Twenty-five US cities with material vacant / abandoned housing stock, ranked by percentage population loss from historical peak. Rust Belt manufacturing collapses, Northeast legacy cities, and Deep South distress. Each row includes documented vacant-unit counts, the responsible land bank or blight authority, the dominant cause, and the Institute's practitioner prognosis. Where the empty houses actually are, who owns them now, and what the recovery trajectory looks like block by block.
The vacancy problem in American cities is heavily documented but poorly aggregated. Every city with a serious vacancy issue has its own land bank, its own municipal blight authority, its own academic partners, and its own newspaper-of-record covering the story. What no single source produces is a cross-city comparison table with population trajectory, vacancy scale, dominant cause, and land-bank contact all in one place. This page is that table.
The Institute's use case: a starting point for practitioners doing serious work in this space — investors evaluating opportunistic entry, family offices building distressed real-estate positions, journalists working comparative pieces, urban-planning graduate students building their thesis comp set, and civic-technology teams whose maps need a source table. Not a property-level target list; that requires a separate build per city. This is the aggregate landscape.
Sources: US Census Bureau (2020 Decennial + ACS 5-year housing vacancy series, Table B25004); municipal and county land-bank public inventories (Detroit Land Bank Authority, Cuyahoga County Land Reutilization Corporation, Genesee County Land Bank Authority, St. Louis Land Reutilization Authority, Baltimore Vacants to Value, Shelby County Land Bank, and others); academic research (Case Western Reserve Center for Community Solutions, Johns Hopkins Urban Health Institute, Youngstown State Center for Working-Class Studies, University of Memphis blight-mapping research); local press of record for each city.
What the Institute adds. We save the reader the time to aggregate the fragmentary sources — land-bank inventories, Census vacancy tables, municipal blight registries, and academic research all live in different formats behind different portals — and we bring an original point of view once the record is on the table. Thirty-one years as a CPA, plus decades as a controller, CFO, and family-office operator, sit behind the practitioner read of what the numbers mean for opportunistic entry, long-term recovery trajectory, and neighborhood-level dynamics. Aggregation is the floor; the practitioner interpretation is the reason to read us.
What the reader should know: Vacant-unit counts are estimates, not audited disclosures. Municipal and county land banks publish inventory data with varying frequency and definitional consistency. The Institute uses the most authoritative available figures per city and cites the source. Where a range is reported, the lower bound is shown.
| # | City | State | Peak Pop. | Peak Yr | Current Pop. | % Loss | Est. Vacant / Abandoned | Vacancy Rate | Dominant Cause | Land Bank / Blight Authority | Prognosis |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | St. Louis | MO | 856,796 | 1950 | 279,000 | 67% | ~11,000+ | ~22% | Manufacturing collapse; white flight; municipal fragmentation (90 munis in county) | Land Reutilization Authority (LRA) — St. Louis; ~11,000+ properties in inventory | Downtown recovery slow; Old North and Ville attracting investors block-by-block |
| 2 | Detroit | MI | 1,849,568 | 1950 | 633,000 | 66% | ~24,000 structures + ~90,000 lots | ~24% | Auto industry collapse; 2013 municipal bankruptcy; population departure | Detroit Land Bank Authority (DLBA); largest US land bank; ~65,000 parcels | Aggressive demolition program (2014-present); DLBA auctions active; downtown and midtown recovering |
| 3 | Youngstown | OH | 168,330 | 1930 | 59,000 | 65% | ~5,000+ | ~19% | Steel industry collapse (Youngstown Sheet & Tube 1977); population departure | Youngstown Neighborhood Development Corporation (YNDC); planned-shrinkage program | Managed shrinkage strategy; low investor interest outside specific pockets |
| 4 | Gary | IN | 178,320 | 1960 | 66,000 | 63% | ~10,000+ | ~28% | US Steel Gary Works decline; racial and economic segregation; municipal insolvency | City of Gary Redevelopment Commission; land bank formation attempts | Extreme distress; some proximity-to-Chicago speculation but low ground-level activity |
| 5 | Cleveland | OH | 914,808 | 1950 | 362,000 | 60% | ~20,000+ structures | ~19% | Manufacturing decline; foreclosure crisis 2007-2010; population departure | Cuyahoga Land Bank (Cuyahoga County Land Reutilization Corporation) — 2009 model | Active demolition + rehab; Slavic Village and Detroit-Shoreway rebuilding; downtown recovery |
| 6 | Flint | MI | 196,940 | 1960 | 78,000 | 60% | ~7,000+ | ~22% | GM plant closures 1980s-90s; water crisis 2014; population departure | Genesee County Land Bank Authority — earliest US model, widely studied | Land bank widely-cited model but recovery slow; specific block-scale rebuilds |
| 7 | Pittsburgh | PA | 676,806 | 1950 | 302,000 | 55% | ~5,000+ (declining) | ~13% | Steel collapse; successful reinvention via UPMC, CMU, tech ecosystem | Pittsburgh Land Bank (city-level, established 2014) | Recovering; East Liberty, Lawrenceville, Strip District gentrifying; Hill District improving |
| 8 | Buffalo | NY | 580,132 | 1950 | 275,000 | 53% | ~10,000+ | ~15% | Manufacturing decline; population departure to Sunbelt | Buffalo Erie Niagara Land Improvement Corp. (BENLIC); city demolition program | Gradual recovery; downtown and West Side improving; East Side remains distressed |
| 9 | Dayton | OH | 262,332 | 1960 | 136,000 | 48% | ~8,000+ | ~15% | GM Delco/Frigidaire closures; NCR relocation to Atlanta; opioid crisis | Montgomery County Land Reutilization Corp (MCLRC) | Wright-Patt AFB stable employer; Oregon District reviving; land bank active |
| 10 | Scranton | PA | 143,433 | 1930 | 76,000 | 47% | ~3,000+ | ~11% | Anthracite coal decline mid-20th century; slow long-term departure | Lackawanna County Land Bank Authority | Modest recovery around Marywood, University of Scranton; specific arts-district rebuilds |
| 11 | Camden | NJ | 124,555 | 1950 | 70,000 | 44% | ~3,500+ | ~15% | RCA and Campbell Soup manufacturing decline; racial segregation; extreme concentrated poverty | Camden Redevelopment Agency | Some Rutgers-Camden and waterfront revival; residential recovery limited |
| 12 | Birmingham | AL | 340,887 | 1960 | 197,000 | 42% | ~7,000+ | ~13% | Steel and industrial decline; suburban flight to Hoover, Vestavia | Land Bank Authority of Birmingham | UAB employer anchor; downtown Loft District recovery; northside distressed |
| 13 | Baltimore | MD | 949,708 | 1950 | 577,000 | 39% | ~14,000-15,000 vacant homes | ~8% | Manufacturing decline; deindustrialization; population departure to Baltimore County | Vacants to Value (V2V) program; Baltimore Housing Authority | Persistent vacancy despite programs; Fells Point / Federal Hill / Canton recovering; West/East distressed |
| 14 | Cincinnati | OH | 503,998 | 1950 | 309,000 | 39% | ~10,000+ | ~12% | Manufacturing decline; suburban flight; racial economic segregation | Hamilton County Land Reutilization Corp (Cuyahoga model applied) | Over-the-Rhine revival dramatic; downtown-adjacent neighborhoods reviving; other neighborhoods distressed |
| 15 | Rochester | NY | 332,488 | 1950 | 208,000 | 37% | ~6,000+ | ~11% | Kodak collapse (2012 bankruptcy); Xerox decline; manufacturing departure | Rochester Land Bank Corp; city acquisition programs | University of Rochester anchor; East Avenue and downtown revitalizing; other neighborhoods distressed |
| 16 | Utica | NY | 101,740 | 1930 | 65,000 | 36% | ~2,500+ | ~13% | Manufacturing decline; long-term population loss | Utica Industrial Development Agency; land bank programs | Refugee resettlement (Bosnian, Somali, Burmese communities) partial stabilization |
| 17 | Syracuse | NY | 220,583 | 1950 | 144,000 | 35% | ~4,500+ | ~12% | Manufacturing decline; General Electric departure; population loss | Greater Syracuse Land Bank | Syracuse University and SUNY-ESF anchors; specific downtown revival |
| 18 | Akron | OH | 290,351 | 1960 | 189,000 | 35% | ~4,500+ | ~10% | Rubber-industry collapse (Goodyear, Firestone); population loss | Summit County Land Bank | University of Akron employer; Highland Square recovery; other neighborhoods distressed |
| 19 | Erie | PA | 138,440 | 1960 | 92,000 | 33% | ~2,500+ | ~10% | Manufacturing decline; GE Transportation reductions | Erie Land Bank Authority | Lake Erie tourism modest recovery; downtown improving; specific pockets distressed |
| 20 | Newark (parts) | NJ | 442,337 | 1950 | 305,000 | 31% | ~4,000+ | ~10% | Manufacturing decline; 1967 unrest; population departure | Newark Community Economic Development Corp (NCEDC) | Downtown Prudential/NJPAC anchor; Ironbound and University Heights improving; West/South distressed |
| 21 | Toledo | OH | 383,818 | 1970 | 265,000 | 31% | ~6,000+ | ~11% | Auto and glass manufacturing decline (Jeep, Owens-Illinois) | Lucas County Land Bank | Jeep plant continued operation stabilizing; specific neighborhoods reviving; land bank active |
| 22 | Jackson | MS | 202,895 | 1980 | 140,000 | 31% | ~5,000+ | ~13% | White flight to Madison County; municipal insolvency risk; water crisis | City redevelopment authorities; state receivership discussions | Distressed with limited near-term recovery signals |
| 23 | Trenton | NJ | 128,009 | 1950 | 91,000 | 29% | ~3,000+ | ~12% | Manufacturing decline; state-capital employment insufficient | City Housing Production Redevelopment Authority | State government employer stable but insufficient; specific pockets improving |
| 24 | Reading | PA | 111,171 | 1930 | 95,000 | 15% | ~2,500+ | ~10% | Historically slow long-term decline; recent Latino population growth partially offsetting | Berks County; Reading Redevelopment Authority | Latino population growth stabilizing; specific downtown improvement; industrial legacy neighborhoods distressed |
| 25 | Memphis | TN | 674,028 | 2000 | 625,000 | 7% | ~30,000+ blighted properties | ~11% | High vacancy despite modest population loss; concentrated blight in specific neighborhoods | Shelby County Land Bank; Memphis 3.0 comprehensive plan | Downtown recovery; Overton Square / Broad Ave revitalizing; Frayser / Whitehaven / Orange Mound distressed |
Ranking sorted by percentage population loss from historical peak decennial Census count. Peak year is the decennial Census year at which the city's population peaked (typically 1950 for northeastern US, 1960-70 for late-manufacturing cities, later for southern cities). Current population is Census 2020 base plus ACS 5-year post-2020 estimates, rounded to nearest thousand. Vacant/abandoned unit counts drawn from municipal land-bank inventories, blight-authority registries, and academic research; where a range is reported, the lower bound is shown. Vacancy rate is ACS 5-year housing vacancy series (includes seasonal, for-sale, and rental listings in addition to structural vacancy).
1. Population loss and vacancy count are different metrics. St. Louis has lost 67% of its peak population but has fewer documented vacant units than Memphis. Memphis has lost only 7% of its peak population but has 30,000+ documented blighted properties. Population trajectory tells you the direction. Documented vacancy count tells you the current inventory. Reading only one metric misses the picture.
2. The land-bank tool matters. The Genesee County Land Bank (Flint, 2004) was the first US model. Cuyahoga County (Cleveland, 2009) codified the county-level land-reutilization-corporation structure that became the widely-adopted template. Detroit Land Bank Authority is the largest US land bank at ~65,000 parcels. St. Louis Land Reutilization Authority holds ~11,000+ properties. Baltimore's Vacants to Value program is the model municipal-blight registry. The specific institutional infrastructure differs city by city and shapes what an outside investor can do — auction cadence, buyer qualification requirements, price floors, rehabilitation requirements.
3. The specific cause matters for the trajectory. Cities whose decline traced to a single dominant employer collapse (Kodak in Rochester, US Steel in Gary and Youngstown, Goodyear and Firestone in Akron, auto in Detroit and Flint) have different recovery trajectories than cities whose decline traced to broad-based deindustrialization (Baltimore, Cleveland, Buffalo). The single-employer-collapse cities are more concentrated in impact and slower in recovery unless a new anchor employer emerges (UPMC and CMU rebuilding Pittsburgh, Wright-Patt AFB stabilizing Dayton).
4. University and hospital anchors are the specific stabilizing forces. Rochester (University of Rochester + Strong hospital system), Cleveland (Cleveland Clinic + Case Western), Pittsburgh (UPMC + CMU + Pitt), Baltimore (Johns Hopkins + University of Maryland), Syracuse (Syracuse University + SUNY-ESF), Birmingham (UAB) all have institutional employer anchors that provide a floor under the local economy. Cities without such anchors (Gary, Camden, Jackson, parts of Youngstown) have less to slow the decline.
5. Population loss can plateau or reverse block-by-block before it plateaus city-wide. Detroit's downtown, Cleveland's Slavic Village and Detroit-Shoreway, Cincinnati's Over-the-Rhine, St. Louis's Old North and Ville, Baltimore's Fells Point and Canton, Newark's Ironbound — specific neighborhoods within distressed cities can be actively appreciating even as the citywide population continues to decline. The investable opportunity is at the neighborhood level, not the citywide level. A block-by-block micro-analysis is the required due-diligence discipline.
6. Reading and Memphis are structural outliers on this list. Reading (15% loss from peak) and Memphis (7% loss) do not fit the general Rust Belt pattern. Reading has been stabilized by material Latino population growth in the last two decades and has the smallest peak-to-current gap on this list. Memphis has never suffered a peak-to-trough population collapse comparable to the northern cities, but has instead accumulated concentrated blight in specific neighborhoods (Frayser, Whitehaven, Orange Mound) even as citywide population held roughly flat. Both cases underscore that the vacancy problem is not exclusively a Rust Belt story.
The Baratelli Institute does not produce original vacancy counts or population projections. The US Census Bureau, municipal land banks, county blight authorities, and academic research groups do that work. We aggregate what those sources publish into a single reference destination and bring a practitioner's read to what the numbers mean — for opportunistic real-estate investing, for civic finance and municipal-bond exposure, for local-economy tracking, and for the broader Institute work on gentrifying small towns.
Every number on this page traces to a filed or published source. Where we discuss land-bank inventories, we source from the land bank's own public reports. Where we discuss Census population and vacancy data, we cite the specific Census / ACS tabulation. Where we discuss academic research, we credit the specific research group. Where we discuss neighborhood trajectory, we source from local press of record and municipal planning documents.
The reader who wants a specialist's read on any single city should start here, then read the primary sources and the local press coverage. We are a starting point that saves the reader the time to aggregate the fragmentary sources — and a practitioner voice once the record is on the table.