Joe Fish
@sadbusdriver.bsky.social
PhD student doing urban econ and industrial organization at Duke. Former highest paid cashier in the Midwest; former RA at Eviction Lab; Macalester College. Not a real bus driver
has anyone done the pay-productivity gap stuff by country? my loose read on the EPI graph was that the gap was mostly wage inequality.
but cross-country comps might shed light on whether this is skill biased technological change or decreasing worker power or both
but cross-country comps might shed light on whether this is skill biased technological change or decreasing worker power or both
January 8, 2026 at 6:54 PM
has anyone done the pay-productivity gap stuff by country? my loose read on the EPI graph was that the gap was mostly wage inequality.
but cross-country comps might shed light on whether this is skill biased technological change or decreasing worker power or both
but cross-country comps might shed light on whether this is skill biased technological change or decreasing worker power or both
Major homebuilders' stocks all down today following the proposed ban on instutional investors owning homes.
probably because the ban would kill buy to rent? If so, that's another reason why the ban seems pretty misguided
probably because the ban would kill buy to rent? If so, that's another reason why the ban seems pretty misguided
January 8, 2026 at 12:53 AM
Major homebuilders' stocks all down today following the proposed ban on instutional investors owning homes.
probably because the ban would kill buy to rent? If so, that's another reason why the ban seems pretty misguided
probably because the ban would kill buy to rent? If so, that's another reason why the ban seems pretty misguided
Most San Francisco neighborhoods haven't meaningfully grown or changed since the 1980s
Keeping neighborhoods in amber remains an insane way to run a city and is a large part of why SF is so unaffordable
(This graph inspired by what will be a titanic fight over a 9-story apartment in west portal)
Keeping neighborhoods in amber remains an insane way to run a city and is a large part of why SF is so unaffordable
(This graph inspired by what will be a titanic fight over a 9-story apartment in west portal)
January 7, 2026 at 1:52 AM
Most San Francisco neighborhoods haven't meaningfully grown or changed since the 1980s
Keeping neighborhoods in amber remains an insane way to run a city and is a large part of why SF is so unaffordable
(This graph inspired by what will be a titanic fight over a 9-story apartment in west portal)
Keeping neighborhoods in amber remains an insane way to run a city and is a large part of why SF is so unaffordable
(This graph inspired by what will be a titanic fight over a 9-story apartment in west portal)
i think there's a very strong "does a fish know it's in water" issue with developers.
they (along with tenant activists, interestingly) often have encyclopedic knowledge of local stuff and are remarkably unfamiliar with what's happening a two-hour drive away.
they (along with tenant activists, interestingly) often have encyclopedic knowledge of local stuff and are remarkably unfamiliar with what's happening a two-hour drive away.
In the housing space, a common refrain from the large scale development community is that broad upzoning doesn’t matter. Instead the challenges are financing and hard costs.
Developers focus on these because they’re looking at pro formas forecasting 2 years out, not 20.
48hills.org/2025/12/the-...
Developers focus on these because they’re looking at pro formas forecasting 2 years out, not 20.
48hills.org/2025/12/the-...
December 31, 2025 at 8:09 PM
i think there's a very strong "does a fish know it's in water" issue with developers.
they (along with tenant activists, interestingly) often have encyclopedic knowledge of local stuff and are remarkably unfamiliar with what's happening a two-hour drive away.
they (along with tenant activists, interestingly) often have encyclopedic knowledge of local stuff and are remarkably unfamiliar with what's happening a two-hour drive away.
my understanding of some of the trade shock literature, from the US and other countries, is reskilling can work well, but it tends to work much better for younger workers.
like the specific thing that's super hard is what to do with a 45 year old who doesnt want to move and cant reskill easily
like the specific thing that's super hard is what to do with a 45 year old who doesnt want to move and cant reskill easily
we verifiably do not do a good job of mitigating the societal impacts of technology on labor. this leaves us with things like the entire state of west virginia yearning for the return of the coal economy, whatever's going on with detroit, etc
This is really grim. www.nytimes.com/2025/12/28/o...
December 30, 2025 at 8:14 PM
my understanding of some of the trade shock literature, from the US and other countries, is reskilling can work well, but it tends to work much better for younger workers.
like the specific thing that's super hard is what to do with a 45 year old who doesnt want to move and cant reskill easily
like the specific thing that's super hard is what to do with a 45 year old who doesnt want to move and cant reskill easily
what's funny about the CA wealth tax is that doing a one time tax, with a credible commitment to not do it again, is the *correct* way to do this... assuming people can't avoid the tax.
and yet because it's California, i imagine this will work out in the dumbest way possible
and yet because it's California, i imagine this will work out in the dumbest way possible
December 27, 2025 at 1:10 AM
what's funny about the CA wealth tax is that doing a one time tax, with a credible commitment to not do it again, is the *correct* way to do this... assuming people can't avoid the tax.
and yet because it's California, i imagine this will work out in the dumbest way possible
and yet because it's California, i imagine this will work out in the dumbest way possible
The MA ballot initiative on rent stabilization, which limits hikes to the minimum of 5% and the CPI, with no vacancy decontrol, is really close to rent control
If Boston had enacted this in 2011, by 2023 this would be binding in 100% of zip codes, and 60% would have rents over the cap by > 30%
If Boston had enacted this in 2011, by 2023 this would be binding in 100% of zip codes, and 60% would have rents over the cap by > 30%
December 22, 2025 at 10:39 PM
The MA ballot initiative on rent stabilization, which limits hikes to the minimum of 5% and the CPI, with no vacancy decontrol, is really close to rent control
If Boston had enacted this in 2011, by 2023 this would be binding in 100% of zip codes, and 60% would have rents over the cap by > 30%
If Boston had enacted this in 2011, by 2023 this would be binding in 100% of zip codes, and 60% would have rents over the cap by > 30%
Reposted by Joe Fish
I'll do an addendum/correction on this:
If you look at changes at the MSA level (so adding up CoCs), it's easier to see medium + long run affordability impacts.
First, note that the correlation in *levels* is still strongly positive. Higher prices correlate with higher homelessness.
If you look at changes at the MSA level (so adding up CoCs), it's easier to see medium + long run affordability impacts.
First, note that the correlation in *levels* is still strongly positive. Higher prices correlate with higher homelessness.
December 22, 2025 at 7:19 PM
I'll do an addendum/correction on this:
If you look at changes at the MSA level (so adding up CoCs), it's easier to see medium + long run affordability impacts.
First, note that the correlation in *levels* is still strongly positive. Higher prices correlate with higher homelessness.
If you look at changes at the MSA level (so adding up CoCs), it's easier to see medium + long run affordability impacts.
First, note that the correlation in *levels* is still strongly positive. Higher prices correlate with higher homelessness.
I'll do an addendum/correction on this:
If you look at changes at the MSA level (so adding up CoCs), it's easier to see medium + long run affordability impacts.
First, note that the correlation in *levels* is still strongly positive. Higher prices correlate with higher homelessness.
If you look at changes at the MSA level (so adding up CoCs), it's easier to see medium + long run affordability impacts.
First, note that the correlation in *levels* is still strongly positive. Higher prices correlate with higher homelessness.
December 22, 2025 at 7:19 PM
I'll do an addendum/correction on this:
If you look at changes at the MSA level (so adding up CoCs), it's easier to see medium + long run affordability impacts.
First, note that the correlation in *levels* is still strongly positive. Higher prices correlate with higher homelessness.
If you look at changes at the MSA level (so adding up CoCs), it's easier to see medium + long run affordability impacts.
First, note that the correlation in *levels* is still strongly positive. Higher prices correlate with higher homelessness.
TIL the wealth of nations came out in 1776. A little on the nose, universe
December 21, 2025 at 7:26 PM
TIL the wealth of nations came out in 1776. A little on the nose, universe
The real STEM vs humanities is social sciences (can read a graph and a methodology report) vs everyone else
December 20, 2025 at 8:46 PM
The real STEM vs humanities is social sciences (can read a graph and a methodology report) vs everyone else
one bugbear of mine is people who insist that the government is hiding the "real" unemployment rate or focuses on the one that's artificially low
1. this is not true, the government publishes lots of different unemployment metrics
2. if you know the headline one, you basically know all the others
1. this is not true, the government publishes lots of different unemployment metrics
2. if you know the headline one, you basically know all the others
December 20, 2025 at 6:39 PM
one bugbear of mine is people who insist that the government is hiding the "real" unemployment rate or focuses on the one that's artificially low
1. this is not true, the government publishes lots of different unemployment metrics
2. if you know the headline one, you basically know all the others
1. this is not true, the government publishes lots of different unemployment metrics
2. if you know the headline one, you basically know all the others
it's true that employment in taxis (so, ride-shares) has grown way faster than overall employment, but it's not clear to me why someone who works full-time as an Uber driver is a "gig" worker.
if gig means "alternative work arrangement" i can see it; less so if gig is "contingent work"
if gig means "alternative work arrangement" i can see it; less so if gig is "contingent work"
December 19, 2025 at 9:55 PM
it's true that employment in taxis (so, ride-shares) has grown way faster than overall employment, but it's not clear to me why someone who works full-time as an Uber driver is a "gig" worker.
if gig means "alternative work arrangement" i can see it; less so if gig is "contingent work"
if gig means "alternative work arrangement" i can see it; less so if gig is "contingent work"
trying to find trashy romance novels for my book club and goodreads is recommending Anna Karenina right between Isn't It Bromantic and You Had Me at Hola (Primas of Power, #1)
December 18, 2025 at 7:19 AM
trying to find trashy romance novels for my book club and goodreads is recommending Anna Karenina right between Isn't It Bromantic and You Had Me at Hola (Primas of Power, #1)
> never join threads
> try to click on threads link
> accidentally try to join threads
> account supended
> never joining threads
> try to click on threads link
> accidentally try to join threads
> account supended
> never joining threads
December 16, 2025 at 5:48 AM
> never join threads
> try to click on threads link
> accidentally try to join threads
> account supended
> never joining threads
> try to click on threads link
> accidentally try to join threads
> account supended
> never joining threads
also, lots of people are calling this a natural experiment. no!!!!
whatever regression you run is going to be horribly contaminated by spillover effects, both for rent prices and permits. in 2026 we shall teach the world about SUTVA
whatever regression you run is going to be horribly contaminated by spillover effects, both for rent prices and permits. in 2026 we shall teach the world about SUTVA
WSJ has a piece on Saint Paul's 2022 rent stabilization ordinance, concluding it killed production.
To be clear, I think the ordinance, especially as it was originally written, is bad, but if you actually plot the permit data, they are, in my professional opinion, noisy as shit.
To be clear, I think the ordinance, especially as it was originally written, is bad, but if you actually plot the permit data, they are, in my professional opinion, noisy as shit.
December 15, 2025 at 7:09 PM
also, lots of people are calling this a natural experiment. no!!!!
whatever regression you run is going to be horribly contaminated by spillover effects, both for rent prices and permits. in 2026 we shall teach the world about SUTVA
whatever regression you run is going to be horribly contaminated by spillover effects, both for rent prices and permits. in 2026 we shall teach the world about SUTVA
San Francisco loves to do the "we're a small working class city" shtick anytime anyone proposes completely normal apartment complexes.
As such, this will be a thread of my favorite charts that show SF is incredibly rich and this charade is insulting
First, a map of incomes in SF vs Oakland"
As such, this will be a thread of my favorite charts that show SF is incredibly rich and this charade is insulting
First, a map of incomes in SF vs Oakland"
December 15, 2025 at 6:40 PM
San Francisco loves to do the "we're a small working class city" shtick anytime anyone proposes completely normal apartment complexes.
As such, this will be a thread of my favorite charts that show SF is incredibly rich and this charade is insulting
First, a map of incomes in SF vs Oakland"
As such, this will be a thread of my favorite charts that show SF is incredibly rich and this charade is insulting
First, a map of incomes in SF vs Oakland"
WSJ has a piece on Saint Paul's 2022 rent stabilization ordinance, concluding it killed production.
To be clear, I think the ordinance, especially as it was originally written, is bad, but if you actually plot the permit data, they are, in my professional opinion, noisy as shit.
To be clear, I think the ordinance, especially as it was originally written, is bad, but if you actually plot the permit data, they are, in my professional opinion, noisy as shit.
December 15, 2025 at 5:12 PM
WSJ has a piece on Saint Paul's 2022 rent stabilization ordinance, concluding it killed production.
To be clear, I think the ordinance, especially as it was originally written, is bad, but if you actually plot the permit data, they are, in my professional opinion, noisy as shit.
To be clear, I think the ordinance, especially as it was originally written, is bad, but if you actually plot the permit data, they are, in my professional opinion, noisy as shit.
Given how easy it is to estimate upward sloping demand curves that fit the data really well, I'm kind of surprised at the "AI will / can solve causal inference" takes (on twitter, mostly).
E.g., a regression of changes in home values on changes in supply gets an R2 of 0.88 and the wrong sign!
E.g., a regression of changes in home values on changes in supply gets an R2 of 0.88 and the wrong sign!
December 11, 2025 at 5:10 PM
Given how easy it is to estimate upward sloping demand curves that fit the data really well, I'm kind of surprised at the "AI will / can solve causal inference" takes (on twitter, mostly).
E.g., a regression of changes in home values on changes in supply gets an R2 of 0.88 and the wrong sign!
E.g., a regression of changes in home values on changes in supply gets an R2 of 0.88 and the wrong sign!
The Brian caplan “college wage premium is all signaling” stuff is always funny because there are a number of papers that show, pretty definitively, there are substantial returns to education… but only if you assume labor markets are (close to) perfectly competitive
Unclear if this satisfies Brian!
Unclear if this satisfies Brian!
December 11, 2025 at 2:52 AM
The Brian caplan “college wage premium is all signaling” stuff is always funny because there are a number of papers that show, pretty definitively, there are substantial returns to education… but only if you assume labor markets are (close to) perfectly competitive
Unclear if this satisfies Brian!
Unclear if this satisfies Brian!
Everyone wants to ban corporate ownership of housing until this guy is your landlord
www.nytimes.com/2025/12/09/u...
www.nytimes.com/2025/12/09/u...
December 9, 2025 at 8:31 PM
Everyone wants to ban corporate ownership of housing until this guy is your landlord
www.nytimes.com/2025/12/09/u...
www.nytimes.com/2025/12/09/u...
i'll do a longer thread on this because these articles get a lot of things wrong, but the fact that the share of properties owned by corporations has increased is almost entirely due to the fact that it's way easier to own through an LLC now.
the underlying asset ownership is much more durable
the underlying asset ownership is much more durable
"Private-equity might not be causing the housing crisis, but corporate owners could end up making it a lot worse—for everyone," @annielowrey.bsky.social argues:
Private Equity Is America’s New Landlord
Is there someone to blame for the housing crisis?
bit.ly
December 9, 2025 at 1:53 AM
i'll do a longer thread on this because these articles get a lot of things wrong, but the fact that the share of properties owned by corporations has increased is almost entirely due to the fact that it's way easier to own through an LLC now.
the underlying asset ownership is much more durable
the underlying asset ownership is much more durable
is there a high quality lit review on inclusionary zoning? I'm skim reading some papers and a bunch have kind of bad statistical errors...
there's a lot of interpreting noisy point estimates as "no statistical relationship" and doing goofy things like not taking logs of housing starts
there's a lot of interpreting noisy point estimates as "no statistical relationship" and doing goofy things like not taking logs of housing starts
December 8, 2025 at 1:24 AM
is there a high quality lit review on inclusionary zoning? I'm skim reading some papers and a bunch have kind of bad statistical errors...
there's a lot of interpreting noisy point estimates as "no statistical relationship" and doing goofy things like not taking logs of housing starts
there's a lot of interpreting noisy point estimates as "no statistical relationship" and doing goofy things like not taking logs of housing starts
A bonus on voucher lease up rates: They suck and they suck more in tight markets.
Something like 50% of households that get vouchers can’t use them, either because their search failed or because the waitlist was so long by the time they got their name called they had left the metro
Something like 50% of households that get vouchers can’t use them, either because their search failed or because the waitlist was so long by the time they got their name called they had left the metro
I’ll add this article by @salimfurth.bsky.social which I liked a lot.
To YIMBYs, the low prices of Houston are maybe not the story. Instead, maybe it’s the extra bedrooms and surplus or ready-to-move-in apartments that matters (plus cheap rent == more vouchers)
worksinprogress.co/issue/why-ho...
To YIMBYs, the low prices of Houston are maybe not the story. Instead, maybe it’s the extra bedrooms and surplus or ready-to-move-in apartments that matters (plus cheap rent == more vouchers)
worksinprogress.co/issue/why-ho...
Why housing shortages cause homelessness
Why do high-cost cities have more homelessness? It's not just about rents — it’s also about the rooms friends and family can’t afford to share.
worksinprogress.co
December 6, 2025 at 4:20 PM
A bonus on voucher lease up rates: They suck and they suck more in tight markets.
Something like 50% of households that get vouchers can’t use them, either because their search failed or because the waitlist was so long by the time they got their name called they had left the metro
Something like 50% of households that get vouchers can’t use them, either because their search failed or because the waitlist was so long by the time they got their name called they had left the metro
Some scattered thoughts on homelessness:
People get tripped up thinking about homelessness as a stock of people. It’s not! It’s *very* transient; most people go in and out of homeslessness.
This ends up being very important for designing good policy
People get tripped up thinking about homelessness as a stock of people. It’s not! It’s *very* transient; most people go in and out of homeslessness.
This ends up being very important for designing good policy
Those homeless numbers come from the cost of giving every homeless person a housing voucher.
Which would definitely help a lot!
But it relies on their being a healthy vacancy rate to accommodate the flow of people and “healthy vacancy rate” doesn’t describe most places with high homelessness
Which would definitely help a lot!
But it relies on their being a healthy vacancy rate to accommodate the flow of people and “healthy vacancy rate” doesn’t describe most places with high homelessness
December 6, 2025 at 3:45 PM
Some scattered thoughts on homelessness:
People get tripped up thinking about homelessness as a stock of people. It’s not! It’s *very* transient; most people go in and out of homeslessness.
This ends up being very important for designing good policy
People get tripped up thinking about homelessness as a stock of people. It’s not! It’s *very* transient; most people go in and out of homeslessness.
This ends up being very important for designing good policy