Neighborhoods hit hardest by the 2008 foreclosure crisis have recovered less — and several remain below their pre-crisis peak once inflation is counted

Summary

Not fully, and unevenly. Nationally, home prices are up about 95.5% in nominal terms since the January 2007 peak — but after adjusting for 19 years of inflation, that is only +22.4% in real (constant) dollars. The metros that were the epicenter of the 2008 foreclosure crisis (Las Vegas, the California Inland Empire/Central Valley, and Southwest Florida) fell far further than the nation (50-63% peak-to-trough, vs. -21% nationally) and have not made up that ground: Las Vegas, Riverside-San Bernardino, Stockton-Lodi and Cape Coral are still below their 2007 peak in real, inflation-adjusted terms as of mid-2026, even after the largest nominal home-price boom in FHFA's history. Detroit, Orlando, Phoenix and Miami have exceeded their 2007 peak in real terms, but by less than the national gain, except Miami and Phoenix which are close to or above it. Within these metros the picture is worse still: ZIP-code-level research (Zillow) finds the specific neighborhoods that had the highest foreclosure rates recovered far less than low-foreclosure ZIPs in the same metro — gaps of 25-60 percentage points in recovery rate in cities like San Diego, San Francisco, Atlanta, Detroit and Cleveland — meaning the true neighborhood-level shortfall is larger than even these already-lagging metro averages show.

Method and grain limitation

This corpus's finest housing-price geography is the Core-Based Statistical Area (metro), via FHFA's House Price Index (housing.house_price_index, purchase-only, seasonally adjusted, quarterly). There is no ZIP-code or Census-tract price index in this warehouse, so 'neighborhood' here is approximated by the eight U.S. metro areas most identified in the post-crisis literature as foreclosure epicenters: Las Vegas-Henderson NV, Riverside-San Bernardino-Ontario CA (Inland Empire), Stockton-Lodi CA, Cape Coral-Fort Myers FL, Phoenix-Mesa-Chandler AZ, Orlando-Kissimmee-Sanford FL, Miami-Miami Beach-Kendall FL, and Detroit-Dearborn-Livonia MI. This is a coarser unit than 'neighborhood' and understates true dispersion (see the Zillow ZIP-level findings below, which show the gap between high- and low-foreclosure ZIPs within the same metro can exceed 50 percentage points). For each metro, peak = max quarterly index value 2005-2008; trough = min value 2008-2013; latest = 2026 Q2. Real terms use the BLS CPI-U cumulative deflator from 2007 to 2026 (partial-year, ~59.7% cumulative inflation) via this connector's adjust_inflation tool.

How far these markets fell, and how far they've come back

Every one of these eight crisis-epicenter metros still trails the national real recovery rate (+22.4%), and half of them (Las Vegas, Riverside-San Bernardino, Stockton-Lodi, Cape Coral) have never gotten back to their 2007 peak once inflation is accounted for — meaning a home bought at the 2007 top in one of those markets and held for 19 years has still lost purchasing-power value, even through the 2020-2022 pandemic housing boom.

Within-metro neighborhood disparities are larger still

Zillow Research's ZIP-code-level analysis (grouping ZIPs within each metro into high-, medium-, and low-foreclosure tiers based on peak 2007-2009 monthly foreclosure rates) found that the metro-level averages above mask much bigger internal gaps: in San Diego, only 28.2% of homes in high-foreclosure ZIPs had recovered their pre-recession peak value vs. essentially all homes in low-foreclosure ZIPs — a 59.8 percentage-point gap, the largest of any major metro studied. San Francisco (49.9pp gap), Atlanta (38.1pp), Detroit and Cleveland (both >25pp) showed the same pattern: metros that look fully recovered on a median-home-value basis still have specific formerly-hard-hit neighborhoods where a majority of homes have not regained 2007 values. This is consistent with, and larger than, the metro-level real-dollar shortfalls computed above — the actual neighborhood-level story is worse than the metro averages suggest, not better.

Caveats

• Metro grain, not neighborhood/ZIP grain — this connector has no ZIP or tract-level home price index; the Zillow ZIP-tier findings are cited from that firm's own published research, not recomputed here.
• The eight metros were selected from the post-crisis foreclosure literature (Pew, FCIC final report, Brookings) as commonly cited epicenters, not from a formal foreclosure-rate ranking inside this corpus, since no foreclosure-rate-by-metro/ZIP series is loaded here either.
• FHFA's purchase-only index reflects repeat sales of the same homes and excludes cash/distressed-sale-only transactions in some periods; it is the standard federal home-price series but is not identical to Case-Shiller or Zillow's own ZHVI, which can show different peak/trough magnitudes for the same metro.
• 2026 figures use a partial-year (7-month) CPI-U average for the deflator; the ~59.7% cumulative inflation figure would shift slightly once the full 2026 CPI year is published.

Every query behind this report

6 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

SELECT DISTINCT place_id, place_name FROM housing.house_price_index WHERE level='MSA' AND (place_name ILIKE '%Las Vegas%' OR place_name ILIKE '%Phoenix%' OR place_name ILIKE '%Riverside%' OR place_name ILIKE '%Stockton%' OR place_name ILIKE '%Modesto%' OR place_name ILIKE '%Merced%' OR place_name ILIKE '%Cape Coral%' OR place_name ILIKE '%Miami%' OR place_name ILIKE '%Detroit%' OR place_name ILIKE '%Orlando%')

SELECT hpi_flavor, frequency, level, COUNT(*), COUNT(index_sa) FROM housing.house_price_index WHERE level='MSA' AND hpi_type='traditional' GROUP BY hpi_flavor, frequency, level

SELECT place_id, place_name, data_year, period, index_sa
FROM housing.house_price_index
WHERE level='MSA' AND hpi_type='traditional' AND hpi_flavor='purchase-only' AND frequency='quarterly'
AND place_id IN ('38060','19804','29820','33700','40140','15980','33124','44700','32900','36740')
ORDER BY place_id, data_year, period

SELECT place_id, place_name, data_year, period, index_sa
FROM housing.house_price_index
WHERE level='USA or Census Division' AND hpi_type='traditional' AND hpi_flavor='purchase-only' AND frequency='quarterly' AND place_id='USA'
ORDER BY data_year, period

WITH d AS (
  SELECT place_id, place_name, data_year, "period", index_sa,
    ROW_NUMBER() OVER (PARTITION BY place_id ORDER BY data_year DESC, "period" DESC) AS rn_latest
  FROM housing.house_price_index
  WHERE level='MSA' AND hpi_type='traditional' AND hpi_flavor='purchase-only' AND frequency='quarterly'
  AND place_id IN ('38060','19804','29820','33700','40140','15980','33124','44700','32900','36740')
),
peak AS (
  SELECT place_id, MAX(index_sa) AS peak_idx FROM d WHERE data_year BETWEEN 2005 AND 2008 GROUP BY place_id
),
trough AS (
  SELECT place_id, MIN(index_sa) AS trough_idx FROM d WHERE data_year BETWEEN 2008 AND 2013 GROUP BY place_id
),
latest AS (
  SELECT place_id, place_name, index_sa AS latest_idx, data_year, "period" FROM d WHERE rn_latest = 1
)
SELECT l.place_id, l.place_name, peak_idx, trough_idx, latest_idx, l.data_year, l."period"
FROM peak p JOIN trough t ON p.place_id=t.place_id JOIN latest l ON l.place_id=p.place_id
ORDER BY l.place_id

SELECT place_id, place_name, MIN(data_year), MAX(data_year), COUNT(*) FROM housing.house_price_index
WHERE level='MSA' AND hpi_type='traditional' AND hpi_flavor='purchase-only' AND frequency='quarterly'
AND place_id IN ('33700','32900')
GROUP BY place_id, place_name