Current · housing · macro · cultural
US House Shortage
The homes the country needs to build, and the homes it cannot afford.
Momentum
· Steady
+1 pts belief · 90 days · as of Sep 26, 2026
Belief
22 / 100
skeptic-dominated · as of Sep 26, 2026
Maturity
Consensus
Curated · revised May 23, 2026
Numen reads this Current
The American housing affordability collapse of 2020–2023 has plateaued at a painful level. Median home prices roughly 4–10x median income across major metros — three times the historical norm of 3x. The underlying mismatch is structural: two decades of under-building, a tax code that subsidizes incumbents, zoning regimes that block density, and a labor force trained for the volume of homes the country built a generation ago.
Unlike inflation, this shortage cannot be fixed by raising rates. Higher mortgage rates make existing homeowners refuse to sell — locking them into 3% notes — and make new construction less attractive, so supply tightens further. The window for this Current to close is measured in years, possibly a decade. The arc holds high across the entire planning horizon.
For stewards: this changes labor markets (workers cannot move for jobs), capital allocation (housing absorbs disposable income that would otherwise reach consumption), and politics (housing scarcity radicalizes the under-35 voter). It is not a temporary distortion — it is the operating environment.
Believers
Case-Shiller home prices, 12-month change
Weight 4/5 · 12-month change
FHFA house prices, 12-month change
Weight 3/5 · 12-month change
Median new-home price, 12-month change
Weight 2/5 · 12-month change
Skeptics
Months' supply of new homes
Weight 4/5
Building permits, 12-month change
Weight 3/5 · 12-month change
Housing starts, 12-month change
Weight 3/5 · 12-month change
Methodology
How the three reads are made
Belief is the balance of live evidence. Every believer and skeptic signal on this page is ranked against its own history (Custom Indices are read on their 0–100 scale; a few are read as their 12-month change, or with a low reading supporting their side, as marked). Each side is weight-averaged and pulled gently toward neutral so thin evidence can't pin the read; Belief is the believer side's share, from 0 to 1.
A signal counts only when it is fresh by its own publishing cadence. If less than 60% of the weight is fresh, or one side has no fresh evidence, Belief is withheld for the day rather than carried forward.
Momentum is Belief today minus Belief 90 days ago, computed with the same evidence as it stood then. +3 points or more reads as accelerating, −3 or less as fading, anything between as steady.
Maturity is where the motion sits on the adoption curve. It is an editorial judgement, not a computation, and it shows the date it was last revised.
Evidence and context
Items marked with a weight are scored. Items marked Context · not scored are the reporting and history that frame the Current; they inform the essay, not the numbers.
Belief and Momentum are also published as signals (current.us-house-shortage.belief, current.us-house-shortage.momentum) and can be charted, correlated and woven like any other.
Council Inquiry™
Book a 30-min with a Council researcher.
The synchronous analyst-call analog. Bring one question; a named researcher walks the lattice with you for thirty minutes. Stewards (Enterprise) include this in every tier; Creators + Researchers book on a per-session basis.