How Long Does It Take To Build Housing?

Last month I posted a series about the difficulty and expense of creating new housing, a problem that drives inflation in the housing economy, an effect that disproportionately consumes the incomes of people with less money and inspires politicians to dump more money onto the problem in the form of subsidies. How long exactly does it take to produce a unit of housing in the United States? The question seems important at a time when rents appear to be rising and forming a bigger part of overall inflationary pressures for consumers. Yet, it is a question that isn’t easy to answer. Producing housing faster will be impossible until we can measure how long it actually takes and why.

Perhaps the most endearing formulation of the “How long does it take question?” is the owl from the ad I remember from childhood for Tootsie Roll Tootsie Pops. How long does it take to get to the center of a Tootsie Roll Tootsie Pop?  

 Mr. Owl’s answer – three licks – leads to the punchline of the ad, “The world may never know.” Unfortunately, that’s how I have felt digging into that question at the national and local level. Before jumping into my foray into finding the data, let’s first acknowledge the generally accepted fact that time is indeed money. Brookings, hardly a paid shill of greedy developers, found in a review of the cost drivers of housing that,

“A key principle of finance is the time value of money: A dollar of income received in the future is worth less than a dollar received today. In real estate development, project expenditures (land purchases, entitlements, constructing infrastructure and buildings) are frontloaded, while income derived from selling or leasing homes may not occur for several years. Regulations that lengthen the development process exacerbate the problem, causing developers to rack up expenses long before earning income from the completed homes.”

I searched, though, for more specific data on exactly how long it takes and I have yet to find a definitive and quantitative analysis in a single market or across several markets. There are many acknowledgements of what Brookings found. For example, one journal article asserts that “the most highly regulated places have project approval lags that average ten months in length, which is three times longer than in the least regulated one-third of communities.” Makes sense, but there is no footnote. And while the figure of ten months tells us something, it doesn’t tell us when that clock starts ticking.

I want to know how long does it take from the moment a project seeks a permit to the time an individual or family could move in. Maybe the federal government or census has data. Unfortunately, the best thing I could find was the promising sounding “Start to Completion” data from the United States Census’ regular construction survey. The problem with this data is that it is far too broad in its focus. Sure, you can find out that 33% of projects with 20 or more units took 13 months or more, and that by 2020, the last year reported, that percentage had more than doubled, to 75%. But these aren’t especially useful numbers in terms of looking at the specific issues or how many months longer than 13 it takes on average to produce a unit.

So, my next step was to start asking cities to share their information. You’d think that at the city level, government would be able to run a simple query on their data base asking something like the following.

This was an example I made and sent with a request to the city of Nashville and a similar request to the city of Albuquerque. The staff in Albuquerque didn’t respond except to send a link to an already publicly available spreadsheet with all building permits. This doesn’t have the basic information to find to determine start and finish of a housing project.

When I sent the request to Nashville, I got an auto response that said,

“This question has been assigned to a zoning examiner and should hear from them by e-mail on 2/17. Questions are being assigned in the order in which they are received. Please keep in mind the volume of requests are high at this time and because of that there are not any earlier dates available.”

A couple of days later I got a response from an “examiner.”

“You would need to do submit the information request as that would be a lot of data if you are needing it for all permits that have been issued/in review/ etc.”

Umm, well, yeah. It would be a lot of data. Where is it? I never heard back.

I haven’t given up the quest, but there are likely three reasons why this is going to be a challenging task. My list comes from this initial ask but also from more than twenty years of working for and with cities. First, cities are not tracking this information. Usually, their data comes from software that manages permits for specific departments. A city will usually have different permitting steps for utilities and land use, for example. Sometimes the fire department is the final sign off. It’s possible to find out, but it likely would require a query of multiple databases.

Second, while people who work in planning and permitting are solid and upstanding citizens, they are generally not evaluating the process or outcomes for efficiency. The job of a reviewer or examiner is to avoid making a mistake that might accidentally permit something that shouldn’t – they are not worried about holding back a permit that should be granted. As I pointed out in a in a post at the San Francisco Business Times a few years back, these reviewers should be incentivized to issue permits fast. They aren’t and they aren’t incentivized to ask “how long is it taking?”

Finally, nobody cares. Local government, especially elected officials are not bombarded by emails from social justice activists, neighbors worried about trees, or people who don’t want anything built when a project slows down or doesn’t happen. It’s the old rule that you can’t fix what you don’t measure. If nobody makes noise about how long it takes to build housing and how that is make it more expensive, then it won’t get tracked. But there is one piece of good news in all this, I did find out the answer to one important question: 678. That is the latest estimate based on extensive research of how many licks it takes to get to the center of a regular Tootsie Pop. How long does it take on average to build a unit of housing in Albuquerque or Nashville? The world may never know.