I publish the data behind everything on this site. The datasets below are free to download and free to reuse with attribution. The three primary datasets are archived with permanent DOIs, so they can be cited in academic and policy work and the links will not rot. No signup, no email, no license fee.
The datasets
| Dataset | What it covers | Formats | License | DOI |
|---|---|---|---|---|
| Foothold Index | 83 US metros screened for where an owner occupant of a two to four unit home can buy and let the tenant rent carry part of the payment; 11 are ranked | CSV, JSON | CC BY 4.0 | 10.5281/zenodo.22019647 |
| Down Payment Assistance Survey | 83 US metros checked for down payment assistance an owner occupant can use on a two to four unit purchase | CSV, JSON | CC BY 4.0 | 10.5281/zenodo.22019630 |
| Pittsburgh Short-Term Rental Census | Listing-level count of short-term rentals in Lawrenceville and Bloomfield, 273 live Airbnb listings plus the mid-term furnished market, August 2026 | CSV, JSON | CC BY 4.0 | 10.5281/zenodo.22258181 |
| Foothold rate sensitivity sweep | The Foothold Index re-run at 13 mortgage rates from 5.00% to 7.50%: dollars kept per month for 82 of the 83 metros at each rate (Scranton has no surviving listing) (supplement to the Foothold Index, September 2026) | CSV (the file and the rate page are both the 3 September run, Buffalo row included; the later Buffalo re-run is on the Buffalo page) | CC BY 4.0 | Supplement to 10.5281/zenodo.22019647 |
How the Foothold Index screen works, step by step, is published as a citable methods note: 10.5281/zenodo.23070347.
Foothold Index
A screen of two to four unit listings across 83 US metro areas, built to answer one question: where can a cash limited first time buyer purchase a small multifamily home, live in one unit, and have the tenant rent cover enough of the payment that the purchase holds up against local incomes and rents. Data as of 6 August 2026.
Each row carries the entry price, the typical price, the surviving listing count, the monthly amount kept versus renting, and the ranked Foothold score. The screening method, including what the model cannot tell you, is written out in full on the Foothold Index page.
Download the CSV or download the JSON.
Suggested citation: VanToVault, Foothold Index: rent versus own screen for two to four unit homes across 83 US metros, 2026. DOI 10.5281/zenodo.22019647
Mirrors: GitHub, Hugging Face, Kaggle.
Down Payment Assistance Survey
A hand checked survey of down payment assistance programs across the same 83 US metro areas, limited to assistance an owner occupant can actually use on a two to four unit purchase. This is narrower than it sounds. Plenty of programs exist for single family homes and quietly exclude anything with a second unit, which is the detail that sinks a lot of first purchases.
Each row carries the program name, the advertised maximum assistance, and a link to the source I read, which is the agency itself in all but five metros (Chicago, Portland ME, Salt Lake City, Tulsa and Washington DC, where it is a third-party summary).
As last checked on 4 October 2026 (survey v4.3), a usable program was found in 70 metros and none in 13, five of which could not be confirmed either way.[S1] Unconfirmed metros are recorded as unknown rather than as a no, because absence of evidence is not the same as a program not existing.
Download the CSV or download the JSON.
Suggested citation: VanToVault, Down payment assistance for two to four unit owner occupied purchase, 83 US metros, 2026. DOI 10.5281/zenodo.22019630
Mirrors: GitHub, Hugging Face, Kaggle.
Pittsburgh Short-Term Rental Census (Lawrenceville and Bloomfield)
A hand-checked, listing-level census of 273 live Airbnb listings and 42 mid-term furnished units across four Pittsburgh neighborhoods, snapshot August 2026: counts by neighborhood and type, density against the housing stock, three observed price windows, and host concentration. Nothing is modeled, and every row carries its listing URL. Methodology and summary tables are on the census page. Archived at Zenodo under concept DOI 10.5281/zenodo.22258181, which always resolves to the latest version (version 1.0 is 10.5281/zenodo.22258182).
Download the Airbnb census CSV, the mid-term furnished CSV, or the metadata JSON.
Suggested citation: VanToVault, Pittsburgh Short-Term Rental Census: Lawrenceville and Bloomfield, August 2026. DOI 10.5281/zenodo.22258181
Mirrors: GitHub, Hugging Face, Kaggle.
Two-to-Four-Unit Housing Stock by Metro (Census ACS)
How many homes in each of the same 83 metros sit in two-, three- and four-unit buildings, and how many of those are owner-occupied, from the U.S. Census Bureau’s American Community Survey 5-year estimates for 2020-2024 (tables B25024 and B25032), with the principal city alongside each metro. Every percentage is computed from the raw counts and rounded once. Revised 27 September 2026: percentages now carry one decimal computed from the counts, and San Antonio and Greensboro swap ranks 59 and 60; no count changed. The README lists every column, the Census API request for each row, and the caveats.
Download the housing stock CSV (83 rows) and the README. License CC BY 4.0.
Release calendar
Each dataset is published as a dated edition, so a number you cite stays attached to the edition it came from. The rate-dependent pages (the metro pages and the income and price pages) are re-run quarterly at the Freddie Mac average published on the first Thursday of the month; the datasets and their rankings only change at a new edition, which gets its own DOI version.
| Release | What changes | Planned |
|---|---|---|
| Rate edition, December 2026 | Metro and price pages re-run at the PMMS rate of December 3, 2026. Rankings unchanged. | Week of December 7, 2026 |
| Rate edition, March 2027 | Same pages, PMMS rate of March 4, 2027 | Week of March 8, 2027 |
| Rate edition, June 2027 | Same pages, PMMS rate of June 3, 2027 | Week of June 7, 2027 |
| Foothold Index, 2027 edition | Fresh listing screen across all 83 metros, new rankings, new DOI version under 10.5281/zenodo.22019647 | August 2027 |
| Down Payment Assistance Survey, 2027 edition | Every program re-checked against its current rules, new DOI version under 10.5281/zenodo.22019630 | August 2027 |
| Pittsburgh Short-Term Rental Census, update | Re-count of the same four neighborhoods once the city’s short-term rental rules take effect | Tied to the ordinance, date to follow |
Corrections between editions are logged on the corrections page and do not change an edition’s DOI. If you need a cut of the data on a different schedule, ask through the press and data page.
How you can use this
Every dataset on this page is released under a Creative Commons Attribution 4.0 license. That means you can copy it, republish it, chart it, build on it, and use it commercially, as long as you credit Van to Vault and say if you changed anything.
If you are a librarian, a journalist, a researcher, or a housing counselor and you want to link something, this page is the stable address for the data. The DOI links are the right citation for anything that needs to stay valid for years.
Suggested citation. VanToVault (2026). Foothold Index [dataset], August 2026 edition. CC BY 4.0. https://doi.org/10.5281/zenodo.22019647. Use the same pattern for the other datasets with the DOI from the table above; for a file without its own DOI, cite this page and the date in the file name.
Data licensing and press. The public files stay free under CC BY 4.0. For a commercial license, a custom cut of the data, or a press question, email vantovault@gmail.com.
Corrections
Program rules change and listing data goes stale. If you find a row that is wrong, especially a program you administer that I have recorded incorrectly, tell me and I will fix it and note the change. Accuracy matters more to me than being first.
Why I publish it
Most writing about the affordability crisis lands on one of two poles. One says owning is out of reach forever. The other says it is easy if you just believe. I think the useful ground is in between, and the only way to hold that ground is to publish the listing level data and let people check my work. If the numbers say something I did not expect, that goes in too.
See the ranking these numbers produce.
Writing about this data? The press and data page carries citable figures, the cite-as forms and a contact for custom cuts.
Neighbourhood-level companion: what house hacking costs in Lawrenceville and Bloomfield, built from county sale records and 2026 tax rates.
Get started with the newsletter!
Useful, relevant analysis, tools, tricks, and what you need to know on the journey to financial freedom. From the van years to a duplex, the plain math and none of the hype.
Subscribe and get the free $0-to-First-Property Roadmap to start.
No spam. Unsubscribe in one click.
Want the deeper toolkit? The guides and spreadsheets are in the shop, from $5, and the free tools stay free either way.
Sources
- [S1] Van to Vault, read 24 September 2026: “Albuquerque … 2026-09; Phoenix … unconfirmed … 2026-09; Tucson … unconfirmed … 2026-09” vantovault.com.
If you use Google, you can pick Van to Vault as a preferred source. Google can then mark this site as preferred in your Top Stories and, where it shows them, in AI Mode and AI Overviews.
Add Van to Vault as a preferred source on Google