В Ежедневная газета строителя Focus on Excellence Conference in Denver, we had a chance over breakfast to discuss what we were seeing in off-site housing.
While we have come at this from different angles and have seen different aspects of it, we reached a similar conclusion. Both of us have believed for years that off-site building is simply inevitable—the way we do things has changed so little in a century.
Most people are familiar with a McKinsey study stating that there has been no productivity growth in construction for decades, as well as the demographic projections for our aging workforce. It seems so obvious that off-site modern manufacturing is the solution. But it has not worked yet for production single-family housing—at least not at the full modular level.
Почему?
The core challenges
Let us start with some of the fundamental obstacles:
- Shipping inefficiency. When you are shipping modules, you are shipping a lot of “air.” It is inherently inefficient. Further, you are likely to need more storage space at the site.
- Transportation damage. Let us face it—American roads are not pristine. Trucks bounce and units rack, requiring more on-site work to correct.
- Weather vulnerability. Fully finished interiors can be damaged by rain and the elements if they are not installed and sealed before the weather sets in. One of us has seen severely damaged units when the roofs and wraps were not sealed in time.
- Trade coordination remains complex. You need all the same subcontractors at the job site, just less of their time. But you are not eliminating any trades. Plumbers and electricians still need to make connections, drywallers still need to tape, texture, and paint seams, and foundations (which must be nearly perfect) still need to be completed. Ideally, the same crews would “stitch” the modules together consistently. But that would make them like itinerant oil field workers, moving from location to location until the technology was so widely adopted that they could stay in one place.
Modules aren’t cheaper. When you sum up these factors, modules rarely result in lower costs.
The factory economics problem
What would be the preconditions for effective use of factories?
Factories are generally profitable with high rates of usage, little variation in production, and predictable demand. We have none of that.
An ideal off-site project would work like a Levittown from the past: very few models, no changes, easy-to-develop lots delivered more or less just-in-time, and price points low enough that demand is highly predictable.
What you would not have is municipalities with different standards in the same geography, all with different ideas on what “looks good,” altering each project’s design to suit their tastes. You could not lose six months in getting a grading permit because staff believes standards should change. Imagine Apple getting its iPhone approved in dozens of jurisdictions—in the Denver MSA alone. Think you would have one if that were true? Or one you could afford?
The factory would need to receive orders based on an even-flow sales model that is highly predictable.
See many of those? We don’t.
We are weary of comparing home building to auto manufacturing. When our customers and other stakeholders let us build tens of thousands of standardized homes each year, we can approach that level of efficiency.
Home is personal. We don’t want to live in the same house as every other home on the street. Our local governments regulate against that. What they see as anti-monotony means that, to one degree or another, each home is its own little snowflake. And that snowflake expresses to the world the values and identity of the homeowner who fell in love with it. We aren’t complaining. We are just saying that the “antiquated” delivery system we have offers that flexibility. Manufacturing systems that don’t offer personalization have failed in the past and are doomed to fail in the future.
We offer one more advantage of our “antiquated” delivery system. We build homes in a virtual factory without walls. It may not be as efficient as a factory with walls, but when demand slows, our virtual factory goes home. In that sense, it is more efficient because it is custom-built for our cyclical industry.
That’s why, as antiquated as it seems, it survives.
The mountain town mirage
Many of the use cases to date are in resort areas, with touted success in several Colorado mountain towns. But two things stand out when thinking about core production markets.
First, in the mountains, the cost-per-square-foot can easily be $350 or more, three times that of production markets. And even then, when the previously alluded to on-site work costs are added, the result is often roughly even in cost, but faster. That’s positive, but not game-changing, especially given that, for affordable housing, factories often dictate when deliveries will occur.
For a city building affordable housing, waiting a year for a time slot might work, but it will not work for us.
Imagine calling up the factory and saying, “I think we need units beginning in September 2026. But it might be 2027, or later. We hope we need four per month, but we’re not sure what the market will be like then.”
The reality is, our present system of subcontractors, while not uber-efficient, is uber-flexible, allowing us to move production up or down with demand. It did not get like this by accident.
What is working
There are positive changes in panelization and in how materials are batched and sent to sites. We are routinely using manufactured floor trusses, roof trusses, and wall panels. We are experimenting with adding value to these components by performing more operations in the factory and pre-assembling trusses into cassettes to shorten cycle times.
Arguably, the most efficient builder in the US is NVR in the mid-Atlantic region. They bid subs separately from materials, which they provide in precise amounts to job sites. Work on increasing what is in the panels when they reach the site, and standardized panels that can work for different designs, is encouraging.
But it is also notable that NVR in the mid-Atlantic comes as close as anyone to having the preconditions for modular—high density of projects, huge market share to influence the market and balance out demand and supply. And they have stopped short of modular.
The political distraction
On one level, there is no reason to discourage entrepreneurs from trying new modular solutions. Who knows, one may find the golden answer to all these challenges. So, in that sense, optimism does not hurt.
However, the danger is that many of our politicians want to believe modular will be the game changer that makes houses affordable—because that is much easier to believe in and hope for than to address the real root causes: entitlements and infrastructure.
It is important to help those politicians understand that even if a game-changing innovation were found for modular off-site construction, it would not, in the long run, lower home prices. Why? Because prices equilibrate supply and demand. If we are not increasing supply, a lower cost structure will not lower home prices—it will just raise land residuals.
Thought experiment: If Lennar is the first to find a huge cost advantage, will they sell houses for less than the market price, or use the advantage to buy even more land, outbidding others?
The answer is obvious. While it makes sense for us, as builders, to look for every advantage and cost savings we can, politicians should stop avoiding the hard work of real solutions: entitlement reform and building infrastructure at scale, rather than putting it on the back of every individual project.
English