Joe Fish

if it's just regex / relatively simple filtering id just use the command line. will be faster w/ minimal overhead and no need to set up an sqlite database or anything

there's also a fun one where cities are future/past versions of other cities.

e.g., atlanta is los angeles from 20 years ago

yeah, i think that gets most of it. Otherwise, you'd see a lot more price drops when there are positive demand shocks, but I can't think of any examples of that

"Supply curves slope upwards: evidence from labubu" - forthcoming.

for anything from netflix to a raw denim company, the acquisition costs are absolutely convex.

also, the survey research people use to justify flat or declining marginal cost curves generally ask about 'unit' cost curves, which explicitly abstract from customer acquisition as a cost

niche subtweet:

aside from people confusing short / long run cost curves, isn't an easy way to get upward sloping cost curves to just add "acquisition cost" to the firm's cost function?

ETA until he tries to parlay this into a “breakup UPS” think piece?

Another minor point that they fuck up is that the “healthy” vacancy rate they mention is the *rental* vacancy rate which they then compare to the *overall* vacancy rate

also, let me take a big sip of coffee and look at what the median income is in these west side areas and compare that to Oakland

The "I'm a small bean" energy by San Francisco is incredibly insulting. Of major cities, San Francisco has the second-highest percentage of rich nhoods and the fourth-lowest percentage of low/middle income nhoods.

How can we talk about gentrification when there's like two places left to gentrify?

my contribution to @besttrousers.bsky.social discourse:

1) census contract rents in Brooklyn make sense given observed incomes
2) zillow asking rents give implausible implied levels of rent burdens
3) zillow does way worse at explaining rent:income in high immigrant counties; census does not

a related point is that most new apartments are >200+ units and cost > 50 million to build.

developers don't like to hold on to these and so try to sell them once they lease up. but to sell, you need somebody to buy. who else but large public/private investors has the money to do this? Local gov?

be cool if pe stakeholder published their data so i could check this, but isnt this just a map of which places have built a bunch of new apartments? (relative to the size of their existing housing stock)

Agreeing w/ you tbc, but for other people:

1) RealPage promises users a 3-7% revenue lift
2) estimates of "breakup the large landlords" policies find large effects (~10%) *but only in a very small number of markets*
3) most markets aren't concentrated

bsky.app/profile/sadb...

the issue is that even within year X building type X zip code bins, there will be large differences in prices.

Some units won't be maintained well, some units will be in high crime parts of the neighborhood, etc. Those units will be both lower price and less likely to enter zillow's data

not to say there aren't issues with the survey data (sweetheart deals, rent stabilized units, etc), but the "zillow says rents are X" reports people like to post have selection issues.

E.g., If you read zillow's methodology, they lean really hard on year built + apartment size to capture quality

i made this point on the other site, but those commercial rent data that people cite for "median rental in NYC" have bad coverage of low income areas.

If you look at coverage rates or compare commercial rent averages vs contract rents from eviction filings, you see huge divergences

Per usual, people are bad at reading these articles

1) investors are defined as bought via an LLC, LP, etc. Per the article, institutional investors are net sellers
2) there's no obvious spatial correlation between % investor-owned and prices or price increases

Also, the "break up big landlords" isn't really a viable affordability tool, although it can work in very specific markets

(usually markets with lots of high rises in neighborhoods with few substitutes -- think location specific amenities like great views or beach access)
bsky.app/profile/sadb...

Some papers have looked explicitly at the "break up big landlords" angle. Breaking up big landlords can lead to pretty large price declines! Like ~10%! *But only in a very small number of markets* (about 5% or so)

It's not a general affordability tool
www.dropbox.com/scl/fi/jny7m...

the other thing is RealPage quotes like a 3-7% revenue lift depending on the pitch deck they use. Which is fairly big as far as effects in the academic rental market literature go, but also clearly in the "helpful but not sufficient" bucket

Only one of those prizes was funded by dynamite blood money and it ain’t the Swedish central bank one

Preaching to the choir, but homelessness is mostly a "prices are too high problem"; high prices are mostly a "vacancies are too low problem"; ergo, homelessness is mostly a "there aren't enough vacant homes problem"

pbs.twimg.com/media/GI_Pbl...

Eg if you look at office costs
As a rough approximate for differences in costs for big multi family projects there’s huge implied room for reductions without even touching any labor costs

Kinda depends on what levers you pull. If California waved a wand and got European (or even Texan) costs for multi family you’d probably need prices to fall like 20-50% before projects stopped penciling

Developers don’t internalize whether their projects will push down other landlords rents

If you waved a wand and upzoned a bunch of land, a lot of projects would pencil, developers would build them, and prices of other landlords’ units would fall

apropos of nothing, it's kind of annoying how the flavor-of-the-month criticism of economic methodology is so incurious about why mainstream economists do things the way they do

armchair sociology and aversion to qual research apparently affects economists of all methodological persuasions

Price discrimination gets you some of the way there, I guess

shot in the dark: does anyone know of models that get market power where the market power doesn't come from

1) quantity restrictions
2) bargaining

Motivated by Ticketmaster, which seems to have market power and doesn't seem to obviously restrict quantities or bargain

In general though,
1) the correlation between high rents and high homelessness is real and is a long term phenomenon
2) short run fluctuations can be because of rent hikes, but they're also likely to be kind of random blips / unrelated to rent trends (see all the discourse on the 2024 increases)

Interestingly, you can kind of square the circle by noting that increases in rents do predict increases in homelessness, but only for cities that were already expensive.

It's a lot worse for rent hikes to hit cities with no slack, than to hit relatively affordable ones