Michael Wiebe
michaelwiebe.bsky.social
Michael Wiebe
@michaelwiebe.bsky.social
Economics (UBC), yimby, replication, effective altruism, data science.
If you conflate vacancy chains and filtering, you also mix up how long their effects take.

Filtering takes decades, since homes depreciate slowly. But vacancy chains take only one year (as measured in the Finland study).
November 14, 2025 at 9:26 PM
It's not!
November 12, 2025 at 9:31 PM
Single-family zoning causes gentrification.
November 11, 2025 at 5:33 PM
Single-family zoning causes gentrification.
November 10, 2025 at 6:01 PM
Why do you think developers are uniquely responsible for promoting integration? If integration is a social good, shouldn't all of us be responsible for providing it?
November 9, 2025 at 7:32 PM
In mansion district Shaughnessy, land values increased 5x, just due to rising demand to live in Vancouver. There are no upzonings, since the neighborhood is heritage protected.

IZ does nothing to capture this land value. LVT works, because it's unconditional.

2/
November 5, 2025 at 6:12 PM
Doesn't this answer your question? p.17
November 5, 2025 at 2:34 AM
Land values increased 5x in a mansion district. IZ cannot capture this.

michaelwiebe.com/blog/2025/09...
November 4, 2025 at 5:35 PM
I don't get this distinction. They both do the same thing.

In my terminology, density bonus is a type of IZ.
November 4, 2025 at 3:25 PM
New post on inclusionary zoning, for my housing literature review.

IZ requires developers to provide subsidized homes in their new apartment buildings. Without providing compensation to developers, this unfunded IZ is just a tax on new housing.

1/
November 3, 2025 at 8:03 PM
Local control over zoning is a government failure, because restrictive zoning imposes a negative externality on non-residents who are priced out. Higher-level governments need to set zoning policy to internalize the externality.
October 31, 2025 at 5:20 PM
Instead, it is the market-rate homes across the entire city that do the cross-subsidizing. Ie. the pass-through of the IZ tax is market-wide rather than at the building level.

en.wikipedia.org/wiki/Tax_inc...
2/
October 30, 2025 at 5:52 PM
And the first stage looks good.

Agglomeration effects for innovation: might be real!

4/
October 29, 2025 at 8:10 PM
Moretti basically only does the 'shift' part, so it's not a standard IV.

michaelwiebe.com/assets/moret...

2/
October 29, 2025 at 8:10 PM
Moretti 2021 uses an IV for cluster size based on the number of inventors in the same firm in other cities. But this has a coding error, and correcting the error gives a null result.

But the idea sounds like shift-share IV, so I tried that, and it works:

1/
October 29, 2025 at 8:10 PM
They generally don't find this. (Here, More Constrained = Low elasticity, so this number should be smaller.)

Note: this comparison consolidates the separate predictions about price and quantity, since those are the first-stage and reduced-form terms of the IV.

2/
October 28, 2025 at 10:37 PM
PS. Here's a derivation of the bias in the estimated average supply elasticity when don't assume constancy:
October 28, 2025 at 10:35 PM
C'mon.
October 23, 2025 at 1:44 AM
Rollet 2025 Fig 3b shows that the average upzoning raises FAR by 1.0, but only increases built FAR by 0.10 over ten years.

Upshot: theoretical zoned capacity needs to be at least 10x bigger than the target amount of housing.
October 22, 2025 at 5:16 PM
On the amyloid cascade hypothesis for Alzheimer's

www.astralcodexten.com/p/your-revie...
October 19, 2025 at 9:55 PM
Possibly!
October 17, 2025 at 6:12 PM
But in a large upzoning, apartment-zoned land becomes much cheaper. So the parcel moves from P_H = 2 to P_A = 2.8, for land lift = 0.8.

Cheap apartment-zoned land means cheap homes, but less value capture for the city. Therein lies the misaligned incentives.

6/6
October 17, 2025 at 6:01 PM
Upzoning moves a parcel from the low-price house-zoned market (H) to the high-price apartment-zoned market (A).

In a small upzoning, the own-parcel effect is to increase the land price from (initial) P_H = 2 to (final) P_A = 6. So land lift = 4.

5/
October 17, 2025 at 6:01 PM
Mike Manville on the failure of value capture to actually capture value.
October 16, 2025 at 7:36 PM
Michael Manville points out that a lot of land value is created without upzoning.

I also made this argument in my Condon review: Shaughnessy's land values have 5x'ed, with zero upzoning!
michaelwiebe.com/blog/2025/09...

5/
October 16, 2025 at 12:00 AM