With U.S. lumber prices hitting a record-for-this-time-of-year (going into spring) this is a timely piece from Construction Physics, February 16:
One of the goals of this newsletter is to figure out ways to improve construction productivity and reduce the costs of construction, particularly residential construction (ie: housing) - if you could reduce the costs of construction, housing costs would fall as well. Roughly 15-18% of US GDP goes towards housing costs (including mortgages, rents, and construction of new homes), and fraction of income spent on housing is inversely proportional to income, so being able to reduce costs associated with it would have substantial benefits.
But outside of this little corner of the internet, there’s relatively little focus on construction costs as a driver of housing costs. Instead, discussions of housing cost tend to be focused on what we might call ‘price distortions’ - increases in housing costs caused by things like zoning rules, height caps, onerous permitting requirements, and other policy decisions [0]. These factors are a major focus of urban economists, and of policy advocates like Brookings, Niskanen, and the YIMBY movement. They’re also fairly central in housing discussions in major media outlets such as the New York Times, the Wall Street Journal, and the Atlantic (as well as smaller publications like Works in Progress.)
I suspect this is at least partly due to the fact that urban areas like San Francisco and New York, which have a disproportionately large influence on what gets discussed in major media outlets [1], have both extremely high housing costs, and extremely restrictive building policies that limit housing construction.
This isn’t necessarily unreasonable - supply restrictions are a problem with a clear path towards a solution (let people build more houses!), and it’s a solution that can (theoretically) be reached by getting the right people paying attention and talking about it. So it makes sense for media discussions around housing costs to focus on them - you won’t be able to reduce construction costs by writing an op-ed in the New York Times, but it might help get SB9 passed. But because it’s so central to the conversation, it would be easy to get the idea that policy decisions are the main thing that affect housing costs, and (implicity) that construction costs don’t matter as much. So it’s worth diving into some of the factors that affect the price of housing, and how construction costs fit in.
Land and structureWe can think of housing as a bundle of a parcel of land and a structure (plus other improvements) sitting on top of it. We can thus break down the cost of housing into a) the cost of putting up the structure, and b) the cost of the land it sits on [2]. For single family homes, this breakdown is often done by taking the purchase price of a house, then subtracting the (depreciated) costs of construction, typically determined by using RS Means or similar construction cost data [3]. What’s left is the value of the land itself. When using this method of cost breakdown, outside of price distortions that increase the cost of construction (which we’ll look at), price distortions caused by limits on how much housing can be built will show up as increased land prices, assuming you keep the amount of housing per lot constant (more on this later.)One way to think about it is that land is also a bundle, consisting of a ‘hedonic’ or consumption portion (the value people get from having the land itself) and a ‘permission slip’ that allows you to build a certain quantity of housing on it. In general, the hedonic value of land is fairly low - when Ed Glaeser calculated this for 21 metro areas in early 2000 (by analyzing home price sales and estimating how much a larger lot added to the home’s value), he found that in 16 of them, the hedonic value was less than (inflation adjusted) $1.50 per square foot, or less than $12,000 dollars for the median lot size of ~8,000 square feet (though we should be careful with a simple inflation adjustment of this hedonic value [4].) By comparison, the median new home in 2020 sold for around $330,000.
A similar study of single family homes in Boston found that the hedonic value of land for the average lot was just $11,200, compared to an average house cost of $450,000.
The permission slip, by contrast, is often incredibly valuable, especially if the permitting body limits how many of them it gives out [5]. For instance, in the previous study Glaeser calculated the hedonic value of land for single family homes in San Francisco at $4.10 per square foot (the highest out of any metro examined), or about $10,000 for the average lot size at that time. The ‘zoning tax’ portion, by contrast (the portion of the cost due to various building and supply restrictions), was priced at around $220,000 - in other words, over 95% of the land’s value (and over half the price of the house), was due from being allowed to build on it [6]....
....MUCH MORE
As always, very sharp analysis.
Previous links to Construction Physics:
