When you spend months building an independent model to explain housing affordability, and then one of the twelve Federal Reserve Banks publishes research that confirms your core mechanics from a completely different direction, you pay attention.
That's what happened with the Seattle Housing Affordability Model (SHAM) and two recent papers from the San Francisco Federal Reserve — one on housing affordability and income dynamics (Louie, Mondragon, Najjar, and Wieland, February 2026), and a broader working paper on supply constraints and housing markets (Louie, Mondragon, and Wieland, with a July 2025 FAQ document).
These papers didn't set out to validate SHAM. They don't know SHAM exists. But their findings line up with what SHAM has been showing for 35 years of Seattle data in ways that are worth examining.
What SHAM Does
SHAM is straightforward. It calculates the percentage of gross household income that would go to housing costs — principal, interest, taxes, and insurance — if a median-income buyer purchased a median-priced home at prevailing mortgage rates in a given month.
I've computed this for every year from 1990 to 2025 across multiple markets, using Case-Shiller home price indices, Freddie Mac PMMS mortgage rates, Census SAIPE household income data, NAR buyer survey down payment data, and market-specific property tax and insurance rates from Tax Foundation and Bankrate.
Over 35 years, a consistent pattern emerged. SHAM values for median buyers cluster in a 32–52% channel regardless of which market you're looking at. Below 32%, homes tend to be underpriced relative to income, and appreciation follows. Above 52%, the market is under stress, and corrections tend to materialize. The channel held through the S&L aftermath, the dot-com era, the housing bubble, the crash, the recovery, COVID, and the 2022 rate shock.
When I expanded SHAM to four markets — Detroit, Dallas, Seattle, and San Francisco — the price ranges were dramatically different (a 6.5x spread from Detroit to SF), but the actual SHAM values clustered within a much narrower band. The mechanics were the same. The thresholds differed by market "altitude," but the physics didn't change.
What the SF Fed Found
Paper 1: Income Drives Everything
The February 2026 SF Fed Economic Letter ("Housing Affordability and Housing Demand") examined the relationship between income growth and house price growth across U.S. metro areas. Their central finding was striking: house prices tracked average (per capita) income almost perfectly — not median income, but the income measure that reflects the top of the distribution.
Before 2000, median household income and per capita income moved roughly in lockstep, so it didn't matter much which measure you used. After 2000, they diverged. Per capita income — pulled upward by high earners in tech, finance, and professional services — accelerated away from median household income. House prices followed per capita income, not median.
The SF Fed's Explanation
Paper 2: Supply Constraints Don't Explain Price Differences
The broader working paper ("Supply Constraints Do Not Explain House Price and Quantity Growth Across U.S. Cities") challenged the conventional wisdom that expensive cities are expensive because of restrictive zoning and limited supply. Their finding: empirical measures of housing supply constraints — regulations, land availability, building costs — don't explain the variation in price growth across metro areas. Instead, income growth does.
The FAQ document (July 2025) addresses the obvious objection directly. Question 11 asks: "Are you saying that San Francisco is not more expensive than Houston?" Their answer: SF's real house price growth was about 2.4% annually from 2000–2020 compared to Houston's 1%. But SF's real per capita income growth was 2.2% (99th percentile nationally) versus Houston's 0.83%. In both cities, house price growth was just a bit more than per capita income growth — exactly what you'd expect if supply elasticities were similar.
Where SHAM and the SF Fed Converge
Income Is the Denominator That Matters
SHAM was built around the principle that affordability is a ratio — price relative to income. Not price alone. Not rate alone. Not sentiment. The ratio.
The SF Fed formalized the same insight from the demand side. Their model shows that across cities, house prices are essentially a function of income growth. When SHAM shows that Detroit at $245K and San Francisco at $1.35M produce SHAM values that are closer than you'd expect (roughly a 25-percentage-point range despite a 6.5x price range), it's demonstrating the same phenomenon: income growth absorbs price growth, and markets equilibrate around what buyers can actually pay.
Which Income Measure Matters — and Why
The SF Fed's finding that house prices track per capita income rather than median household income maps directly onto a feature SHAM built independently: the dual-income toggle.
After the SF Fed published their February 2026 Economic Letter, I added a per capita income series to the SHAM engine. In King County (Seattle MSA), the gap between per capita and median SHAM narrowed from +21.70 percentage points in 1990 to +0.18 percentage points in 2023. Per capita income and median household income converged.
For the first time in our data, per capita income in King County now exceeds median household income. The SF Fed would call this "income growth at the top outpacing the middle." I call it the K-shaped recovery. Same observation, different language.
Markets Follow the Same Mechanics at Different Altitudes
SHAM's multi-market analysis found that SF, Seattle, Dallas, and Detroit all follow the same affordability mechanics — the channel just operates at a different level depending on the market. I call this the Altitude Analogy: water boils at 212°F at sea level and 202°F in Denver, but it's the same physics.
The SF Fed reached the same structural conclusion. Their FAQ explicitly states that SF's house price growth relative to its income growth looks essentially the same as Houston's. The "expensive" market isn't anomalous — it just has higher incomes driving higher prices through the same mechanism. Different methodology, same conclusion: the mechanics are universal.
The Residual Income Equalization
SHAM's "Hidden Hand" analysis found that buyers in Dallas ($133K income) and San Francisco ($240K income) end up with nearly identical residual income after housing and fixed costs — roughly $81,500. Despite earning $107,000 more, the SF buyer's additional income gets absorbed by higher taxes and higher housing costs.
This is consistent with the Rosen-Roback spatial equilibrium model that the SF Fed's work builds upon. If labor is mobile and markets are competitive, housing costs will adjust to absorb available household capacity. Both arrive at the same equilibrium: competitive markets price housing to what buyers can bear.
Where They Differ — and Why That Matters
The differences are as important as the convergences, because they reveal complementary strengths.
Macro vs. Payment Level
The SF Fed examines price-to-income ratios across dozens of metro areas. SHAM calculates actual monthly PITI including property tax rates that vary from 0.68% in SF to 1.89% in Detroit, insurance rates that vary by state, PMI tiers based on actual down payment distributions, and mortgage rate environments that change monthly. The SF Fed tells you the big picture. SHAM tells you what it costs on the first of the month.
Financing Conditions
A core SHAM insight is the Exotic Product Overlay (EPO): during 2004–2007, exotic mortgage products — interest-only loans, negative amortization, temporary buydowns — masked actual affordability stress. California absorbed 60% of the $500 billion in Option ARMs originated during that period. The SF Fed's framework captures why prices rose (income growth), but SHAM captures why prices were able to rise beyond what income fundamentals would normally support.
Practitioners vs. Economists
A realtor sitting across from a first-time buyer doesn't need to know about supply elasticity regressions. They need to know that in November 2025, the median Seattle home at $888K with a 6.24% rate and 20.3% down requires $5,354 per month — 48.3% of median income — and that historically, SHAM values above 52% have preceded corrections. That's the translation layer SHAM provides.
What This Means for Buyers, Sellers, and Advisors
If both an independent practitioner model and Federal Reserve research agree that income growth drives housing prices, and that the relationship holds across markets with very different supply conditions, then a few practical conclusions follow.
For Buyers
The question isn't whether your market is "expensive" in absolute terms. It's whether the ratio of housing costs to income in your market is at the high end or low end of its historical range. SHAM can show you that for Seattle, month by month, going back to 1990.
For Sellers and Listing Agents
Price appreciation doesn't happen because "Seattle is desirable." It happens because buyer incomes support higher prices at prevailing rates. When rates rise and incomes don't keep pace, appreciation stalls — not because people stop wanting to live here, but because the math doesn't work.
For Policy Conversations
If the SF Fed is right that supply constraints don't explain price differences across cities, then the "just build more" argument — while potentially valid for other reasons — may not be the affordability solution many assume it to be.
For Anyone Making a Housing Decision
Get the data. Look at the ratios. Compare the current moment to historical patterns. Don't rely on narratives — "prices always go up," "it's different this time," "rates will come down." The numbers tell a story that's less dramatic but more useful than any headline.
A Note on Intellectual Honesty
I want to be clear about what I'm claiming and what I'm not.
I am not claiming that SHAM predicted the SF Fed's findings. I am not claiming equivalence between a practitioner model and peer-reviewed Federal Reserve research. And I am not claiming that either SHAM or the SF Fed has answered every question about housing affordability.
What I am claiming is that when a model built from 35 years of payment-level data in the field, and research built from rigorous economic analysis at the Federal Reserve, arrive at structurally similar conclusions about how housing markets work — income drives prices, the mechanics are universal across markets, and the buyer pool (not the community median) sets the price — that convergence is meaningful. It suggests we're looking at something real.
Barry Varshay
The Mortgage Mechanic | Creator of SHAM Housing Affordability Model
Barry Varshay is a licensed mortgage loan officer (NMLS #583050) with 35 years of industry experience and a continuing education instructor for real estate professionals in Washington state. The Seattle Housing Affordability Model (SHAM) is available at seattlehousingfacts.com.
The views expressed are the author's own and do not represent CMG Home Loans, the Federal Reserve, or any government agency. SHAM is an educational tool, not financial advice.
References
Louie, S., Mondragon, J., Najjar, R., & Wieland, J. (2026). "Housing Affordability and Housing Demand." FRBSF Economic Letter 2026-03. Federal Reserve Bank of San Francisco.
Louie, S., Mondragon, J., & Wieland, J. (2025). "Supply Constraints Do Not Explain House Price and Quantity Growth Across U.S. Cities." NBER Working Paper. National Bureau of Economic Research.
Louie, S., Mondragon, J., & Wieland, J. (2025). "Frequently Asked Questions About and Comments On 'Supply Constraints Do Not Explain House Price and Quantity Growth Across U.S. Cities.'" July 2025.
McClure, K. & Schwartz, A. (2025). "Where Is the Housing Shortage?" Housing Policy Debate, 35(1), 49–63.