iiinova — where your digital self keeps evolving

Joe Fish @sadbusdriver.bsky.social · 11/10/2026

i think these net worth breakdowns are kind of dumb tbh. there's two questions here for the under 35 group: 1. why did wealth skyrocket from 2019-2022? (not shown by WSJ) 2. why did wealth decline from 2022-2025 turns out it's a mix of changes in homeownership, used car prices, and house prices

142

Joe Fish @sadbusdriver.bsky.social · 08/10/2026

From a guy I follow on instagram. This would be why the “tax on applications” makes sense

250

Joe Fish @sadbusdriver.bsky.social · 07/10/2026

yeah, i could probably come up with some story but the time series seems reasonably convincing. although a bit odd the saint paul is back up to where MPLS is but this could be: 1. noise 2. MN data quirks since the data come via the courts and some stuff gets sealed (would need to ask elab people)

000

Joe Fish @sadbusdriver.bsky.social · 07/10/2026

you can kind of squint and see a "rent effect" although it's how believable it seems contigent on what the base period is.

100

Joe Fish @sadbusdriver.bsky.social · 06/10/2026

this is the UI on the case management software Massachusetts prosecutors use nasser.blog/dial-up-just...

2110

Joe Fish @sadbusdriver.bsky.social · 05/10/2026

Even the industry publications don’t really go to bat for them: “They graduate at higher rates and with less debt” (note nothing about earnings) And “there are some schools that need reform” And “our students make more than the min wage and general labors”

100

Joe Fish @sadbusdriver.bsky.social · 05/10/2026

just heads up, this series is bad for your excercise. the bottom 10% are generally not in the labor force so income from wages and salaries isn't a super helpful measure. you'd want like total income bottom 10%. doesnt change the story too much though

120

Joe Fish @sadbusdriver.bsky.social · 01/10/2026

the article does both public and private comparisons.

060

Joe Fish @sadbusdriver.bsky.social · 30/09/2026

www.nber.org/papers/w34072 assuming youve seen this one? quasi-experimental. as an aside, i've kind of thought LLMs would be great for low income countries where a barrier is training on like "how should i do inventory" and bookkeeping best practices

120

Joe Fish @sadbusdriver.bsky.social · 29/09/2026

imo even this is conceding too much. the typical SF condo looks much more like ~800K than it does an 11 million. even 280 spear st, which is as luxury as it's gonna get, is cheaper than a SFH on average. and for the 11 million dollar penthouse, if someone wants to pay taxes on that be my guest.

220

Joe Fish @sadbusdriver.bsky.social · 28/09/2026

(some of this is composition but for all the hay about declining wages in 2022 ish, the trends are pretty striking. there's been like two periods of sustained wage growth in the past fourty years, and we're living in one of them)

2294

Joe Fish @sadbusdriver.bsky.social · 27/09/2026

"everyone should strive to be economists" is what im hearing (or at least the economist keynes would have wanted). and by "the economist keynes would i wanted" I, of course, mean a dentist

250

Joe Fish @sadbusdriver.bsky.social · 26/09/2026

it also like doesnt fit the data particularly well? what's probably happening are egg prices crashed and so the model says sentiment should recover, except it hasnt because people dont actually care much about egg prices.

250

Joe Fish @sadbusdriver.bsky.social · 25/09/2026

i dont think so. im (mentally) eyeballing the cross-country correlation between change in labor share and "more far right" and there seems to be approximately nothing.

140

Joe Fish @sadbusdriver.bsky.social · 25/09/2026

the comp i find helpful is US vs Norway and Finland. Finland and Norway have lower or comprarable shares and have seen (generally) larger declines

2432

Joe Fish @sadbusdriver.bsky.social · 24/09/2026

Here's another version: Because of the glory of the home mortgage disclosure act, we can see the income of homebuyers with mortgages. Within the same county/neighborhood/neighborhood and adjusting for unit characteristics, condo buyers are poorer than single family home buyers. it's a starter home!

210420

Joe Fish @sadbusdriver.bsky.social · 23/09/2026

this was inspired by strongtown's weird anti condo fixation. condos *are* starter homes -- especially in places like san francisco where a 1200 square foot single family home will easily fetch 1.5 million. condos fit *so well* into their "just make it legal* ethos in every way except the vibes

1713

Joe Fish @sadbusdriver.bsky.social · 23/09/2026

the thing with luxury condos is that they're mostly a myth. 1. within the same county, condos sell for 35% less than single family homes 2. within the same neighborhood, they sell for 43% less 3. within the same neighborhood and adjusting for some quality characteristics, they sell for 21% less

717833

Joe Fish @sadbusdriver.bsky.social · 23/09/2026

a bunch of states are toying with the idea of charging companies fees based on the number of employees they have on medicaid. including washington, which wants to do a 7k / employee tax. seems like the response would be "dont hire them" www.healthaffairs.org/content/fore...

040

Joe Fish @sadbusdriver.bsky.social · 23/09/2026

the main point in the paper is that the market works better with fees/quotas. but the data are project level which makes it a bit hard to map to worker welfare highlighted rows show increased payment probability, higher fill, lower asking price. but this also reflects what kinds of jobs are posted

110

Joe Fish @sadbusdriver.bsky.social · 23/09/2026

also there's a paper that looks at job application fees on congestion directly. pubsonline.informs.org/doi/10.1287/...

4111

Joe Fish @sadbusdriver.bsky.social · 21/09/2026

a rule of thumb is that by most academic definitions of gentrification, something like half of neighborhoods will be gentrifiable over a 15-20 year time horizon, somewhere around 10-20% of those neighborhoods will end up gentrifying. so like 5-10% of neighborhoods. you can see this in, say, philly

130

Joe Fish @sadbusdriver.bsky.social · 21/09/2026

if you look at US-born New Yorkers and ask where they're from, surprise, surprise the answer is: 1. New York State (77%) 2. Near New York City (8%) 3. Everywhere else (15%)

1110

Joe Fish @sadbusdriver.bsky.social · 21/09/2026

the whole NYC transplants discourse is so dumb. for one, new york city -- famously a city of transplants! for two, if you're asking "how many people who live in NYC were born in the US and not in New York State" it's like 14%.

1191

Joe Fish @sadbusdriver.bsky.social · 18/09/2026

might do ludwig institute's next. they're probably the worst offender for "man this index sucks" eg in their section comparing their housing inflation measure to CPI they: 1. swapped the two series in their chart (blue is CPI; orange is TLC) 2. their data are still wrong bsky.app/profile/sadb...

120

Joe Fish @sadbusdriver.bsky.social · 18/09/2026

ALICE is still a marked improvement on what the Ludwig Institute puts out, however. Ludwig finally "fixed" their misunderstanding of the housing component of CPI, but in the most misleading way possible However, Ludwig's kept the errors in their underlying rent inflation data and their website

020

Joe Fish @sadbusdriver.bsky.social · 18/09/2026

clothing is an obvious omission in the index. it's an example of something whose price has been driven so far down that we dont even really conceive of it like an essential even though it is.

150

Joe Fish @sadbusdriver.bsky.social · 18/09/2026

kind of telling re vibecession discourse is that i've seen approximately nobody FRED-posting about how inflation adjusted incomes hit all time highs last year fred.stlouisfed.org/series/mepai...

1151

Joe Fish @sadbusdriver.bsky.social · 14/09/2026

So then the issue is for the 7,8,9 deciles, you can kinda draw a straight line from 2019 to 2023. They might address this! I’ve barely skimmed the paper. But treatment onset is 2022. Alternatively:

010

Joe Fish @sadbusdriver.bsky.social · 09/09/2026

not sure? the roomba company went bankrupt, but idk enough about that to be sure whether it's a technology issue or a financing issue or a product demand issue or ...

230

Joe Fish @sadbusdriver.bsky.social · 09/09/2026

i guess housekeeping robots haven't really become a thing? self driving cars and conversational companions definitely are here. so article is 1/3 on predictions. complexity science -- sometimes it's hard to predict the rate of progress!

2200

Joe Fish @sadbusdriver.bsky.social · 08/09/2026

this is one of those where the fact that chatgpt release lines up so well with "information" payrolls should count as evidence *against* chatgpt being the cause of the decline. chatgpt sucked at release and even now AI usage is at like half of information firms. alternatively, interest rate hikes

020

Joe Fish @sadbusdriver.bsky.social · 07/09/2026

fwiw, income growth has been pretty U-shaped more than anything else

110

Joe Fish @sadbusdriver.bsky.social · 07/09/2026

obviously income inequality has grown a lot and this is bad but it seems very weird that NBC is reporting a stock return number that is not adjusted for inflation and comparing it to a wages number that is.

051

Joe Fish @sadbusdriver.bsky.social · 07/09/2026

behold, the soon to be expanded Wharton San Francisco campus! Previously, they occupied the sixth floor of an office building along the embarcadero

020

Joe Fish @sadbusdriver.bsky.social · 03/09/2026

what's funny is there's an old integralds BE comment arguing that the labor share of income has been approximately 2/3 for 5000 years sourcebooks.web.fordham.edu/ancient/meso...

120

Joe Fish @sadbusdriver.bsky.social · 30/08/2026

the author i think basically understands all this (although the occupancy part is, imo, underplayed) but i don't think there's a correction, per se.

110

Joe Fish @sadbusdriver.bsky.social · 30/08/2026

if you narrow in on specific markets, the effects are larger in poorer and less white areas

1132

Joe Fish @sadbusdriver.bsky.social · 30/08/2026

the basic comparison in the paper is to take properties within pretty narrow submarkets and within pretty narrow building types and ask whether the rent stabilized units start doing less maintennce after the law was passed. answer seems like "yes" across a wide variety of outcomes

1132

Joe Fish @sadbusdriver.bsky.social · 30/08/2026

working paper arguing that the 2019 NYC strengthening of rent control lead to substantially less property maintence and lower housing quality. beaubressler.github.io/papers/rent_...

34110

Joe Fish @sadbusdriver.bsky.social · 25/08/2026

imagine that the code changes do nothing but increase costs. if new and old houses are perfect substitutes, this doesn't necessarily change relative prices, but it increases overall prices. because they're looking within markets, they'd find nothing. also their regression equation looks wrong

100

Joe Fish @sadbusdriver.bsky.social · 25/08/2026

wild that a tenure-track professor position at san jose state pays about as much (or less) than a teacher in san jose unified have to imagine if their search works the econ job market is in rough shape

280

Joe Fish @sadbusdriver.bsky.social · 20/08/2026

when you ask "how much did san francisco permit during it's third gold rush (2011 tech boom)?", the answer is: 1. more than it had the decade prior! yay! 2. way, way, way less than seattle 3. about as much as bakersfield unfortunately, it's doing even worse during it's fourth gold rush (now)

1190

Joe Fish @sadbusdriver.bsky.social · 20/08/2026

that era of RHNA was fun because they used past population growth as a measure of current housing need. so cities that built got told to build more and NIMBY cities got told "great job, sweetie!" lead to beverley hills being one of the "success" stories because their requirement was 3 units

1150

Joe Fish @sadbusdriver.bsky.social · 20/08/2026

Re the eternal traffic camera discourse, china experimented with adding traffic violations to people's credit scores, which are then used / reported in job applications and residency permits. this was much better at deterring speeding than monetary fines www.sciencedirect.com/science/arti...

020

Joe Fish @sadbusdriver.bsky.social · 20/08/2026

if you plot out rent to income ratios for different cities over time it quickly becomes very obvious how useless they are. Wow San Francisco must have improved their housing polict so much starting in 2011! Home price to income is a bit better but not my much.

1131

Joe Fish @sadbusdriver.bsky.social · 19/08/2026

this is "illuminated" (sorry couldn't resist) in this paper on the history of the price of light. but "the quality of modern life is generally much higher than modern statistics imply" is generally not what people argue with matt about digamoo.free.fr/nordhaus96.pdf

051