Why It's Still So Expensive to Build Homes in America

Episode Summary

Executive Summary: The episode examines why U.S. housing construction has not become more productive over time, despite broader manufacturing gains. Guest Brian Potter argues that housing is constrained by fragmentation, customization, regulation, transport costs, and risk aversion, making prefab and factory-based approaches harder to scale. The discussion also connects these limits to Boeing, nuclear power, lean manufacturing, and talent allocation in the modern economy.

Main Topics: Why housing productivity has stagnated (Priority: 5/5): The hosts frame housing as a product whose costs should fall with productivity gains, yet construction remains labor-intensive and low-tech compared with most manufactured goods. Prefab housing and its repeated failures (Priority: 5/5): Brian Potter explains that prefab has long been promoted as the next big innovation, but repeated efforts—from Lustron to Katerra—have failed to achieve durable mass-market cost reductions. Structural reasons housing is hard to standardize (Priority: 5/5): Housing is shaped by local codes, site-specific conditions, climate, soil, and thousands of permitting jurisdictions, all of which reduce the feasibility of true mass production. Risk, overruns, and limited incentives to innovate (Priority: 4/5): Because construction cost overruns can be extreme while savings are comparatively small, builders are rationally cautious about adopting unfamiliar technologies or workflows. Scale, transport, and the limits of modular manufacturing (Priority: 4/5): Modular construction can help, but buildings must be broken into pieces, shipped, and stitched together, adding complexity and constraining factory scale to local catchment areas. Broader manufacturing parallels: Boeing, nuclear, and talent (Priority: 4/5): The conversation broadens to compare housing with aircraft, nuclear reactors, and steel, highlighting how scale, safety, and talent distribution shape industrial outcomes.

Key Arguments: Housing construction has not enjoyed the same productivity improvements seen in most other physical goods; at best, productivity is flat and possibly down over time. Prefab is not a universal fix because buildings cannot be fully produced like cars or semiconductors; they must be transported and assembled under site-specific constraints. Local variation in codes, soils, weather, seismic risk, and wind loads makes it hard to build a single standardized housing product at national scale. Construction projects have a fat-tailed cost distribution: upside savings are limited, but downside overruns can be enormous, discouraging experimentation. Shipping modular components is expensive enough that many prefab operations are limited to a roughly one-day-drive market, preventing gigafactory-style scale. Manufacturing sectors with continuous-flow processes, like chemicals or steel, are structurally easier to make efficient than construction, which is intermittent and custom. The same logic helps explain why some large industrial sectors, like commercial aircraft or nuclear power, remain difficult to scale and innovate in. Talent may be drawn away from industrial and engineering fields toward higher-paid software, finance, and AI roles, weakening capability in physical industries. Lean manufacturing is not just about minimizing inventory; it also requires flexibility and strategic stockpiles for hard-to-control bottlenecks. Prefab can still work in narrower niches, but major cost breakthroughs likely require robotics, automation, or some new constraint-relieving technological shift.

Data Points: Decline in construction labor productivity in housing: more than 30% - Richmond Fed paper cited in the intro, comparing 1970 to 2020. Number of permitting jurisdictions in the U.S.: about 20,000 - Used to illustrate fragmented building rules and codes. Katerra venture capital raised: about $2–3 billion - Example of a heavily funded prefab startup that ultimately failed. Katerra outcome: bankruptcy in a few years - Illustrates how difficult it is to scale prefab housing profitably. Lustron Homes period: right after World War II - Early prefab housing example discussed as a historical precedent. Toyota semiconductor lesson: after the Fukushima accident - Toyota stockpiled critical semiconductors after supply disruptions. U.S. aircraft production in World War II: more airplanes than have ever been built for commercial aviation in all of history - Used to show the scale of U.S. industrial mobilization. Number of major commercial aircraft manufacturers: 2, maybe 3, maybe 4 - Boeing, Airbus, Embraer, and possibly COMAC. Manufactured home factory catchment area: about a day's drive - Explains why modular/home factories cannot serve huge national markets efficiently. Typical size of a modular factory's annual output: roughly 500 to 1,000 houses - Contrasted with gigafactory-style scale.

Pivotal Quotes: "Yet, yes, it has not improved unlike almost every sort of other physical good has gotten cheaper and less expensive to make overtime. Housing is not like that at all." — Brian Potter: On the core thesis that housing has not benefited from standard productivity gains. "It all seems very old." — Tracy Alloway: On the surprising low-tech reality of home repair and construction. "You know, if you’re smart enough to be a structural engineer, you’re smart enough to switch to like software development and make a lot more money and have a much less stressful job." — Brian Potter: On talent moving out of industrial and engineering fields toward higher-paying sectors.

Implications: Housing affordability will remain hard to fix without changes that cut actual construction costs, not just zoning or financing. Bigger gains likely require robotics, better standardization, and industry structures that can absorb innovation without extreme downside risk.