ranks

Where an area sits relative to every other area in LA and Orange County, 0 to 100. Useful for comparison, weaker for quoting: they are our construction, not a measurement, so they need a sentence of explanation in copy.

affordability_pctl, environment_pctl, location_pctl, safety_pctl, opportunity_pctl
The average rank of the hexagons making up the area. Higher is cheaper, cleaner, better connected, fewer reported incidents, higher past mobility.This is a mean of hexagon ranks, not a rank recomputed at neighborhood level. The two differ, and the difference matters if you build a ranking.
pollution_exposure_pctl / rent_burden_pctl / combined_burden_pctl
Pollution exposure, rent-to-income, and the combined index. Higher means worse on all three, the opposite direction to the dimensions above.
safety_is_citywide_figure
yes when at least one contributing hexagon carries a uniform citywide incident rate rather than street-level records.Where this is yes, the safety figure describes a whole jurisdiction. Presenting it as neighborhood-level detail would be wrong.
opportunity_pctl
Opportunity Atlas outcomes.Describes adults born 1978 to 1983 measured around 2015, so it is about where past residents ended up, not how children do in that neighborhood today.

What this data cannot tell you

Limitations, stated up front

Several of these are deliberate refusals rather than missing features. Where a number would encode something we think an index has no business encoding, we publish the underlying figure and decline to score it.

No neighborhood scores for schools

School outcomes track race and income closely, so rating a neighborhood on them would encode demographics into a ranking. Per-school state figures are published with their source and year on neighborhood pages; no school figure is read by any scoring code.

Overcrowding is published but never scored

It correlates 0.68 with Hispanic population share in this region, so weighting it into any index would systematically penalize immigrant neighborhoods.

Reported incidents are reports, not crime

They measure what was reported to and recorded by police, which varies with trust and staffing. Blank never means zero incidents, only no coverage. Where safety_is_citywide_figure is yes, the number is a whole jurisdiction's rate.

311 disorder measures reporting as much as conditions

It covers the city of Los Angeles and nowhere else. Absent values mean no 311 feed exists, not that an area is clean. Context only, never scored.

Housing unit counts are not published

Per hexagon they are interpolated estimates of a block-group total, so summing them double-counts and averaging them means nothing recognizable. Omitted rather than published misleadingly.

Park distance is to a point, not a boundary

OpenStreetMap gives park positions rather than outlines, so a large park reads as farther away than its edge actually is.

This is a cross-section, not a time series

Every figure describes the present. The data can characterize pressure — rent against income, tenure, overcrowding — but it cannot show change over time, so it cannot on its own demonstrate displacement or gentrification.

Sources

Where every figure comes from

American Community Survey (5-year)Public domain (US Census Bureau)
Median gross rent B25064, median home value B25077, travel time B08303, tenure B25038, occupants per room B25014, units in structure B25024, bedrooms B25041, year built B25034/B25035, vacancy and broadband. Block-group level.
CalEnviroScreen 4.0Open, California OEHHA
Diesel particulate matter, traffic proximity, PM2.5, ozone, cleanup sites and hazardous waste percentiles, from California OEHHA.
EPA Smart Location DatabasePublic domain (US EPA)
Walkability, transit access and density indices at block-group level.
NLCD tree canopyPublic domain (USGS)
Percent tree canopy cover, USGS Multi-Resolution Land Characteristics Consortium.
Opportunity AtlasOpen, Opportunity Insights
Adult income rank at about age 35 for children raised in the tract, cohort born 1978 to 1983 (Chetty et al., 2018). Tract level, and a description of past residents rather than of the neighborhood today.
Police-reported incidentsOpen government data
LAPD and LA County Sheriff street-level records, CA DOJ OpenJustice agency counts elsewhere, over a three-year window.
LA 311 service requestsCC0
Graffiti and illegal dumping requests per resident, city of Los Angeles only. Context, never scored.
OpenStreetMapODbL 1.0
Counts of restaurants, cafes, groceries, parks and gyms. These columns carry share-alike.

Earlier snapshots

Nothing is ever overwritten, so a figure cited from an older release still resolves.

Ask us for something

We will run the cut for you

If you need a slice that is not here, want a figure checked before you file, or want a chart or map for your story, email support@saktoo.com. We aim to reply within one working day, and faster if you are on deadline: say so in the subject line.

Any Explorer view can be embedded in an article as a live interactive map, free and without a login. Open a view, then use “Copy embed code”.