↖︎ Vishal Singh
Contents

Zillow housing data · Chapter 5 of 5

Zillow housing data · Chapter 5 of 5

Neighborhoods, using New York City

From October 2016 to October 2026 the median Manhattan neighborhood's estimate fell 10.3%, while the medians in the other four boroughs rose 30% to 58%.

Author
AffiliationNYU Stern
Published

Zillow, November 2026 release

From October 2016 to October 2026 the median Manhattan neighborhood's estimate fell 10.3%, while the medians in the other four boroughs rose 30% to 58%.

Author

Vishal Singh

Affiliation

NYU Stern

Published

November 19, 2026

Data

Zillow, November 2026 release

Zillow's November 2026 release estimates a typical home value each month for 228 New York City neighborhoods [1], 30 more than its October release. In October 2026 the estimates ran from $209,100 in Pelham Parkway, in the Bronx, to $3.73 million in Tribeca, a ratio of about 18 to 1. All dollar figures are nominal unless labeled.

Over the ten years from October 2016, 22 of Manhattan's 27 neighborhoods with values in both months ended lower. Outside Manhattan, 19 of 169 did: 6 in Brooklyn, 8 in Queens, 5 in the Bronx and none on Staten Island. Across the 196 neighborhoods with both values, the median change was +46.0%. New York City serves here as the example of Zillow's neighborhood files, which hold 23,568 neighborhood series across the country (Chapter 2); the same steps apply to any of them. The chapter first describes what a Zillow neighborhood is and how its lines relate to census geography, then follows the values on a map, by borough, and neighborhood by neighborhood.

What a Zillow neighborhood is

Zillow has not published a method for drawing its neighborhood lines, and its data page offers no boundary file. The most recent public copy is the one the U.S. Environmental Protection Agency posted in 2017, which says only that Zillow created the boundaries and that they are not formal or jurisdictional lines [2]. For New York City it holds 278 polygons for 274 neighborhoods; 4 IDs are listed twice, under two counties [3]. Each is a named place, such as Cobble Hill or Co-op City, and parks, islands and airports count as neighborhoods. Together they cover 99.65% of the city's land (302.1 square miles) without overlapping. Their areas run from 0.008 to 8.1 square miles.

Table 1 sets them beside the two kinds of small areas that census and city data use. The median Zillow neighborhood covers 0.86 square miles, close to the median Neighborhood Tabulation Area (NTA), which City Planning builds from whole 2020 census tracts [4] [5]. Zillow's lines are not built from tracts. Of the city's 2,325 tracts, 1,786 (76.8%) have at least 95% of their land in one Zillow neighborhood. On Staten Island 42.9% of tracts have less than 80% of their land in any one neighborhood, against 10.3% to 11.9% in the other boroughs. Only 22 of the 274 neighborhoods closely match an NTA (an overlap of 0.8 or more in Table 1).

Table 1. Zillow's neighborhoods are about the size of the city's NTAs and 12 times the size of its census tracts, and are not built from either
Three sets of small areas in New York City: count, median land area, who draws them, and how they line up with Zillow's 2017 neighborhoods
AreasNumber in NYCMedian land area, sq miDrawn byFit with Zillow neighborhoods
Zillow neighborhoods, 20172740.86Zillow, as named places (parks, islands and airports included)199 have a series in the November 2026 release; 75 do not
Census tracts, 20202,3250.07U.S. Census Bureau, as statistical units1,786 (76.8%) have at least 95% of their land in one Zillow neighborhood; the median neighborhood touches 7 tracts
Neighborhood Tabulation Areas (NTAs), 20202620.89NYC Department of City Planning, from whole tractsMedian overlap with the best-matching NTA 0.41; 22 of 274 neighborhoods reach 0.8 or more
Table 1. Zillow's neighborhoods sit between tracts and NTAs in how they are drawn and match the size of NTAs, but most do not line up with either. Overlap is the area a Zillow neighborhood shares with an NTA divided by the area the two cover together (1 means the same shape). Areas are land only, using City Planning's shoreline-clipped tracts.
Source: Zillow neighborhood boundaries, EPA copy of the 2017 release [2]; overlap and nesting from nyc/geo/out/summary.json [3], using NYC City Planning 2020 tracts [4] and NTAs [5]; series counts from the Zillow Home Value Index by neighborhood, November 2026 release [1]. Counts of areas, unweighted.

To move tract data, such as American Community Survey counts, onto Zillow neighborhoods, nyc/geo/out/zillow_tract_crosswalk.csv lists each piece of a tract inside a neighborhood (4,155 pieces) with its share of the tract's land, 2020 population and 2020 housing units [3]. The population and housing shares come from placing each of the city's 37,984 2020 census blocks in the neighborhood that contains the block's Census internal point [6]. Counted by people, 84.7% of the 2,292 tracts with residents have at least 95% of their population in one neighborhood, and 15.8% of New Yorkers live in tracts below that line.

Home values cover only some of these polygons. Of the release's 228 New York City series, 199 share a RegionID with a 2017 polygon. The other 29 are Hell's Kitchen and Hudson Yards, whose IDs were created after 2017, and 27 of the 30 series added in November (10 in the Bronx, 7 in Manhattan, 7 in Queens, 3 in Brooklyn and 3 on Staten Island). With no public boundaries, those 29 cannot be mapped. All the added series are short. Of the 30, 26 begin in October 2024 and the other 4 later: Arverne (December 2024), Van Cortlandt Park (December 2025), Marble Hill (October 2026) and Civic Center (October 2026). The release also adds a Little Neck series while the 2017 polygon named Douglaston-Little Neck carries the Douglaston series, so Zillow's current lines probably differ from the 2017 copy.

Of the 274 polygons, 75 have no series. In 2020 they held 953,310 people, 10.8% of the city [6]. Of these, 23 are parks, airports and islands with fewer than 200 housing units; the most populous of the rest are South Bronx (98,828 residents), Murray Hill in Queens (78,573) and University Heights (54,522).

Neighborhood values on the map

The highest October 2026 estimates were in Tribeca ($3.73 million) and SoHo ($3.24 million) in Manhattan, followed by Nolita ($2.75 million) and Central Park South ($2.21 million), series added in this release, and by Cobble Hill ($2.19 million) and Carroll Gardens ($2.11 million) in Brooklyn. The lowest were in the Bronx: Pelham Parkway ($209,100, an added series), Parkchester ($263,100) and Bedford Park ($263,600). In Figure 1 the slider steps through each October from 2000 to 2026, and the buttons switch to the 1-year and 10-year changes.

For 48 of the 75 polygons without a series, Figure 1 shows an estimate built from Zillow's ZIP-code series and marks it with dots. Another 23 are parks, airports and islands, and 4 have no estimate. The map also fills a neighborhood's years before its own series begins with the ZIP-based estimate. Every number in the text comes from Zillow's 228 neighborhood series.

In October 2026, estimates ranged from $209,100 in Pelham Parkway (not on the map) to $3.73 million in Tribeca
Oct 2026
Figure 1. Lower Manhattan and northwest Brooklyn have the highest estimates, and the Bronx the lowest. Each shape is a Zillow neighborhood, colored by Zillow's typical home value in October of the chosen year or by its percent change over 1 or 10 years to October 2026. The value scale is logarithmic: equal steps in color are equal percent differences. Dotted shapes have no Zillow neighborhood series for that measure; their values are estimated from Zillow's ZIP-code series, weighting each ZIP code by the neighborhood's 2020 housing units in it. Gray shapes are parks, airports and islands with fewer than 200 homes in 2020, and hatched shapes have no estimate. The 29 series without a 2017 boundary appear in Table 2 but not on the map; the 2017 polygon Clinton, with 51% of its land in City Planning's Hell's Kitchen NTA, probably covers much of the same area. Hover for details; click a neighborhood to add it to Figure 2.
Source: Zillow Home Value Index, all homes, mid tier, smoothed and seasonally adjusted, nominal dollars, October 2000 to October 2026, November 2026 release [1]; Zillow neighborhood boundaries, 2017 release [2]. Dotted areas: Zillow Home Value Index by ZIP code (176 ZIP codes) [1], 2020 ZIP Code Tabulation Areas [7] and 2020 Census housing units by block [6], weighted by housing units. Other areas: one Zillow estimate per neighborhood, no weighting. Change colors stop at ±75% (10-year) and ±15% (1-year). Basemap © OpenStreetMap contributors, © CARTO.

Manhattan's median fell while the other boroughs' medians rose

Figure 2 follows each borough's median neighborhood: for every month, the estimate of the neighborhood in the middle when the borough's neighborhoods are ranked by value. It uses the 196 neighborhoods with a value in October 2016, so the series added in this release do not shift the lines in their first month. Manhattan's median peaked at $1.38 million in August 2017. By October 2026 it was $1.19 million, 14.0% below the peak and 10.3% below October 2016 ($1.32 million).

From October 2016 to October 2026 the Brooklyn median rose 29.7% (from $853,300 to $1.11 million), Queens 48.6% ($505,100 to $750,800), the Bronx 48.2% ($408,200 to $604,900) and Staten Island 58.3% ($467,100 to $739,600). The Brooklyn, Queens and Staten Island medians were at their highest since 2000 in October 2026. The Bronx median was 0.6% below its August 2026 high. Brooklyn's median was still below Manhattan's, but the gap had narrowed from $469,000 to $79,600.

Manhattan's median neighborhood has fallen since 2017; the other four boroughs' medians rose to or near their highest since 2000
Median across each borough's neighborhoods, monthly, nominal dollars
Borough medians as a table
BoroughOct 2000Oct 2016Highest (month)Oct 2026Oct 2016 to Oct 2026
Manhattan$425,700$1.32 million$1.38 million (August 2017)$1.19 million−10.3%
Brooklyn$241,500$853,300$1.11 million (October 2026)$1.11 million+29.7%
Queens$245,900$505,100$750,800 (October 2026)$750,800+48.6%
Bronx$223,100$408,200$608,500 (August 2026)$604,900+48.2%
Staten Island$214,200$467,100$739,600 (October 2026)$739,600+58.3%
Figure 2. The Manhattan line peaks in August 2017 and declines; the other four lines rise through 2026. Each line is, for every month, the median across the borough's neighborhoods of Zillow's estimate, in nominal dollars, in October 2026 dollars, or as an index with October 2016 set to 100. The dashed line marks October 2016.
Source: Zillow Home Value Index, all homes, mid tier, smoothed and seasonally adjusted, March 2000 to October 2026, November 2026 release [1]; CPI-U, seasonally adjusted [8]. Unweighted medians across the 196 neighborhoods with a value in October 2016 (27 to 49 per borough); each neighborhood counts once regardless of size.

Neighborhoods with lower 2016 values gained more

Figure 3 plots each neighborhood's October 2016 estimate against its change to October 2026. Across the 196 neighborhoods with both values, the correlation between the logarithm of the 2016 value and the 10-year change is −0.66 (a correlation runs from −1 to 1, and 0 means no linear relationship). It is negative in every borough: −0.77 in Brooklyn, −0.77 on Staten Island, −0.53 in Manhattan, −0.51 in the Bronx and −0.22 in Queens.

The largest gains were in Jamaica, Queens, where the estimate doubled from $354,700 to $713,200 (+101.1%), Parkchester in the Bronx (+97.6%, from $133,100 to $263,100) and Morris Heights in the Bronx (+86.0%). The largest declines were in Highbridge in the Bronx (−54.4%) and in four Manhattan neighborhoods: the Financial District (−38.9%, from $1.70 million to $1.04 million), Hell's Kitchen (−35.0%), the West Village (−34.4%) and the Flatiron District (−30.1%).

In every borough, neighborhoods with lower 2016 values rose more by 2026
Horizontal: typical home value, October 2016 (log scale). Vertical: percent change, October 2016 to October 2026.
Figure 3. In each panel the dots slope down from left to right: the lower a neighborhood's October 2016 estimate, the larger its change over the next ten years. Each dot is a neighborhood. Colored dots belong to the panel's borough and gray dots show the rest of the city for reference; r is the correlation within the borough. The largest and smallest changes in each borough are labeled.
Source: Zillow Home Value Index, all homes, mid tier, smoothed and seasonally adjusted, nominal dollars, October 2016 and October 2026, November 2026 release [1]. 196 neighborhoods with both values; unweighted; r is the Pearson correlation of the log 2016 value with the percent change.

A second summary of the distance between expensive and inexpensive neighborhoods is the ratio of the 90th-percentile value (higher than 90% of neighborhoods) to the 10th-percentile value. Among the 159 neighborhoods with an estimate in every month since March 2000, the ratio rose from 2.32 in March 2000 to 3.14 in May 2015, stood at 3.04 in October 2016, and fell to a low of 2.43 in November 2024 (Figure 4). It then rose to 2.68 in October 2026; between those two months the 10th-percentile value went from $523,800 to $530,300 and the 90th-percentile value from $1.28 million to $1.42 million.

The 90th-percentile neighborhood value fell from 3.04 times the 10th percentile in October 2016 to 2.43 times in November 2024, then rose to 2.68 times by October 2026
Ratio of the 90th- to the 10th-percentile typical home value across 159 neighborhoods, monthly
Figure 4. The ratio climbs from 2000 to a high in May 2015, falls to a low in November 2024 and rises after that. Each month, the 90th-percentile neighborhood value divided by the 10th-percentile value, computed over the same 159 neighborhoods throughout.
Source: Zillow Home Value Index, all homes, mid tier, smoothed and seasonally adjusted, March 2000 to October 2026, November 2026 release [1]. Fixed panel of 159 neighborhoods with no missing months; unweighted.

The 12 months to October 2026

Across the 225 neighborhoods with values in both October 2025 and October 2026, the median 12-month change was +4.5%. Estimates fell in 27: 9 in Manhattan, 9 in Queens, 6 in the Bronx, 2 in Brooklyn and 1 on Staten Island. Of the 186 neighborhoods with a series since March 2000, 123 reached their highest estimate in 2026, including 47 of Staten Island's 48 and 1 of Manhattan's 23 (Battery Park).

The Bronx had the three largest declines, in Bedford Park (−10.3%), Kingsbridge (−8.9%) and Concourse (−8.4%), and the two largest gains, in Van Nest (+17.2%) and Morrisania (+16.8%), a series added in this release. Among the other boroughs, the largest gain was in Borough Park, Brooklyn (+15.1%).

All 228 neighborhoods

Table 2 lists every neighborhood with a Zillow series, including the 29 that are not on the map. Select a column heading to sort.

Table 2. October 2026 estimates ranged from $209,100 in Pelham Parkway to $3.73 million in Tribeca
Typical home value and change by neighborhood, nominal dollars
Source: Zillow Home Value Index, all homes, mid tier, smoothed and seasonally adjusted, nominal dollars, November 2026 release [1]. Changes run October to October. A dash means the series starts after the base month. Sparklines show March 2000 to October 2026, each on its own scale.

What these data cannot show

The Zillow Home Value Index is Zillow's model estimate of the value of a typical home, built from its estimates for individual homes [9]. The series here covers all homes (single-family, condominium and co-op) in the middle third of values, smoothed and seasonally adjusted. Where few homes sell, the estimate rests on fewer transactions and can move for reasons other than market prices. Zillow publishes no margin of error.

The inflation adjustment uses the national CPI-U, which rose 39.6% from October 2016 to October 2026; prices in the New York area may have risen at a different rate. Borough medians count each neighborhood once, so a neighborhood with 2,000 residents weighs as much as one with 200,000. The set behind them grows from 187 neighborhoods in October 2000 to 196 in October 2016.

The map uses Zillow's 2017 boundaries, and Zillow does not say which boundaries the 2026 estimates use [2]. The 30 series added in this release have at most 25 months of history, so they enter only the October 2026 values and, where an October 2025 value exists, the 12-month changes.

The ZIP-based values on the map are approximations. For the 198 neighborhoods that have both a Zillow series and a ZIP-based estimate, the estimate differs from Zillow's October 2026 value by a median of 7.3%, and by 29.8% or more for the tenth with the largest differences. Their 10-year changes differ by a median of 7.3 percentage points (correlation 0.82). ZIP Code Tabulation Areas also only approximate the ZIP codes Zillow uses.

The ten-year comparisons depend on the starting month. Manhattan's median in October 2016 was 4.2% below its August 2017 peak. The figures describe where estimated values moved and do not identify why.

Exercise

The five-neighborhood exercise (nyc/predict_2021.py, results in nyc/README.md) picks five neighborhoods by each of four rules that use only values through December 2021, such as the largest two-year gain or the lowest value, and scores the picks on growth from December 2021 to September 2026. It was built on the October 2026 release (data through September 2026). Rerun it on the November release with October 2026 as the end month; none of the 30 added series has a December 2021 value, so the set of neighborhoods it can pick from does not change.

Data sources

  1. Zillow Home Value Index (ZHVI), all homes, mid tier, smoothed and seasonally adjusted, by neighborhood (Neighborhood_zhvi_uc_sfrcondo_tier_0.33_0.67_sm_sa_month.csv) and by ZIP code (Zip_zhvi_uc_sfrcondo_tier_0.33_0.67_sm_sa_month.csv), November 2026 release (server date November 16, 2026), downloaded November 19, 2026; data through October 31, 2026. Neighborhoods with City "New York" and State "NY", borough from CountyName. The October 2026 release (downloaded November 11, 2026) is used only to identify added series. zillow.com/research/data
  2. Zillow neighborhood boundaries, 2017 release hosted by the U.S. EPA (layer 3, Neighborhood Areas), queried for the five NYC counties on November 18, 2026 (nyc/geo/sources/zillow_neighborhoods_nyc_epa2017_accessed_2026-11-18.geojson). catalog.data.gov
  3. Crosswalk and comparison outputs built on November 18, 2026 by nyc/geo/zillow_geo_crosswalk.py: nyc/geo/out/summary.json, zillow_neighborhoods.csv, tract_fit.csv, zillow_tract_crosswalk.csv and block_assignment.csv; described in nyc/geo/README.md.
  4. NYC Department of City Planning, 2020 census tracts clipped to shoreline and 2020 Neighborhood Tabulation Areas, accessed November 18, 2026 (nyc/geo/sources/).
  5. 2020 Census block population (POP100), housing units (HU100) and internal points from TIGERweb, accessed November 18, 2026; 2020 ZIP Code Tabulation Areas from TIGERweb for the NYC area, accessed November 18, 2026.
  6. CPI-U, all items, U.S. city average, seasonally adjusted (CPIAUCSL), via FRED, accessed November 18, 2026 (booklet/sources/cpi_u_sa_CPIAUCSL_fred_accessed_2026-11-18.csv).

Methods

Sample. 228 Zillow neighborhoods in the five boroughs with a ZHVI series in the November 2026 release, monthly from March 2000 to October 2026 (198 in the October 2026 release; 30 added, none dropped). 27 series are missing 1 to 3 interior months; these are filled by linear interpolation. Series are left empty before their first month.

Changes. October to October: 1 year (2025 to 2026), 10 years (2016 to 2026) and since 2000 (October 2000 to October 2026). "Highest in 2026" counts use each neighborhood's highest monthly estimate since March 2000, among the 186 series that start in that month.

Summaries. Borough and citywide medians are unweighted medians, each month, across the 196 neighborhoods with a value in October 2016 that have a value that month. Including the series added in this release would change the October 2026 borough medians by −7.9% (Manhattan), −0.5% (Brooklyn), −0.7% (Queens), 0.0% (Bronx) and −0.5% (Staten Island); the difference starts in October 2024, when they begin. The 90th/10th percentile ratio uses a fixed panel of 159 neighborhoods with no missing months. The inflation adjustment divides by CPI-U and rescales to October 2026. The FRED file has no value for October 2025; for the October 2026 dollar lines in Figure 2 that month is filled by geometric interpolation between its neighbors. Correlations are Pearson correlations of the natural log of the October 2016 value with the 10-year percent change. No survey weights are involved.

Section 1. Geometry facts (areas, tract and NTA overlap, block assignment) are from the November 18, 2026 run of nyc/geo/zillow_geo_crosswalk.py; tract shares by borough are recomputed from tract_fit.csv. Counts of series and of matches to the 2017 polygons are recomputed here for the November 2026 release. Matching is on RegionID only.

Map. Zillow's 2017 polygons (278 features; 4 IDs listed under two counties are merged, leaving 274) are matched to series on RegionID, which links 199 of 228 series. Boundaries are simplified with a 25-foot tolerance that keeps shared edges aligned (from 49,754 to 22,154 vertices). Population and housing units are from 2020 Census blocks, each assigned to the neighborhood containing its internal point. Built with MapLibre GL JS 4.7.1 on a CARTO basemap. Highbridge's neighbors are the polygons within 50 feet of its 2017 boundary.

Map gaps. Each 2020 block is also assigned to a 2020 ZIP Code Tabulation Area by its internal point. Where a polygon has no Zillow series, or before its series begins, the map shows the housing-unit-weighted mean of Zillow's ZIP-level ZHVI (176 NYC ZIP codes in the November release, 172 in the file the earlier article used) over the polygon's blocks. The estimate is shown only when ZIP values cover at least 50% of the polygon's 2020 housing units. Polygons without a series and with fewer than 200 housing units in 2020 are shown as parks, airports and islands. These estimates appear only on the map and its tooltips. The method follows nyc/article/build_article.py, which built an earlier version of this analysis on the October 2026 release with September as the reference month.

How to cite

@misc{singh2026zillowch5,
  author = {Singh, Vishal},
  title  = {Neighborhoods, using New York City},
  year   = {2026},
  note   = {In: Zillow housing data. NYU Stern. Data: Zillow Research},
  url    = {https://vishalsingh.org}
}

Acknowledgments

Home-value estimates and neighborhood boundaries are from Zillow Research; Zillow requires clear attribution when its data are used.

References

  1. Zillow Research. Housing data: Zillow Home Value Index (ZHVI), neighborhood and ZIP code levels. https://www.zillow.com/research/data/ (accessed November 19, 2026).
  2. U.S. EPA Office of Environmental Information. Neighborhoods, US, 2017, Zillow. Boundaries created by Zillow. https://catalog.data.gov/dataset/neighborhoods-us-2017-zillow-segs
  3. Singh, V. Zillow neighborhood boundaries in NYC, compared with census tracts and NTAs. nyc/geo/README.md and nyc/geo/out/, November 18, 2026.
  4. NYC Department of City Planning. NYC Census Tracts for 2020 US Census (clipped to shoreline). services5.arcgis.com
  5. NYC Department of City Planning. NYC Neighborhood Tabulation Areas 2020. services5.arcgis.com
  6. U.S. Census Bureau. TIGERweb, Census 2020 blocks, with 2020 Census population (POP100) and housing units (HU100). https://tigerweb.geo.census.gov/arcgis/rest/services/TIGERweb/Tracts_Blocks/MapServer/12
  7. U.S. Census Bureau. TIGERweb, 2020 Census ZIP Code Tabulation Areas. https://tigerweb.geo.census.gov/arcgis/rest/services/TIGERweb/PUMA_TAD_TAZ_UGA_ZCTA/MapServer/7
  8. U.S. Bureau of Labor Statistics. Consumer Price Index for All Urban Consumers: All Items in U.S. City Average (CPIAUCSL), retrieved from FRED, Federal Reserve Bank of St. Louis. https://fred.stlouisfed.org/series/CPIAUCSL
  9. Olsen, S. Zillow Home Value Index methodology, 2023 revision: what's changed? Zillow Research, February 10, 2023. https://www.zillow.com/research/methodology-neural-zhvi-32128/