It’s not technically Monday but it feels like it is, so here’s a map from indefatigable Tweetable map-maker Scott Shaffer. It shows the number of “new construction permits” (i.e. bulldozing of an existing building slash parking lot) in Minneapolis over the last eight years or so.
As you can see, the vast majority of new construction permits are located in areas zoned for single-family housing. This is also known as the “teardown problem” and there are boatloads of digital ink devoted to debating this issue online on various neighborhood group pages and forums.
Shaffer leaves the following comment on Twitter:
livability advocate: the comp plan is proposing we bulldoze the city! if only we keep our low-density mansion zoning, we can stave off the bulldozers!
SW Mpls: [see map]
Personally, I’m a bit agnostic about SFH teardowns. In Saint Paul, I’ve listened to hours of testimony about teardowns in Ward 3, and the city passed some new rules that were intended to serve as an experiment. Based on some anecdotal feedback that I’ve heard, the new rules haven’t really satisfied neighbors who are upset about other people tearing down old homes and building new ones.
I sincerely wish concerned neighbors the best in dealing the teardown issue and coming up with regulations that are sensible and actually work. But I’m also skeptical that any regulations will accomplish anything, because the root of the problem is huge demand for urban housing in wealthy enclave-like neighborhoods combined with an endlessly hot housing market. No matter what regulations you come up with, apart from a very restrictive full moratorium of some kind, crafty developers or home-owners will find a way to work with or around the letter of the law.
The root of the matter is private property itself, and people have been expanding and improving their private homes since the dawn of civilization. Short of a socialist revolution, I don’t see much hope for changing those fundamental dynamics.
(Tearing down an old structure for multi-unit housing is another story, to me, because in that case you are actually helping to create new housing. In theory, and most of the time, in practice, creating new housing alleviates rising housing prices. To me, that’s a different conversation.)
Anyway, cool map!
To me tear downs likely draw some single family homewner from suburbs into city….but hard to know without real data, it just may be long time city residents who have done well and want more..
I saw in L.A. in 80s that both tear downs and apartment building went out at same time as response to increased demand (when it wasn’t disallowed by zoning).
But clearly if people wanted to preserve old, less.dense neighborhoods, they could start with stopping tear downs and address 80-90 percent of the problem in their minds… rather more harmful to rents/supply policy of stopping four plexes, bigger apartments at nodes, ADUs etc.
“the vast majority of new construction permits are located in areas zoned for single-family housing”
The “low density” zoning in yellow is not the same as SFH zoning. In my neighborhood R2B is the lowest density zoning that exists and much of the neighborhood is yellow on this map.
There are a lot of multifamily dots, and those are over 5 units each, so it looks like most housing getting built is multifamily, not tear-down SFH replacement.
Can we have the numbers please? It would useful to see the number of permits and the number of units.
Ah. Thanks for the correction. Hopefully Scott weighs in. Certainly the # of households and the # of permits are not equivalent.
Units ≠ permits, and it would, I agree, be interesting to see what percentage of new housing *units* built are in single-family vs. small multifamily vs. larger multifamily structures.
However, I think Scott’s map still makes a compelling and interesting point in its own right. A significant stated reason that homeowners object to liberalized zoning for 2- through 4-plexes is often that they don’t want to experience the disruptive impacts of construction in the vicinity of their homes. (See Carol Becker’s recent piece invoking the specter of “bulldozers.”)
The most immediately eye-grabbing feature of this map is that cluster of purple dots in southwest Minneapolis, which does a lot to debunk the notion that single-family zoning either protects single-family homeowners from construction on their block, or protects single-family homes from demolition.
Also, that it protects them from larger structures. Those new single family homes are large enough to be multiple units.
Dan, I think you have missed the main thrust of Carol Becker’s piece. The discussion of bulldozers therein is brief, dramatic, but not her point. The concern is what happens when new 4plexes are of a height and massing as to impose externalities on their neighbors.
It’s a fair concern. Another concern I have is that by moving away from transit oriented development (encouraging the multi-family to be near transit) the City will set up a perverse incentive for those owning the lowest value homes (think affordable) to defer maintenance and let those properties rundown until they have little or no economic value because then they can be torn down and replaced with a 4plex.
That’s a lesson I learned here the hard way. You use inflammatory language to try to make your point and people focus on the language instead of the points.
Except that the plan directs the vast bulk of multi-unit units toward transit, so…
I don’t see the big difference in incentive to letting your rundown house be replaced with a McMansion versus a fourplex. And I don’t think anyone lets their house get rundown for the future payout.
I don’t see why people who want smaller apartments should be forced to live on the most polluted, highest-traffic streets. Everyone, no matter whether they rent or own, whether they have a house or apartment, should have the choice of being 1/4 block from a bus stop, or on a quiet neighborhood street.
Certainly, four-plexes are more likely to offer family housing than larger apartment buildings, and many families like to be a bit away from busy streets, possibly even across the street from a neighborhood park.
As I mentioned in the subtitle of the map, I categorized R1, R1A, R2, and R2B zones as low-density, which is how the City of Minneapolis does it. So your neighborhood being yellow isn’t a mistake.
I made this map because an author on this site claimed that upzoning would lead to “bulldozing the city,” and the low-density zoning would prevent that. The point of the map is to show that most construction sites (i.e., bulldozers) are for single-family homes, and I think that’s born out in the numbers (hope the formatting isn’t too bad):
Type | New Construction Sites
Here’s a spreadsheet with a chart, details, and a summary: https://docs.google.com/spreadsheets/d/1bVUC6G4QtWJ3HfTLh0-6CmuMDubTYJYRq4l2UuRaoOU/edit?usp=sharing
I got the raw data here: https://gisdata.mn.gov/dataset/us-mn-state-metc-econ-residential-building-permts
Thanks for the data Scott. I seem to be missing one permit when I copied your data set, I get 921 (not 922) permits issued.
Those 921 permits were for 14,750 units. While SFUs dominate the number of permits, they are not a very significant amount of housing being built.
716 permits (and units) were SFUs. If you add the 28 ADUs (a pathetically low number), and the 15 town homes that were a single unit, you get 759 SFUs or 5% of the units permitted.
162 of the permits ( less than 18%) produced 13,990 of the 14,750 total units permitted, or 95%.
That seriously challenges the drumbeat I hear often on this site that tear-downs to build new houses is preventing construction of new apartments. I doubt it. If someone can afford to buy a house to tear it down and build a bigger one, the property is probably sufficiently valuable that a four-plex on the same site would be be very high rent. Only where larger structures can be built is multi-family after teardown (in the high amenity neighborhoods where it is desired) going to be practical.
The data shows that the “missing middle” (and ADUs) are simply not a significant way to add units to the total housing stock: 2-4 unit construction are only 22 permits and 72 units.
The data shows that the missing middle is missing.
And, to follow onto Adam Miller, the missing middle is missing because construction of the missing middle is too onerous from a zoning perspective. Of the 716 SFH’s built, 589 could not have been otherwise due to R1 zoning restrictions. Some number of additional homes probably could not be effectively made into multifamily due to other property restrictions (FAR, parking, setbacks, etc).
If upgraded 700 SFH to 2-4 unit structures, this would amount to ~1600 more housing units, a 10 percent boost on the total supply, not insignificant. That is less than 100 SFH teardowns upgraded to small apartments per year.
Blaming the “missing middle” on zoning suggests a simple solution to building more 2-4 unit buildings that may not exist.
Yup, 587 of the permits issued were for SFH zoned areas, but I am more interested in the 16 SFHs built in R2B and the 21 built in R4. This suggests the issue is not zoning but economics. Developers accessed risk and return and decided to build single family homes even when the zoning allowed multifamily construction. They are building what they believe they can make money on and even where it is already zoned for multifamily they are building SFHs.
So perhaps the real scarcity in our market is for SFHs for the increased number of high income families that want homes in the City?
Instead of rezoning R1 and R1A for 4-plexes, why not change the requirements of those zones so tear-downs and remodels have limitations on height and massing that match the neighbors’ context, but provide a bonus for adding a second unit?
I suspect it would yield more units than the current plan. As the separate ADU thread discusses the cost of detached ADUs makes them cost prohibitive on most lots, but maybe a new SFH size bonus for including a 2nd unit (keeping the owner occupied ADU requirement for detached ADUs only) would be a more practical idea.
That strikes me as a “too” instead of an “instead.” Let’s both allow small multi-unit and provide size bonuses for multiple units.
And, of course, the current zoning works that way. You’re allowed more height and FAR when you build multifamily over single family in an area zoned R5, for example.
I believe that up until 2016, the minimum lot size in an R2B to build a duplex was 10,000 square feet (now it is 5,000 square feet). My house is on an R2B-zoned lot, but with just under 6,000 square feet of lot, a single family home is all I could have built anew here.
How many of the 16 SFHs built in R2B zoned areas were on lots larger than 10,000 sq. ft? And how many R2B lots are over 10,000 square feet to begin with?
Is there any data on demo permits in R2B for duplex construction since 2016, when the lot size requirement was halved?
And looking at Scott’s data set, it looks like there were 16 NON-SFHs built in R2B and 21 NON-SFHs built in R4. By comparison, there were 105 SFHs built in R2B and 26 in R4.
Then again my question is how many of the 131 lots on which SFHs were built had enough square footage to built a multifamily building?
Present zoning doesn’t allow (except as an ADU) more than 1 housing unit on a lot in R1 and R1A, the debate about the new plan is that it would. I am suggesting an alternative that would create a market incentive to encourage new development to include “Granny” units.
For reasons to lengthy to get into here the presently popular argument that zoning is responsible for the “missing middle” and that exclusionary zoning is driving housing costs up seems to me mostly wrong.
So in my view if advocates for the new plan get what they want it will result in more replacement of affordable housing with expensive housing and will benefit developers but not the City residents proponents of the plan changes say they are trying to serve.
I’m suggesting a solution that recognizes the economic issues (rapid increases in 1) the number of high income households, 2) the income of those high income households, and 3) income disparity between households) creating our present imbalances and uses those imbalances to create more affordable housing.
For instance: If you want to tear down an existing 1600 foot home and build a new 3800 foot home it seems safe to assume:
1) the massing and height of that building might not fit in with its neighbors;
2) it can still be built without variances;
3) the allowable structure could also fit 3 separate 1200 foot apartments.
But market forces aren’t going to going to give you that triplex even if you rezone the lot.
My proposal might play out like this:
You can build a 2000 foot new home because that is what fits in with your neighbors, but if you include a 600 square foot granny flat you can build that 3800 square foot home. So the builder gets an additional 1200 square feet in the home and the City gets an additional housing unit that is likely affordable.
An alternative would be to charge a significant development fee for tear-downs that goes into an affordable housing fund, but reducing that fee for each additional unit of housing built on the lot.
Sure it’s a compromise, but a compromise that might actually accomplish what is wanted.
I like the idea of allowing larger structures as unit counts go up. I don’t even think that’s inconsistent with the plan (which doesn’t specify that level of detail). And, as I mentioned, existing zoning does that in places.
I think it’s a stretch to say that market forces won’t provide triplexes in a new unit, or that they won’t be “affordable.” While the rent you’d need to cover op expenses for 2 (if owner-occupied) or 3 (if landlorded) units in a triplex would be more per square foot than the equivalent SFH rental (same goes for sales price comparison), it would be less expensive per housing *unit*.
Here’s a blog post breaking down the economics of these different scenarios in Portland https://medium.com/@pdx4all/portlands-residential-infill-project-still-has-major-flaws-housing-advocates-say-6a225ec290e Their policy proposal assumes allowing skinny lot single family homes (something we’re not) but the other possibilities of lot redevelopment are pretty similar to ours (including lot costs, construction costs, etc). Their analysis shows that a new triplex is as likely as a new large home, both of which are less likely than a duplex, and all of them are far less likely than the existing home just being bought (and maybe remodeled). That last point is important and jives with the overall tone of your other comments – that small infill may not have a huge impact on supply (this ignores converting SFHs to 2-4 units, though).
But it’s worth repeating that the areas of town you’re saying have prices beyond the likelihood of smaller developments like fourplexes are only this way to begin with because of decades of our current zoning regime. There are plenty of large homes that *should* already be duplexes or triplexes. There are many lots that should be small apartments or condos. The fourplex proposal probably doesn’t go far enough for this reason – we should be talking about 8-12 units by right on every lot rather than 4! And yeah, sometimes developers will buy a crummy house on a lot and choose to build a really nice and large SFH on it instead of an 8-plex, and the pro-housing crowd should be fine with that, too.