Methodology

How the Underbuilt vs. Overpriced Atlas scores housing pressure across the 75 largest U.S. metropolitan areas.

What this tool does

The atlas is a screening tool for housing pressure. It does not model rents or predict prices. Instead, it flags where two forces — weak new construction and high rent benchmarks relative to income — appear to overlap.

A high score means the inputs suggest pressure. It does not prove a housing crisis or assign blame. Rankings should be investigated, not cited as conclusions.

Data sources and vintages

Source	Vintage	Coverage
Census Building Permits Survey (BPS)	2024 annual	Permitted housing units by CBSA and county
American Community Survey (ACS 1-year)	2024	Metro-level median income, rent, tenure, cost burden
American Community Survey (ACS 5-year)	2024	County-level income and cost burden
HUD Fair Market Rents (FMR)	FY2026	County-level 2-bedroom rent benchmarks
HUD Small Area FMRs (SAFMR)	FY2026	ZIP-level 2-bedroom rent benchmarks
Census delineation	2023 (List 1)	County-to-CBSA membership crosswalk
Census ZCTA boundaries	2020 generalized	ZIP geometry for detail maps
Census county boundaries	2023 generalized	County outlines for detail maps

The meta.json file shipped with every build records the exact vintages used. The source_catalog.json file contains direct download URLs for each input.

Score construction

Underbuilt component (0–100)

Measures how thin new housing construction is relative to the metro's existing housing stock.

Input:

permits_per_1000_households = (annual_permitted_units / total_households) × 1,000

Scoring: Inverse percentile rank across all 75 metros. Metros with fewer permits per household score higher.

A score of 90 means only 10% of the metros in the panel had weaker permitting intensity.

Overpriced component (0–100)

Measures how expensive benchmark rents are relative to local incomes.

Inputs:

FMR-to-income ratio = (2BR Fair Market Rent × 12) / median household income
Renter cost-burden share = percentage of renters paying 35%+ of income on housing

Scoring:

overpriced = mean(percentile(fmr_to_income), percentile(rent_burden_35plus)) × 100

A score of 80 means the metro is in the top quintile on the combined rent-to-income and burden measures.

Overall score (0–100)

overall = mean(underbuilt, overpriced)

Score range	Label
75–100	Acute pressure
55–74	Elevated
35–54	Watch
0–34	Lower pressure

ZIP-level detail

For each of the 75 metros, the atlas shows ZIP-level Small Area Fair Market Rents (SAFMRs) mapped onto Census ZCTA boundaries. SAFMRs are HUD's ZIP-level rent benchmarks, calculated from local rent distributions. They are used in the Housing Choice Voucher program to set payment standards that vary by neighborhood.

County boundaries are overlaid for geographic context, with county-level permits, income, and cost-burden statistics shown alongside the map.

Limitations

Single-year snapshot. There is no time series — you cannot see whether a metro is getting worse or improving. A trailing 3-year permitting average would smooth annual volatility and strengthen the underbuilt signal.
Permits are a noisy proxy for supply. Metros with large multifamily pipelines can look well-supplied on permits while still having affordability problems, because permitted units are not necessarily affordable units.
FMR is a lagging indicator. Fair Market Rents are backward-looking and tuned for the voucher program, not real-time market rents. Metros with fast rent growth may appear better than they are.
No vacancy, household growth, or structure mix. These factors shape housing pressure but are not yet included. Their absence means the screen can misread metros with unusual housing stock characteristics.

Caveats

Permits are not completions. The BPS counts authorized housing units. Not all permitted units get built, and there is a lag between authorization and occupancy.
FMR is a program benchmark, not observed market rent. HUD calculates Fair Market Rents for the Housing Choice Voucher program. They approximate the 40th percentile of gross rents for recent movers, but they are not the same as asking rents on listing sites.
ACS products differ. Metro-level figures use the ACS 1-year product (available only for areas with 65,000+ population). County-level figures use the ACS 5-year product (broader coverage, more stable, but slower to reflect recent changes). Comparing across these products requires care.
Metro definitions vary. Census CBSAs and HUD Metro FMR Areas do not always share the same boundaries. Some large metros are split into sub-areas by HUD.
This is a screen, not a causal model. A high score flags a pattern worth investigating. It does not establish that housing costs are rising because of insufficient construction, or vice versa.

Why this design

Every component is easy to explain in plain language.
No black-box regression or machine learning is required.
Every input is downloadable as CSV so anyone can check the math.
The score separates supply-side (underbuilt) from demand-side (overpriced) so users can see which dimension is driving a metro's ranking.

Possible future enhancements

Trailing 3-year permitting average to smooth annual volatility.
Compare permits to household growth, not just household stock.
Add vacancy rate and structure mix terms.
Sensitivity analysis by metro size bucket.
Confidence intervals and missingness flags.