Permits Pop, Single-Family Slips: April’s Housing Starts Lean on Backlog, Not New Demand

In the official press release dated June 16, 2026, the headline claims a broad April rebound in residential construction. But the body of the document reads “FOR RELEASE AT 8:30 AM EDT, THURSDAY, MAY 21, 2026” and presents April 2026 data with “Next Release: June 16, 2026.” We flag the date mismatch up front because it matters for record-keeping and comparisons. Our analysis below references the April 2026 figures contained in the body text and the May 21, 2026 release timing, while acknowledging the June 16, 2026 header date.

Here’s what the data actually shows:

• Total permits rose to 1,442,000 SAAR in April, up 5.8% m/m, but single-family authorizations fell 2.6% m/m to 872,000; the rebound is overwhelmingly multi-family (5+) with 514,000 permits and Table 1a showing +22.7% m/m and +11.5% y/y.

• Starts printed 1,465,000 SAAR, down 2.8% m/m (±11.0%) and up 4.6% y/y (±13.9%)—both changes are statistically indistinguishable from zero. The one move that does clear the significance bar: single-family starts fell 9.0% m/m (±7.5%) to 930,000.

• Completions rose to 1,449,000 SAAR, +4.8% m/m (±18.2%) and -2.0% y/y (±10.7%)—again, not statistically significant. Single-family completions dipped 1.0% m/m (±10.4%) to 903,000.

• Starts exceeded permits (1,465,000 vs 1,442,000 SAAR), implying a draw on the backlog rather than fresh permit momentum. The not-seasonally-adjusted “authorized but not started” pool remains elevated at 260.6k units (April) vs 265.7k (March).

• Seasonally adjusted back-revisions are extensive: permits revised back to January 2020 and all activity series (authorized but not started, started, under construction, completed) revised back to January 2021. March 2026 comparators were updated to starts 1,507,000, permits 1,363,000, completions 1,382,000, but the release offers no consolidated view of how the historical trendlines changed.

Note: “—” indicates not cited with a confidence interval in the release.

Permits Up, Single-Family Down: The Composition Problem

A 5.8% jump in total permits makes a clean headline. The composition tells a different story. Single-family authorizations slid 2.6% m/m to 872,000 and are down 5.5% y/y (Table 1a). The entire “up” narrative rests on multi-family: 5+ permits soared 22.7% m/m to 514,000 and 11.5% y/y. Year-to-date (NSA), the split is just as stark: total permits -1.6%, single-family -6.8%, 5+ +12.1% (Table 1b).

That isn’t just composition—it’s the market speaking. With mortgage affordability still tight, the demand-side pressure pivots toward rentals and dense supply, while single-family intent softens.

Starts Outrun Permits—for the Wrong Reason

Starts at 1,465,000 SAAR outpaced permits (1,442,000). In a healthy expansion, permits lead starts. Here, starts are leaning on pipeline conversion. The not-seasonally-adjusted backlog remains high at 260.6k units (April) vs 265.7k (March). Seasonally adjusted signals point to easing—Table 2a shows the total backlog -1.9% m/m (with wide ±4.4%), including 1-unit -2.1% and 5+ -5.3%—but from elevated levels.

The statistically significant move here is the wrong one for growth momentum: single-family starts fell 9.0% m/m (±7.5%) to 930,000. Multi-family starts held at 529,000, sustaining activity even as the single-family engine misfires.

Significance Is the Story: Confidence Bands Turn Headlines to Haze

Several marquee changes are not statistically significant:

• Total starts: -2.8% m/m (±11.0%), +4.6% y/y (±13.9%)—both include zero.

• Total completions: +4.8% m/m (±18.2%), -2.0% y/y (±10.7%)—also include zero.

In other words, the “decline in starts” and “rise in completions” aren’t clear signals—they’re noise within wide sampling bands. The one result that does clear the bar is the -9.0% m/m drop in single-family starts, which appears without the asterisk in the tables, implying statistical significance. If you’re trading the headline, you’re trading the error bars.

Regions Tell a Different Tale

National aggregates hide where the ground is actually shifting (YTD NSA, Table 1b):

• South: the heavyweight region is weak—total -6.7%, 1-unit -6.8%. That’s a broad-based drag.

• Northeast: total +14.2%, but 1-unit -11.7%—apartment-led strength, not a single-family revival.

• Midwest: total +7.3%, 1-unit -0.5%—modest resilience with mixed composition.

• West: total +0.7%, 1-unit -9.1%—again, multi-family doing the heavy lifting.

For builders and suppliers geared to single-family (especially in the South and West), this is a headwind masked by national totals.

Revisions: Moving Targets Masquerading as Trends

The May 21, 2026 body text announces sweeping methodology updates: seasonally adjusted permits revised back to January 2020, and all seasonally adjusted activity series (authorized but not started, started, under construction, completed) revised back to January 2021. The only explicit anchors are the revised March 2026 levels—starts 1,507,000, permits 1,363,000, completions 1,382,000—but we’re not told how the rest of the history shifted.

That matters. Stretch back to the already-cooling trajectory:

• August 2025 (release dated Sept 17, 2025): starts 1,307,000 SAAR (-8.5% m/m); single-family authorizations 856,000; 5+ permits 403,000—multi-family resilience was visible then.

• 2025 completions were 1,497,800, 7.9% below 2024’s 1,626,900 (release dated Feb 18, 2026). December 2025 permits printed 1,448,000, 2.2% below December 2024—a slowing supply handoff into 2026.

April 2026 extends that through-line: single-family permits -5.5% y/y, single-family starts -9.0% m/m, and total completions still -2.0% y/y. Revisions may tweak the contours, but the melody—single-family softness, multi-family resilience—hasn’t changed.

The Investor Takeaway

• Single-family weakness is real; the rest is mostly sampling noise. The statistically significant decline in single-family starts (-9.0% m/m) contrasts with non-significant total moves. Treat total starts and completions headlines with caution.

• Composition, not magnitude, is the April story. Multi-family permits (+22.7% m/m; +11.5% y/y) carried the permits rebound, while single-family authorization fell 2.6% m/m and 5.5% y/y.

• Starts > permits = backlog drawdown. With 260.6k units still authorized but not started (NSA) and SA backlogs easing y/y (total -5.6%, 5+ -18.2%), the pipeline can support starts near-term—but it’s not a fresh-demand signal.

• The South—the largest region—is soft YTD (-6.7% total permits; -6.8% 1-unit), undercutting national aggregates. The Northeast’s strength is apartment-heavy (total +14.2%; 1-unit -11.7%).

Positioning Ideas

• Homebuilders: Prefer operators with pricing power via rate buydowns, quick spec turns, and entry-level exposure that can flex incentives. Be selective on land-heavy, move-up–centric builders in the South and West where 1-unit permits are down 6.8% and 9.1% YTD, respectively.

• Building products: Tilt toward categories with multi-family leverage (commercial-grade HVAC, fire protection, elevators, select finishes) and away from single-family–centric SKUs (premium fixtures, discretionary remodel) until rate relief improves absorption.

• Multi-family REITs and developers: Near-term construction flow is supported by 5+ permits at 514,000 and starts at 529,000, but note the -18.2% y/y decline in the SA backlog—2027–2028 deliveries could tighten, favoring stabilized, high-occupancy portfolios over development-heavy stories.

• Regional banks and CRE lenders: Monitor construction and development loan performance in the South where permitting is weakest. Underwriting standards should reflect widening regional dispersion and a cooling single-family pipeline.

• Rates and macro: With total starts/completions moves not statistically significant, April doesn’t deliver a growth shock. Housing remains a modest drag until single-family stabilizes—incrementally supportive of a slower Fed trajectory rather than a hawkish pivot.

What to Watch Next

• Mortgage rates and affordability metrics (payment-to-income, buy-down adoption rates).

• Permits-to-starts ratio: a sustained gap with starts > permits confirms pipeline drawdown over new intent.

• Multi-family absorption, lease-up velocity, and vacancy; whether 5+ permits stay elevated or revert.

• NAHB/Wells Fargo HMI for signs of single-family sentiment turning.

The headline says permits are up. The numbers say the single-family core is cooling, starts are riding the backlog, and the South—the market’s center of gravity—is soft. Until affordability improves enough to lift single-family permits and reverse that -9.0% m/m hit to starts, treat total rebounds as composition effects, not a new housing cycle.