Why home prices are still so high in July 2026

1. The July 2026 snapshot

As of mid-2026, national medians for new and existing homes both sit above $400,000. Harvard’s Joint Center for Housing Studies (2026 State of the Nation’s Housing) notes existing-home prices are up about 54% since 2020 and remain near five times median income — far above the roughly three-times ratio that held for much of the 1990s.

Sticker prices are only half the pain. With mortgage rates still holding above 6%, the monthly principal-and-interest payment on a median-priced home was about $3,100 in late 2025, up from roughly $1,700 in early 2020. Affording that payment requires something like a $120,000+ household income, versus about $66,000 five years earlier.

Listing prices have softened in places — Realtor.com’s May 2026 report showed the national median listing price down about 2.4% year over year to roughly $430,000, the steepest annual listing decline in years of their series — but “sellers pricing to sell” is not the same as “housing became cheap.” Payments, down payments, insurance, and property taxes still price a large share of households out.

Bottom line for July 2026: prices are no longer sprinting higher everywhere, but the level of price and payment remains historically high relative to incomes. Soft landings do not reset a 50% pandemic-era jump overnight.

2. Decompose the house: land, materials, labor

A house price is not one number. Economists usually split it into structure (the reproducible building — materials, labor, overhead, profit) and land (the location, including the right to build there). New homes add a third practical layer: soft costs and regulation (entitlements, impact fees, code-driven design, delay).

National Association of Home Builders (NAHB) cost surveys of new single-family homes show construction hard costs now dominating the builder sales price:

That table is easy to misread. It describes what builders charge for a newly built house on a finished suburban lot, not what buyers pay for an existing bungalow in San Francisco or Boston. In constrained coastal metros, land’s share of total market value is often 50–80%+ for existing stock. Nationally, research using FHFA appraisal data puts land’s share of single-family value in the high-30s percent range in the 2010s–early 2020s — and much higher in the densest, richest counties.

So the answer to “why are homes expensive?” depends on which homes:

• Expensive metros / scarce lots: mostly land and the right to build (zoning, NIMBY process, limited zoned capacity).
• New construction everywhere: materials + labor + regulation, with lots still expensive where growth is hot.
• Existing stock nationally: pandemic demand shock, low inventory, and rate lock-in layered on decades of underbuilding relative to household formation.

3. Land over thirty years: the big story

If you only track lumber futures, you miss the main long-run driver. Work by Morris Davis and coauthors (including the classic Davis–Palumbo large-city series and later FHFA/AEI land indicators) shows that residential land values have outpaced structure costs for decades in high-demand places.

• In a sample of large U.S. metros, land's average share of home value rose from about 32% in 1984 to about 50% by 2004 (Davis & Palumbo). West Coast land shares were already high by the mid-1980s; Midwest shares were still single-digit to low-teens and climbed from a low base.
• Aggregating more carefully across the national stock, land has often accounted for roughly a third to half of total housing value depending on the year and method — with a clear upward trend in superstar metros.
• FHFA research for 2012-2022 finds national land prices rose about 8% per year, faster than house prices overall, lifting land's national share from about 37% to ~40%. Growth was heavily concentrated: a handful of high-value counties (especially California) dominate the aggregate land pile.

Why does land run away? Supply of desirable locations is almost fixed once you fix commute sheds, school districts, and coastal amenity. When more households want to live in the same places — and local rules block denser redevelopment — the price adjustment happens in land rents, not in a factory that can mint more “San Francisco half-acres.”

Regulation is part of the land story even when it shows up as a building cost. NAHB regulation studies put government-imposed costs (lot development requirements, fees, code changes, delay) on the order of a quarter of the price of a typical new home in recent years — tens of thousands of dollars that are not lumber or carpenters’ wages, but still show up in the check the buyer writes.

Thirty-year takeaway: land (and the right to build on it) is the component that can double or triple without a matching physical improvement. Materials and labor cannot do that forever; they face competition and substitution. Scarce urban land under tight zoning can.

4. Building materials: up, volatile, not the whole plot

Producer prices for construction materials (BLS special index, 1982 = 100) tell a clear but limited story:

From the mid-1990s to mid-2026 that is roughly a 2.5–2.6× nominal increase in materials prices. Over the same window, overall U.S. consumer prices rose on the order of ~2×. So materials are modestly more expensive in real terms than thirty years ago — important for builders, not enough alone to explain tripling house prices.

The path mattered more than the trend. Softwood lumber’s 2020–21 blowout (triple-digit year-over-year spikes at the worst) made every framing package a news story. Many inputs partially normalized, then re-accelerated: industry trackers saw construction materials PPI up several percent in 2025 alone, with trade policy on softwood, steel, and aluminum still in the mix in 2025–26.

Materials also interact with size and code. The typical new American house is larger and more heavily specified (HVAC, insulation, electrical, accessibility) than a 1990s starter. Even flat unit prices would yield a more expensive structure. That is quality and regulation, not just “wood got expensive.”

5. Labor: scarce trades, rising pay, slower productivity

Construction wages have risen sharply in nominal terms. Average hourly earnings for all construction employees were about $41 / hour in mid-2026 (BLS). Production and nonsupervisory construction pay sits a bit lower but in the same high-thirties band. In the mid-1990s, comparable construction hourly earnings were roughly in the mid-teens — call it about 2.5× over thirty years, similar to materials and only somewhat ahead of general wage growth.

The binding constraint is often bodies and skills, not the sticker wage. After the 2008–12 bust, residential construction employment cratered; many workers left for other industries and did not return. Demographic aging, training bottlenecks, and immigration policy all hit a sector that still relies heavily on immigrant labor in many regions. When starts pick up, bid prices for framers, electricians, and plumbers rise faster than average wages because schedules slip and overtime becomes normal.

Residential construction productivity has been notoriously weak compared with manufacturing. Building a house is still a site-specific craft product. Factory-built modules and panelization help at the margin, but most U.S. single-family volume is still stick-built on site. Weak productivity means wage gains pass more fully into structure costs.

Materials and labor explain why building a new house is expensive. They do not fully explain why a 1955 ranch on a coastal lot sells for a million dollars. That is mostly land and scarcity of permission to add more homes.

6. Case study: suburban Kansas over thirty years

National averages hide geography. Suburban Kansas is a useful stress test of the land-vs-structure story because the Midwest has historically been the elastic, structure-heavy end of American housing — more room to build, lower land shares, and prices that are high by local standards but still far below the coasts.

What a typical house cost then and now

Start with the metro envelope that includes the big Kansas suburbs (Overland Park, Olathe, Shawnee, Lenexa) as well as Kansas City, Missouri. The FHFA all-transactions house price index for the Kansas City, MO-KS metro is rebased to 100 in 1995:Q1. By 2026:Q1 it was about 374 — roughly a 3.7× rise in same-property values over about thirty-one years. Wichita’s parallel index sits near 312 (~3.1×), a reminder that even within Kansas the growth rate is not uniform.

Dollar levels, not just indexes:

So a mid-1990s suburban Kansas City house that felt like a $130,000 purchase is, at a ~3.7× constant-quality path, a ~$480,000 asset today — right on top of recent Johnson County medians near $485–495k. That is not California math. It is also not free: in local income terms, and especially in monthly payment terms after the 2022 rate reset, suburban Kansas still feels much tighter than it did when 30-year mortgages printed near 8% on a six-figure house.

Quality and size matter in the dollar comparison. Many 2026 Johnson County sales are larger, better-appointed houses than the 1995 starter. The FHFA index tries to hold quality closer to constant; raw medians mix composition change with pure price change. Use the ~3–3.7× band as the honest thirty-year appreciation story, and the absolute medians as what a buyer actually writes a check for.

How that squares with land, materials, and labor

Recall the national cost curves: materials ~2.5–2.6×, construction labor ~2.5×, CPI ~2×, land much more than that in constrained metros. Suburban Kansas sits in the middle of that sandwich:

• House prices (~3–3.7×) outrun materials and labor (~2.5×) and general inflation (~2×), so structure costs alone cannot explain the full run-up — there is still a residual of land, lot improvements, fees, larger/more coded houses, and post-2020 demand/inventory effects.
• But land is not the coastal story. Davis–Palumbo found Midwest large-city land shares starting in the low teens or below in the mid-1980s and rising from a low base, versus West Coast shares already near half of home value by then. FHFA land research likewise shows national land-price growth concentrated in high-value coastal counties. Johnson County lots got more expensive as the metro grew and good school districts attracted demand — they did not become San Jose parcels.
• New construction is where materials and labor bind. A builder putting up a spec home on the Olathe or Wichita fringe faces nearly the same lumber, drywall, HVAC equipment, and trade-crew market as the rest of the country. NAHB’s structure-heavy cost stack (construction often ~60%+ of a new-home price) is more descriptive of suburban Kansas than of an infill bungalow in a coastal city. When materials and wages step up 2.5×, the floor under new KC-suburb prices steps up with them.
• Existing stock still prices off replacement cost plus land. A 1990s ranch in Overland Park is not rebuilt every year, but buyers and appraisers anchor to what a comparable new house would cost nearby. Higher structure costs therefore lift older houses even when the lot was cheap to buy in 1995.

Suburban Kansas confirms the breakdown: it is not a land-only market like coastal California, and it is not a pure materials-and-labor market either. About two-thirds of a thirty-year nominal price path lines up with structure-cost inflation; the rest is land, location, regulatory soft costs, bigger/better product, and the 2020s demand/inventory squeeze. That is why a Johnson County median near half a million still looks affordable next to the coasts — and still feels expensive next to a mid-90s paycheck.

Contrast the same residual logic on the coasts: if the house price multiple is 6–10× over thirty years while structure costs are ~2.5×, almost all of the gap is land and entitlement. Kansas keeps the residual smaller; the national essay’s ranking (land first in scarce places; structure costs as the universal floor) holds.

7. Why the whole market is still high in 2026

Even after construction cost inflation cooled from its 2021–22 peak, existing-home prices stayed elevated. Several demand-and-inventory mechanisms did the work:

1. Chronic underbuilding. After the Global Financial Crisis, U.S. housing starts ran below household formation for years. Estimates of cumulative shortage often land in the low millions of units. You do not clear a multi-year supply deficit in one soft spring market.
2. Pandemic demand shock. Remote work, low rates, and stimulus compressed years of price growth into 2020–22 (~40% national house-price jump in a bit over two years in some indexes). By mid-decade, growth rates normalized; the level did not reset.
3. Rate lock-in. Owners who refinanced near 3% are reluctant to sell into 6–7% mortgages. Turnover falls, listings stay thin, and the few homes that list clear at high prices even if volume is low.
4. Local barriers to density. Single-family zoning, minimum lot sizes, parking mandates, and multi-year entitlement fights keep new supply from responding where prices signal shortage. That converts income and population growth into land appreciation.
5. Carrying costs beyond the mortgage. Insurance (especially in disaster-exposed states), property taxes, and HOA fees rose with rebuild costs and assessed values. Monthly ownership pain can rise even when list prices flatten.

Finance is the amplifier. Cheap credit (2020–21) allowed buyers to bid prices up; expensive credit (2022–26) freezes mobility without forcing a 2008-style crash while unemployment stays low and sellers are not forced. High prices can be a sticky equilibrium.

8. Putting the three cost curves side by side

A simplified mental model for 1995 → 2025/26:

• Materials index: roughly +150–160% nominal (~2.5–2.6×). Real increase: moderate. Volatility: high in commodities like lumber.
• Labor rates: roughly +150% nominal for average construction hourly pay. Real increase: moderate. Effective project labor cost can be higher because of overtime, subcontractor margins, and longer schedules.
• Land: multiplies far more than 2.5× in constrained metros; closer to structure-like growth in elastic, low-demand places. National aggregates sit between those poles, pulled up by coastal wealth concentration.
• House prices: roughly triple nominally nationally; worse vs. incomes; far worse in superstar cities.

So the honest ranking for “why is housing so expensive in July 2026?” is:

1. Location scarcity + land-use rules (especially for existing homes in desirable places).
2. Too few homes built for a decade-plus, then a demand spike that inventory never fully absorbed.
3. Mortgage-rate regime that simultaneously juiced prices (when rates were low) and now freezes supply (when rates are high).
4. Structure costs — materials and labor — setting a higher floor under new construction and rebuild/insurance costs, with a permanent step-up after 2020–22.

9. What would actually bring prices down

None of this is a forecast. It is a constraint list:

• Allow more homes where people already want to live — ADUs, duplexes, mid-rise near jobs and transit. That attacks the land premium directly.
• Shorten entitlement and fee stacks so soft costs stop acting like a shadow land tax.
• Expand the construction labor pipeline (training, immigration, factory methods) so wage growth shows up as higher living standards for workers without endless schedule slippage.
• Materials: trade and capacity policy matter at the margin, but even free lumber will not make San Jose cheap if lots and permits stay scarce.
• Rates: lower mortgage rates improve payments but, with tight supply, often re-bid into prices. Affordability needs supply response, not only cheaper credit.

Fossall’s adjacent interest — low-cost mobility and open systems — is a reminder that shelter and transport are joint household budgets. Expensive land near jobs forces longer commutes or exotic housing forms. Understanding which cost is land versus structure is how you stop solving the wrong problem with the wrong tool.