Joe Fish

so for this san francisco upzoning, what's actually gonna get built in practice?

my quick read is that only vacant lots, commercial buildings, and single family homes (maybe duplexes?) are legally + financially feasible to redevelop? and single family homes have bldg code issues

Why is the labor share of income declining? An informal meta-analysis Seven years ago, Thomas Piketty’s Capital in the Twenty-First Century landed on coffee tables across the English-speaking world.

i liked this substack on the labor share. especially noteworthy since the declining labor share is really limited to
1) the US and Canada
2) the period following 2001.

Complicates a lot of narratives!

jzmazlish.substack.com/p/why-is-the...

same thing, but with 4 different base years. 2021 is the worst year for this exercise.

if you pick any other year, I think the most straightforward conclusion is "good wage gains but man does shelter inflation suck"

it's also a bit weird because if you index to 2020 instead of 2021 (and use median wages instead of the wage series they use) it looks pretty different.

Which isn't to say it's wrong, but the "one chart to kill the vibescession" people are usually overblown.

bsky.app/profile/sadb...

although, it also looks like they're using mean wages and not median wages, the former of which have grown slower since 2021

this chart is a bit weird because it *really* matters which year you set as your baseline.

if you reindex to 2020 instead of 2021, you get that, with the exception of shelter, wages have kept pace with essentials.

On 2) you basically have to have something where the option value of developing later at 85+ feet is both so strong and so currently infeasible that developers would prefer to land bank as oppose to develop at 65 ft.

I don’t think this is true nor is this what’s being argued

This is why people need to write down models.

1) It’s possible for a spot up zoning to be ~entirely capitalized into land values but still drive down price of apartment-eligible land via spillovers
2) it’s ~impossible for production to go down bc the increase in value comes from higher profits

Upzoning Fantasies Last week I invited a market-rate developer as guest speaker in my graduate housing class at the University of San Francisco. I told him…

People, like in this article, make the argument that (spot) up zoning increases land values, but even if it increases the value of the upzoned property it should decrease the value of “land you can build apartments on”, which is what multi fam developers would care about

medium.com/@el.compay.n...

maybe you just tell teachers not to design tests where time is the limiting factor and give everyone extra time?

Admittedly, this is way easier in a pre-LLM world, but it seems like an okay compromise

the accommodations stuff is similar to regrade requests in that when you zoom out, it's disproportionately getting used by higher SES students.

which means, even if everyone using them is genuine, it's a huge benefit to higher SES students. maybe the solution is everyone gets accommodations, idk.

"we have a small army of NGOs because we were anti-union in the 1970s and 1980s" is absolutely a take you here often in San Francisco politics (and is, IMO, correct)

For instance, geography classic, City as a Growth Machine
relies *super* heavily on a spatial equilibrium assumption tucked away in a footnote.

(the assumption here being that bc labor is perfectly mobile, local development is zero-sum; note also, no agglomeration effects are implicitly assumed)

my contribution to the "equilibrium" discourse raging on econ twitter is a that:

1) everyone has to "close" their model
2) i think econ is relatively *more* aware of what assumptions are being made when doing so

@alexandersahn.bsky.social cool paper!

I'm assuming the main reason the focus of the paper was on current % zoned multifamily was data availability?

Any hope on being able to look at anything related to down zoning specifically?

A cool paper I wish I had done argues a lot of these down zonings (broadly, increases in exclusionary zoning) happened in response to more Black people migrating to non-Southern cities.

www.journals.uchicago.edu/doi/full/10....

the hedonic treadmill is definitely real. if someone hasn't written this essay, though, i'd be super interested in reading if this kind of "the past was always way better" was as prevalent then as it is now.

was it always like this, or is the post-COVID period just uniquely weird?

The period from 2015 to 2025 has been the best decade in the past 50 years for large and broad-based income gains

This is even true if you zoom in on young households. To explain the expressed economic malaise, you either need to go Full Stancil or argue that something, eg housing, is super salient

Mamdani seems thoroughly willing to try a bunch of shit, with full knowledge that some of that shit might not work (and also seems willing to course correct).

Incredibly commendable in a world of do-nothing, overly cautious politicians!

How do you think about vacancy chains in terms of how the “effects of local supply” papers get identification?

Is it just that those local papers’ results are probably attenuated by spillovers?

People are purportedly losing faith in the value proposition of a college degree, however, it is simultaneously true that:

1. the price of tuition has been falling rapidly relative to CPI
2. the college wage premium has been pretty flat

so it might be a particularly *good* time to go to college?

the issue with Bruce's data, to close the circle, was that there are people, particularly those who are homeless, who are very challenging to survey and who likely have very low incomes.

Although the data we have on them tends to suggest that most homeless people make ~400 / month

My recollection of the debate was correct, and the debate largely revolved around how to account for and measure in-kind benefits, particularly in survey data.

If you use Bruce Meyer's data, you get an extreme poverty rate of 0.11%.

web.archive.org/web/20250409...

i would bet a non-insubstantial amount of money that the answer to "why is 1% of the US living under $3/day" is "the survey wasn't counting in-kind benefits correctly".

iirc there was a whole debate spurred by similar claims made by Kathy Edin and Angus Deaton

China has brought millions out of poverty. The US has not – by choice Despite the US’s economic success, income inequality remains breathtaking. But this is no glitch – it’s the system

the guardian article is writing a takedown of the US vis a vi China, but the fact that they could have written the exact same article but with the UK, Austria, Sweden, Japan, or Italy should be a sign that the data are weird not that their argument is strong.

www.theguardian.com/us-news/2025...

aside from being a chart crime, I'm think the answer to "why are fewer people living in extreme poverty in china than the US" is "the data aren't comparable".

China is reporting 0% extreme poverty in a consumption based survey; the US reports ~1% in an income based one. *Sweeden* reports a 0.75%.

the long view of shelter inflation also makes a much stronger case for the primacy of supply constraints in determining housing affordability

if you think the stagnation + drop beginning in the late 1960s reflects adoption of growth controls, that's kind of the whole story!

@aarmlovi.bsky.social

of course, this doesn't really vindicate the doomers in the way they think because today is still better than any point in the past 45 years (2019 weirdness excluded). but the gains have certainly been more modest under these measures.

this is almost entirely driven by shelter inflation tracking mostly 1 for 1 with income gains

Housing theory of everything strikes again. I'd be hesitant to go fully to bat for this, but if you think the 1970s drop was partly urban growth controls, they matter a lot more than 2008!

People seldomly make this point correctly, but you can make a mostly-coherent argument that if you look at income gains in terms of the prices of "essentials", they're somewhere between modest and stagnant, depending on which year you pick for your base.