As of now, the Census Bureau is estimating that 647,000 single-family homes were started in 2014 (the December numbers will get one more revision before they’re final). A question that has always interested NAHB is how many of these starts are teardowns—homes built on site where a previous structure had to be torn down first.
Trying to get information like this from builders can be a bit tricky. It’s not unusual for one of NAHB’s members to contract to build on a site without knowing the site’s full history. However, NAHB has contended that, had a previous structure been there, it would in most cases leave some evidence that a builder would notice. We’ve had discussions with the Census Bureau about adding a question like this to its Survey of Construction (the instrument used to generate the official starts numbers). In the interim, we decided to put the question to builders ourselves.
Hence, our survey for the February NAHB/Wells Fargo Housing Market Index asked single-family builders, “Of the homes you started in 2014, approximately what share were on a site where a previous structure, or evidence of a previous structure, was present before you started?”
Nearly all of the 349 single-family builders responding to the survey answered this question. On average, weighted by starts, they said that just under five percent of their starts were teardowns according to the survey’s criterion. This works out to a total of 31,800 single-family teardown starts in 2014, divided among the four principal Census regions as shown below:
In the West and South regions, the teardowns are close to proportional to total construction (as the South usually has over half of all single-family starts, and the West has about one-fourth). In these two regions, according to the survey results, teardowns accounted for a little over 5 percent of single-family starts in 2014. Elsewhere, teardowns accounted for only about 3 percent of single-family starts in the Midwest, and over 6 percent in the Northeast. Of the four regions, the Northeast has the oldest housing stock, and many of its desirable locations are in densely settled cities, or near the ocean, where buildable open space may be limited.
We know these are difficult numbers to come by. As you know, today, most of the residential infill projects are being done in the most affluent communities where small privately held builders build between 1-3 homes a year. A lot of these builders are not easy to find.
(We are however, starting to see some of the large high profile builders enter the market).
We’ve seen much higher quoted numbers. For instance, NAR’s new-home panel prediction from 2014 suggests: “by the end of 2014, 30 percent of all new construction would be due to “infill” properties”.
What I would like to see is the demolition permit count from the top U.S. zip codes sorted by per-capita income…more $, more teardowns.
In the most active infill communities, 100% of the new construction is the result of a teardown. The builders interviewed in those communities would offer a different perspective of the trend.
Thanks for your comments. Definitions aren’t standard, but most people probably mean something different by “infill” and “teardown” development. Wikipedia, for example, has a definition of infill (construction on any undeveloped land that is not on the urban margin) which would exclude teardown starts as defined in this post. The permit demolition count by zip code is a good idea and something many people would no doubt like to see, but as far as I know it’s not possible to do since the Census Bureau stopped collecting demolition permit data.
As you can imagine we’ve been after accurate data on the subject for the 14 years we have been exclusively focused on it.
What concerns me about the accuracy of the data from this report is this: “Nearly all of the 349 single-family builders responding to the survey answered this question. On average, weighted by starts, they said that just under five percent of their starts were teardowns according to the survey’s criterion”.
Prior to adopting a more Internet-centric model during our stint as a National brokerage specializing in marketing teardowns and land-valued infill redevelopment real estate (residential), we found that the builders building in the 400 plus U.S. communities in which we did brokerage business, the builder’s projects in those communities is 100% of their business. The participants in this survey say only a small fraction of their business is the result of a teardown. So, clearly, the survey doesn’t include the response from the builders in the most active teardown markets therefore…IMO, the data, while interesting, is incomplete.
We had a substantial number of builders tell us that most of their starts were on teardowns, but they tended to be smaller builders (only 4 with 100+ starts reported this, for example). So the weighted average is as low as 5 percent because of weighting by starts. Obviously, if we’re applying a percentage to starts to estimate the number of them on teardowns, we have to use a percentage that takes the number of homes started by each builder into account. In a different venue we may have devoted some space to describing the weighting procedure and its effect on the estimates, but this is a greater level of technical detail than we usually include in a blog post.
Thanks Paul…keep up the great work. We’re always trying to get a reliable “market size” number and without actual demo permits we struggle to obtain it.
That said, the one stat that keeps us believers in a long-term infill redevelopment trend is the fact that of the 132 million homes in the U.S., 31% are 55-95+ years old* – that’s enough older housing stock to keep redevelopment, either in whole or part, a highly active sector of the market for a very long time.
*U.S. Census Bureau’s 2011 Housing Profile
It’s great to see more light shone on the importance of infill construction. However, this study should not be positioned as an accurate reflection of teardown/rebuild activity. The actual volume of infill construction activity we see is much higher and is obviously being undertaken by many builders, developers and investors who are not participating in this housing market index.