Nobody Publishes What It Costs to Live in Hancock Park. So It Isn't on This List.

We scored eighteen places out of ten on rent, reported crime and whether you can live here without a car, using only public-domain federal and state data. Koreatown finished first, Culver City finished last, and two well-known names could not be scored at all.

Start with the place we could not rank. Hancock Park is one of the wealthiest districts in Los Angeles, and no publisher in America reports what it costs to live there. Zillow keeps 97 neighborhood series for the City of Los Angeles and Hancock Park is not one of them. The Census Bureau has no place and no ZIP code tabulation area that matches it. HUD's finest geography is coarser than the district. The only housing figure covering its ground belongs to ZIP 90004 — which also contains a dense multifamily belt running east toward Virgil Village and East Hollywood, where nearly all the rental listings are. Print that number under a Hancock Park headline and one of the most expensive districts in the city reads as mid-affordable. We would rather leave a gap than fill it with a number from the wrong side of a ZIP code, so Hancock Park is not ranked, and neither is Larchmont Village, for the same reason. That refusal is the most important thing on this page, which is why it is at the top of it and not in a footnote.

What follows is eighteen places scored out of ten on three things we could measure with public data: what households pay in rent, how much crime was reported in 2025, and how easily you can live here without a car. The recipe is printed below and every component is printed in every entry, so you can disagree with our weighting rather than with our taste. Koreatown finished first. Culver City finished last. Six of the eighteen are whole incorporated cities and twelve are Los Angeles neighborhoods, which is not a fair comparison and we could not make it one.

The one number that isn't here, and what its absence did to the ranking

This list is built entirely on public-domain federal and state data. There is no Zillow figure anywhere in it, and that is a deliberate hole with a consequence worth stating before you read a single rank.

The obvious way to score housing cost in Los Angeles is asking rent — what a new listing wants — and the standard source for it is Zillow's Observed Rent Index. We pulled it. We did not publish it. Zillow's terms-of-use page returned an HTTP 403 to every automated read on the day this was reported, so nobody at this paper has established that we may republish its figures. Being unable to read a licence is not permission to ignore one. Every Zillow number is out of this piece until a human opens that page and reads it.

So housing cost here is the Census Bureau's median gross rent from the American Community Survey 2020–2024 five-year estimates, which is a work of the United States Government and free of copyright under 17 U.S.C. §105. That is a different measurement, not a substitute one. Asking rent is what a newcomer will face. Median gross rent is what the households already there pay, including their utilities, averaged over five years and expressed in 2024 dollars. In a city where the Rent Stabilization Ordinance covers most multifamily built before October 1978, those two numbers describe two different populations, and the gap between them is the single most important fact about the Los Angeles housing market.

Dropping Zillow moved the ranking, and it moved it most where the gap is widest. On asking rent, West Adams looked expensive — new and renovated stock listing at a premium into a historically low-income Black and Latino neighborhood — and it placed seventh. On what its households actually pay, ZIP 90016's median gross rent is $1,815, and West Adams rises to joint second. Mar Vista rises from eleventh to seventh. In the other direction, South Pasadena falls from joint eighth to joint twelfth and Pasadena from joint fifth to joint eighth, because both are places where sitting tenants pay close to what new tenants are asked. The top and the bottom barely moved: Koreatown was first on both measures, and Culver City, Manhattan Beach and Santa Monica finished eighteenth, seventeenth and sixteenth either way.

Read that as the honest summary of the switch: a ranking built on asking rent is a ranking for people about to move, and a ranking built on median gross rent is a ranking of what a place currently costs the people in it. West Adams is the row where the difference is not academic. A 78 percent gap between what sitting tenants pay and what new units ask has a name, and the name is displacement, not "West Adams is expensive."

How we built this, and what it can't tell you

We scored eighteen places out of ten on three things we could measure with public data: what it costs to rent, how much crime was reported in 2025, and how easily you can live here without a car. The recipe is printed below and the components are printed in every entry, so you can disagree with our weighting rather than our taste.

We did not score restaurants, bars or things to do. No free, licensable index of either exists at this geography, and a count of Michelin stars would tell you where L.A.'s tasting menus are, not whether you can eat well near your door. Those judgments are labelled as judgments, and no restaurant and no bar is named anywhere on this page.

Three things you should hold against us. Six of these places are whole cities and twelve are neighborhoods, which is not a fair comparison and we could not make it one. Every neighborhood rent figure is a ZIP-code figure wearing a neighborhood's name, because nobody publishes rent by neighborhood. And crime counted per resident overstates risk badly wherever a shopping district sits inside the line.

Two places are missing on purpose. Nobody publishes what it costs to live in Hancock Park or Larchmont Village. We would rather leave a gap than fill it with a number from the wrong side of the ZIP.

The recipe, and the bands, so you can re-run us

Housing is worth 4.0 of the 10 points because it is the axis with the real spread — median gross rent across this set runs from $1,612 to $3,449 — and because it is measured the same way in every row. Violent crime is worth 2.0 and property crime is capped at 1.0 on purpose: the property numbers are the ones ruined by daytime population, and we could not correct for that because no daytime-population data exists on these boundaries. The remaining 3.0 splits three ways across car-free living, because the walkability index alone cannot carry a ranking — all eighteen places fall between 13.41 and 16.51 on a 20-point scale, and ranking on a column that tight would manufacture precision that is not there.

The bands, in full:

Median gross rent: under $1,800 → 4.0 · $1,800–2,099 → 3.2 · $2,100–2,399 → 2.4 · $2,400–2,699 → 1.6 · $2,700–2,999 → 0.8 · $3,000 and over → 0.0. Every band is $300 wide, and the reason is the Census margin of error: it is under $90 on fourteen of these eighteen figures, so most rows sit well inside the band they landed in. Two do not, and we name them below — Santa Monica, whose figure is two dollars from a boundary, and Studio City, whose ±$163 is wide enough to reach either one.
Violent offenses per 1,000 residents: under 2.0 → 2.0 · 2.0–2.9 → 1.6 · 3.0–3.9 → 1.2 · 4.0–5.4 → 0.8 · 5.5–7.9 → 0.4 · 8.0 and over → 0.0.
Property offenses per 1,000: under 15 → 1.0 · 15–24.9 → 0.7 · 25–34.9 → 0.4 · 35 and over → 0.0.
EPA National Walkability Index: 15.26 and over → 1.0 · 14.50–15.25 → 0.7 · under 14.50 → 0.4.
Residents within half a mile of a Metro rail station: 50% and over → 1.0 · 25–49.9% → 0.75 · 10–24.9% → 0.5 · above zero → 0.25 · none → 0.0.
Households owning no car: 12% and over → 1.0 · 8–11.9% → 0.75 · 5.5–7.9% → 0.5 · 3.5–5.4% → 0.25 · under 3.5% → 0.0.

Every crime band is at least 1.0 per 1,000 wide, deliberately. Our neighborhood method runs about 7 percent low on violent crime against California's summary definition, and a band that wide absorbs a 7 percent instrument bias. A 1-to-18 rank does not.

Where a neighborhood spans several ZIP codes we scored the one matching the residential core, named it in the list, and published below what the other ZIP would have done to the rank.

The ranking

Every figure below is printed with the component it earned, in the order the recipe states them. The chart at the top of this page is the same eighteen rows drawn to scale.

1. 1. Koreatown — 7.80. Median gross rent $1,715 (ZIP 90005) · violent 7.04 and property 25.02 per 1,000 · walkability 15.45 · 80.2% within half a mile of rail · 24.4% of households car-free.
2. 2=. Los Feliz — 7.70. Median gross rent $1,940 (ZIP 90027) · violent 2.39 and property 18.75 per 1,000 · walkability 14.57 · 28.4% within half a mile of rail · 10.8% of households car-free.
3. 2=. West Adams — 7.70. Median gross rent $1,815 (ZIP 90016) · violent 5.32 and property 22.31 per 1,000 · walkability 16.51 · 67.4% within half a mile of rail · 14% of households car-free.
4. 4. Highland Park — 7.60. Median gross rent $1,913 (ZIP 90042) · violent 3.34 and property 11.85 per 1,000 · walkability 14.85 · 36.3% within half a mile of rail · 9.1% of households car-free.
5. 5. Eagle Rock — 7.00. Median gross rent $1,808 (ZIP 90041) · violent 2.30 and property 14.59 per 1,000 · walkability 14.69 · no resident within half a mile of rail · 7.7% of households car-free.
6. 6. Leimert Park — 6.45. Median gross rent $1,612 (ZIP 90008) · violent 12.06 and property 42.88 per 1,000 · walkability 14.78 · 44% within half a mile of rail · 14.5% of households car-free.
7. 7. Mar Vista — 6.30. Median gross rent $2,279 (ZIP 90066) · violent 1.85 and property 14.42 per 1,000 · walkability 14.40 · no resident within half a mile of rail · 5.9% of households car-free.
8. 8=. Glendale — 6.10. Median gross rent $2,182 (city) · violent 2.24 and property 17.77 per 1,000 · walkability 13.41 · no resident within half a mile of rail · 12% of households car-free.
9. 8=. Pasadena — 6.10. Median gross rent $2,265 (city) · violent 4.06 and property 22.32 per 1,000 · walkability 14.62 · 27.2% within half a mile of rail · 10.6% of households car-free.
10. 10. Silver Lake — 6.00. Median gross rent $1,929 (ZIP 90026) · violent 3.15 and property 32.87 per 1,000 · walkability 14.90 · no resident within half a mile of rail · 7.1% of households car-free.

The cut line. Everything below is reported, not recommended — the same recipe, lower scores.

1. 11. Atwater Village — 5.90. Median gross rent $2,250 (ZIP 90039) · violent 2.28 and property 17.07 per 1,000 · walkability 14.80 · no resident within half a mile of rail · 6.5% of households car-free.
2. 12=. Echo Park — 5.85. Median gross rent $1,929 (ZIP 90026) · violent 5.44 and property 27.73 per 1,000 · walkability 15.17 · no resident within half a mile of rail · 11.3% of households car-free.
3. 12=. South Pasadena — 5.85. Median gross rent $2,284 (city) · violent 2.76 and property 18.35 per 1,000 · walkability 14.14 · 25.7% within half a mile of rail · 3.3% of households car-free.
4. 14. Sherman Oaks — 5.35. Median gross rent $2,277 (ZIP 91423) · violent 2.85 and property 27.32 per 1,000 · walkability 14.75 · no resident within half a mile of rail · 4.5% of households car-free.
5. 15. Studio City — 4.90. Median gross rent $2,535 (ZIP 91604) · violent 3.20 and property 29.94 per 1,000 · walkability 14.67 · 10% within half a mile of rail · 5.8% of households car-free.
6. 16. Santa Monica — 4.50. Median gross rent $2,402 (city) · violent 5.95 and property 38.85 per 1,000 · walkability 16.33 · 26.3% within half a mile of rail · 10.7% of households car-free.
7. 17. Manhattan Beach — 3.40. Median gross rent $3,449 (city) · violent 1.16 and property 21.07 per 1,000 · walkability 15.24 · no resident within half a mile of rail · 2.2% of households car-free.
8. 18. Culver City — 2.95. Median gross rent $2,737 (city) · violent 6.42 and property 47.10 per 1,000 · walkability 15.40 · 9.6% within half a mile of rail · 6.6% of households car-free.

Not ranked, because nobody publishes what they cost. Hancock Park: violent 3.89 and property 41.10 per 1,000, walkability 15.46, 22.5% within half a mile of rail, 3.5% car-free. Larchmont Village: violent 1.33 and property 21.49 per 1,000, walkability 15.22, no resident within half a mile of rail, 8.4% car-free.

A yardstick: the City of Los Angeles as a whole ran 6.74 violent and 22.56 property offenses per 1,000 residents in 2025, against a Department of Finance population of 3,839,536.

1. Koreatown — 7.80

Twenty-four point four percent of Koreatown households own no car. That is not a walkability model's opinion; it is what people did, and it is nearly double the next-highest figure in this set. Three D Line stations sit inside the boundary and sit comfortably inside it — Wilshire/Normandie a full kilometre in, Wilshire/Western 496 metres, Wilshire/Vermont 228 metres and doubling as a B Line transfer — which is why 80.2 percent of residents live within half a mile of rail and 100 percent within a mile, the best coverage of any candidate. At 38,552 people per square mile it is the densest place here by a distance, and it recorded one homicide across 103,977 residents in 2025.

The violent rate of 7.04 is the second-highest on this list and we are not hiding it: 232 robberies and 457 aggravated assaults are real, and they are the reason Koreatown scores 0.4 of a possible 2.0 on the measure most readers care about most. It finishes first anyway, on the strength of the cheapest quartile of rent in the set and the only genuinely car-optional geography in Los Angeles. Note the ZIP: we scored 90005, at a median gross rent of $1,715. Koreatown spans four ZIP codes and three of them put it first.

2= Los Feliz — 7.70

Los Feliz contains no Metro station and has better rail access than three places that do. Vermont/Sunset on the B Line sits 1,041 metres from its population-weighted centre — the shortest such distance of any place here without a station of its own — putting 28.4 percent of residents within half a mile and 72.7 percent within a mile, ahead of Culver City and Studio City, both of which have a platform inside the line.

Two honest deductions. The $1,940 is a ZIP 90027 figure, and 90027 also holds part of East Hollywood and Little Armenia, which drags it down; the true Los Feliz figure is higher and this second place is partly an artifact of the ZIP. And 120 burglaries against 32,701 residents is 3.67 per 1,000, three times Highland Park's rate and well above Eagle Rock's or Atwater Village's. What earns the rank is the combination: 2.39 violent offenses per 1,000 with one homicide, 10.8 percent of households car-free, and rent that the Census Bureau puts below the median of this set.

2= West Adams — 7.70, and the row that moved

On the measures we take best, West Adams wins outright: the highest walkability score of all eighteen at 16.51, with a median of 17.17 across its block groups so it is not one good street dragging up an average; the shortest mean walk to any transit stop in the set at 193 metres; 67.4 percent of residents within half a mile of rail and 100 percent within a mile, because the E Line runs down Exposition Boulevard with Expo/La Brea and La Cienega/Jefferson on the southern edge; 14 percent of households car-free; zero homicides; and 42 burglaries across 21,066 residents, a burglary rate of 1.99 per 1,000 — fifth-lowest of the eighteen.

Then the money, and this is the row where the Zillow decision changed the answer. ZIP 90016's median gross rent is $1,815 — among the lowest in the set — while new units in the same ZIP were asking more than in Mar Vista, Sherman Oaks or Eagle Rock. Both figures are true and they describe two different West Adamses. We are ranking on the first, which is what the households there pay, and we are telling you the second exists, because a 78 percent gap between them is the most important thing happening in this neighborhood.

One more hinge, and it is the largest on the page. Score West Adams on ZIP 90018 instead of 90016 — the eastern half, toward Jefferson Park, where median gross rent is $1,509 — and it scores 8.50 and finishes first outright. The Los Angeles Times polygon is 1.48 square miles, far smaller than the district most listings mean by West Adams. We cannot tell you which line is right, so we are printing both.

4. Highland Park — 7.60

If you want one place on this list whose every number was cleanly measured, this is it. ZIP 90042 maps closely onto Highland Park and Garvanza, and the Times polygon does not fight the ZIP. And the finding is a good one: the lowest property-crime rate of all eighteen places, 11.85 per 1,000, and it is not close — 58 burglaries across 50,906 residents, with zero homicides in 2025.

It owns the burglary floor too, at 1.14 per 1,000. Yet it ranks eleventh of eighteen on violent crime at 3.34, which is precisely why we refuse to publish a single crime score. Two A Line stations sit inside it, Highland Park Station a solid 690 metres in, and since the Foothill extension that platform is a one-seat ride out to Pomona and, through the Regional Connector, into downtown — none of which the 2020-vintage walkability data knows about, so the 14.85 understates it.

5. Eagle Rock — 7.00

Eagle Rock is the one neighborhood figure in this dataset we would defend without an asterisk on geography: ZIP 90041 is essentially the neighborhood, boundary for boundary. That matters, because it means the numbers describe the place rather than a postal convenience. Violent crime 2.30 per 1,000; property crime 14.59, third-lowest of eighteen; two homicides in 2025, which is the detail that keeps this honest, because a low rate is not zero.

The cost of all this is rail. There is none: no station inside, no resident within a mile of one, nearest platform 2,926 metres away at Highland Park. Colorado Boulevard carries the walking, and the EPA index cannot tell that from an equally gridded street with nothing on it.

6. Leimert Park — 6.45

Read the composition before you read the rate. Leimert Park has the highest violent-crime rate of all eighteen places, 12.06 per 1,000 — and a burglary rate of 2.23, 30 burglaries across 13,432 residents, lower than twelve of the other seventeen. Both facts, one neighborhood. The property figure that looks catastrophic is 463 larcenies concentrated on the Crenshaw commercial corridor and around the K Line station, counted against the residents of a 1.17-square-mile neighborhood, most of whom are not who those larcenies happened to. One homicide in 2025.

And then the part almost nobody prints: at $1,612, ZIP 90008 has the lowest median gross rent in this entire set. Three stations sit inside the boundary — Leimert Park and Expo/Crenshaw on the K Line, Expo/Crenshaw on the E — giving 44 percent of residents rail within half a mile and 100 percent within a mile. The entire K Line opened after the 2020 transit data our walkability index runs on, so a full third of that 14.78 is computed as though those platforms do not exist. Leimert Park scores zero on both crime components and finishes sixth regardless, which is the clearest illustration on this page of what our weighting actually does.

7. Mar Vista — 6.30

If the ranking needs one crime-based pick it can defend to anybody, Mar Vista is it. Second-lowest violent rate of eighteen at 1.85 and second-lowest property rate at 14.42, zero homicides — and, unlike Manhattan Beach or Culver City or Santa Monica, it has no large visitor or commercial district loading its numerator against a resident denominator. Its number is closer to an actual risk-to-a-resident figure than almost anything else on this page.

What it costs in the other currency: no rail, no resident within half a mile of a station, only 10.9 percent within a mile, and the longest mean walk to any transit stop of any Los Angeles neighborhood here at 450 metres. The E Line runs along Mar Vista's northern periphery through Sawtelle and West L.A. It does not run through Mar Vista. Five point nine percent of households own no car, and that number is the honest summary of the place.

8= Glendale — 6.10

Glendale finishes dead last of eighteen on the walkability index and the rank is wrong, and we can show you why with the city's own residents. Twelve percent of Glendale households own no car — fourth-highest of all eighteen places, above Los Feliz, above Santa Monica, above Pasadena. The 13.41 is an average taken across 78.9 square kilometres that include the Verdugo Mountains, with block groups running from 4.17 to 18.67; the half-mile core at Brand and Broadway scores 15.86, inside the EPA's most-walkable band.

What Glendale actually is, in the data: 193,810 residents and one homicide in 2025, the third-lowest violent rate of eighteen and the fifth-lowest property rate, the only large jurisdiction here that manages both, with both categories down from 2024. It has no Metro rail at all; the nearest station is eight kilometres away in Highland Park. It does have the Glendale Transportation Center on Metrolink's Antelope Valley and Ventura County lines. We gave it no points for that, because Metrolink is not turn-up-and-go. We are telling you we thought about it.

8= Pasadena — 6.10

A real result held back by the shape of the question. One homicide in a city of 141,466; violent and property crime both down from 2024; six A Line stations inside the city limits — Memorial Park, Del Mar, Fillmore, Lake, Allen and Sierra Madre Villa — putting 27.2 percent of residents within half a mile of rail.

The walkability number, 14.62, is close to meaningless: across 59.5 square kilometres the block groups run from 1.00, the literal floor of the national scale, up in the foothills against the Angeles National Forest, to 19.83, and the half-mile core at Colorado and Fair Oaks scores 19.06 — the highest core figure of any place in this study, above Santa Monica's 18.30. The problem is the same one that lifts it: this is one number for a city that contains both Bungalow Heaven and Northwest Pasadena, and no sub-city breakdown was obtainable, because Pasadena's police department publishes monthly incident PDFs rather than neighborhood rates. Ranking a whole city against individual L.A. neighborhoods is the structural weakness in this candidate list, and Pasadena is where it bites hardest.

10. Silver Lake — 6.00

Silver Lake is this list's clearest proof that a single crime number lies. It sits ninth of eighteen on violent crime at 3.15, the exact middle of the field, and fourth-worst on property at 32.87 — 689 larcenies and 161 vehicle thefts against 30,484 residents. A composite would place it mid-table and tell you nothing true about either risk.

Its rent has a deeper problem: Silver Lake has no ZIP code of its own. It is split between 90026, which it shares with Echo Park, and 90039, which it shares with Atwater Village, and on median gross rent the two differ by 17 percent — $1,929 against $2,250. Neither is a Silver Lake figure. We scored the lower one and printed both; scoring the other drops Silver Lake from tenth to fourteenth. No station sits inside it; Vermont/Santa Monica on the B Line is 1,867 metres out, with 43.7 percent of residents within a mile. And the walkability index counts an intersection on a 15 percent grade exactly as it counts one on the flat.

The hinges — where one choice flips a rank

A ranking that hides its hinges is not worth reading. These are ours, and each one is a place where a defensible alternative choice moves a place materially.

• West Adams, 90016 against 90018. Score the eastern ZIP and West Adams goes from joint second to first outright, at 8.50. The largest single hinge on this page.
• Santa Monica sits two dollars from a band boundary. Its $2,402 is $2 above the $2,400 line, and the Census margin of error on that figure is ±$88. Cross the line and Santa Monica scores 5.30 and finishes fifteenth instead of sixteenth. This is genuinely a coin flip and we are not pretending otherwise.
• Koreatown on 90010 instead of 90005. Median gross rent there is $2,603 and Koreatown falls from first to thirteenth. Three of its four ZIPs put it first; the fourth is also the thinnest sampled.
• Silver Lake on 90039 instead of 90026. Tenth becomes fourteenth.
• Studio City is inside its band by less than its own error. Its $2,535 sits $135 above the $2,400 line and $164 below the $2,700 one, against a Census margin of error of ±$163 — so the true figure could belong in either neighbouring band. And scoring it on ZIP 91602 instead of 91604 gives $2,246, and fifteenth becomes fourteenth.
• Sherman Oaks on 91403 instead of 91423. $2,417 instead of $2,277, and fourteenth becomes fifteenth. We excluded 91436 deliberately: it is mostly Encino, and folding it in would inflate a Sherman Oaks figure past recognition.

Two stress tests, because the recipe is a choice

Take the property-crime column out and rescale. Koreatown still leads at 8.22, Los Feliz and West Adams tie at 7.78, Highland Park is fourth at 7.33 — and Leimert Park moves from sixth to fifth, while Manhattan Beach falls to last at 3.00. One column, one place, and the reason is entirely the larcenies on Crenshaw.

Now take the housing column out entirely and rescale — the answer to the objection that we only ranked expensive places low because they are expensive. Los Feliz and West Adams lead at 7.50, Highland Park is third at 7.33, and Culver City is still last, at 3.58. Santa Monica moves from sixteenth to fourteenth, which is not a rescue. Manhattan Beach jumps from seventeenth to eleventh, which is: cost really is Manhattan Beach's whole problem, and it is not Culver City's. Leimert Park falls from sixth to seventeenth, which tells you exactly how much of its rank is affordability.

The exclusions — the argument, not the leftovers

Santa Monica — 4.50, sixteenth

Santa Monica is the only large place in this study that holds a most-walkable average across its entire municipal boundary rather than in a core: 16.33, with the median exactly equal to the mean, which means it is genuinely uniform rather than one good district lifting a big average. Three E Line stations sit deep inside it. And then the crime, at 5.95 violent and 38.9 property per 1,000, with 596 burglaries and 2,662 larcenies. Both fell sharply from 2024. But 93,876 residents against the Pier, Third Street Promenade and a daily visitor population many times the resident base makes 38.9 a badly misleading answer to the question of whether your house is safe here, and we say so rather than pretending to fix it. Its $2,402 median gross rent against far higher asking rents is not a contradiction either — one of California's strongest rent-control regimes means the two numbers describe two different populations.

Manhattan Beach — 3.40, seventeenth

Manhattan Beach has the lowest violent-crime rate of all eighteen places, 1.16 per 1,000, the only one under 1.5, and zero homicides in 2025. It is also only eighth of eighteen on property crime, at 21.07 — higher than Highland Park's 11.85. The safest-place-in-the-region framing survives only if you say which crime you mean.

It scores sixth on walkability, and that score is picking up a very fine beach-town street grid, because intersection density is a full third of the index — while 2.2 percent of Manhattan Beach households own no car, the lowest figure of all eighteen places, roughly one household in forty-five. There is no Metro rail in the city and no resident within half a mile of any; the nearest platform is Douglas Station on the K Line, 2,358 metres from the centre of town. At $3,449, its median gross rent is the highest in the set by $712, and that is what puts it seventeenth. One number we will not print: the ACS median home value for Manhattan Beach reads $2,000,001, which is the Census top-code rather than a measurement — the margin-of-error field returns a not-applicable flag and the true median is simply unknown.

Culver City — 2.95, last of eighteen

It finishes last, and the interesting part is why, because the obvious answer is wrong. Yes, it posts the highest property-crime rate of all eighteen places at 47.10 per 1,000, and yes, 1,559 of its 1,914 property offenses are larceny in a five-square-mile city whose retail and studio districts draw a daytime population far exceeding its 40,640 residents — a resident denominator against a regional numerator. But robbery is genuinely elevated at 119, against 148 in Glendale, a city with nearly five times the population, so this is not purely an artifact. Zero homicides.

Now the two things Culver City is famous for. Walkability: 15.40 average with a median of 16.67, so most of the city is better than its headline — except that the half-mile core around Culver Boulevard and Main Street scores 15.20, slightly below the city average. Rail: the E Line's station called Culver City Station is inside city limits by 42 metres and sits 2,696 metres from the population-weighted centre of town, serving the Hayden Tract and the Palms edge. Only 9.6 percent of residents live within half a mile of it. Do not write that downtown Culver City is on the E Line. It is not. Delete the housing column entirely and Culver City is still last.

Hancock Park — not ranked, and this is the lead

The reasoning is at the top of this piece. What we can establish is not an argument for a high placement anyway: 103 burglaries against 11,826 residents — 8.7 per 1,000, the highest burglary rate of all twenty places we examined, by a wide margin — a property rate of 41.10, third-worst overall, and 3.5 percent of households car-free, second-lowest in the set, against a walkability score of 15.46 that ranks third and is rewarding a tight pre-war grid and Wilshire frontage rather than car-independence. On the five components we can measure, Hancock Park scores 2.95 of a possible 6.0.

Two fair hedges: the Times polygon includes Wilshire Boulevard frontage whose non-resident traffic loads the numerator, and 46 violent offenses on 11,826 people is a small count. And one real correction in its favour: Wilshire/La Brea on the D Line is inside the boundary, five metres inside, at the far southwest corner — and it opened on 8 May 2026, so the walkability index is scoring Hancock Park as though its subway station does not exist.

Larchmont Village — not ranked

No publisher breaks it out either: no Census place or ZIP code tabulation area, no HUD sub-ZIP geography. It is roughly six commercial blocks and the streets around them, entirely inside ZIP 90004 — smaller than the finest geography any of these publishers reports. Its crime rate rests on eleven violent offenses inside a 0.48-square-mile polygon; three more incidents move that rate by 27 percent. Its walkability average rests on seven block groups, the smallest sample of the twenty places we examined. The recommendation from the reporting was to merge it into Hancock Park with a line in the methodology or drop it. We dropped it, and said why.

Bel Air — no data, and the absence is the finding

Bel Air appears in none of the four data pulls behind this piece: no crime rate, no walkability score, no rent figure, nothing. We name it as an absence, not as a rank, and we put no number next to it.

Sixteen things this data does not say

These are the caveats that would embarrass us if a reader found them before we printed them.

1. Median gross rent is not asking rent. It is what current tenants pay, including tenant-paid utilities, averaged over five years and expressed in 2024 dollars. A newcomer signing a lease this month will pay more, in some places far more.
2. Never subtract one from the other and call it rent growth. The gap between a five-year average and a current listing is mostly definitional.
3. ZIP codes are not neighborhoods, and three pairs collide. Echo Park and southern Silver Lake are both 90026. Atwater Village and northern Silver Lake are both 90039. Hancock Park and Larchmont are both 90004. Where two places tie on rent, the tie is about the post office.
4. HUD's fair market rents are not market rents and are not scored here. They are 40th-percentile gross rents set administratively to price housing vouchers.
5. The eighteen places are not measured on the same instrument. Six are whole incorporated cities reporting summary UCR to the California Department of Justice; twelve are Los Angeles Times neighborhood polygons where we extracted and geocoded LAPD's 2025 NIBRS records ourselves.
6. Our neighborhood method runs about 7.1 percent low on violent crime — 24,022 offenses against 25,862 citywide — essentially all of it in aggravated assault, because California's summary definition captures cases NIBRS codes as simple assault. Do not read a 0.3-per-1,000 gap between a neighborhood and a city as real.
7. The denominators are different vintages — 2020 Census for neighborhoods, Department of Finance January 2025 for cities. The City of Los Angeles fell 1.5 percent between those dates, so neighborhood rates run slightly high.
8. These are crimes per resident, not risk to a resident. In six places — Culver City, Santa Monica, Leimert Park, Hancock Park, Larchmont and Koreatown — a large commercial, retail, transit or visitor district sits in the numerator while only residents sit in the denominator. We could not correct it; daytime-population data does not exist on these boundaries.
9. Small counts are not stable rates. Larchmont's rests on 11 offenses, Atwater Village's on 26, Manhattan Beach's on 40, Hancock Park's on 46, South Pasadena's on 73. Never write that one of them is safer than another.
10. There is no trend column and we could not build one. LAPD's 2024 records are an artifact of a March 2024 system cutover — 43,242 geocoded records for 2024 against 110,434 for 2025. Six cities have a usable 2024. Twelve neighborhoods do not. A trend column would have been real for six places and fabricated for twelve.
11. The EPA walkability index does not know about any transit built since 2020, and transit is a third of it. The K Line, the Regional Connector, the A Line's Foothill extension and the D Line extension to Wilshire/La Brea — which opened 8 May 2026 — are all invisible to it. Leimert Park, Highland Park and Hancock Park are all scored lower than they should be.
12. A whole-city average is not a neighborhood. Pasadena's 14.62 spans 59.5 square kilometres and runs from 1.00 to 19.83. Glendale's 13.41 spans 78.9.
13. The index measures street form, not whether the walk is any good. Nothing in it captures sidewalk condition, shade, slope, lighting, traffic speed or whether there is a grocery store at the end. Where the index and the car-free share disagree, the car-free share is observed behaviour and the index is a model — and the car-free share reflects low incomes as well as good transit.
14. Containing a station is not being served by one. Eight of our in-boundary stations sit within 100 metres of a border. Culver City Station is 2.7 kilometres from downtown Culver City. Universal/Studio City is across the 101 from Ventura Boulevard. Los Feliz contains no station and has better rail coverage than three places that do. Rail figures are our own computation from Metro's published schedule feed, pulled 20 August 2026; Metro is the provider of that information and is not responsible for what we did with it.
15. Reported crime is reported crime. It measures what residents call the police about and what the police then record. Reporting rates differ by income, insurance, immigration status and trust in the department. A low number can mean a safe place or a place that does not call, and this data cannot tell them apart — which matters most in a ranking that will be read as a verdict on neighborhoods with very different relationships to the LAPD.
16. Nothing here measures schools, air quality, heat, tree canopy, flood or fire risk, park access, or pedestrian injury. Several of them would reorder this list completely.

What happens to this page next

This ranking is not going to be replaced. It lives at one address and it gets updated in place, with the date of its last real check printed on it. Rent is the fastest-moving figure under it, so the sources are re-fetched quarterly; the crime file is annual and the walkability index has not been republished since 2021, so those are re-checked by hand on their own clocks. If a source moves, the page changes and the date changes with it. If a source cannot be reached, nothing changes and nothing is re-dated — an unread source is not a check.