iiinova — where your digital self keeps evolving

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

We talk about the possibility of means testing later, like clawing it back from ineligible as part of income tax filing, so delivery of payments is still as fast as possible.

000

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

In the long run, preventative investments or rezoning or things like that could bring down the overall cost and you wouldn't need to target. But that's the long run.

010

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

In the sense that high income people might but support a program they don't qualify for? Potentially, yeah. But most states can't afford a program like this on their own, which is the long-standing policy debate. so the only thing to do in the short run is cut generosity or cut eligibility.

110

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

We also had some Exogenous Constraints to look only at a single support program...

010

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

Two things. First, we did look at unemployment generally, but it was extremely noisy. admin delays in UI delivery creates noise in timing of receipt, and not all UI comes via direct deposit. Besides UI, job separation only shows up as a lack of paycheck, which is the metric we used.

110

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

Making high-inc people ineligible can significantly reduce the program budget w/o leaving people unable to buy necessities. But people need money fast, and adding red tape to the application could slow that down.

000

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

Policymakers and wonks have long discussed moving more of the fiscal burden of these programs to states to help spur preventative investment. But most states would have trouble funding these programs. Can income targeting reduce the budget w/o leaving people vulnerable? Maybe...

100

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

And people need that money fast for home repairs, replacing ruined groceries and other essentials. People in the disaster area got a $500 relief payment. High-inc people had a rush of spending as soon as the storm was over, while low-inc had to wait until they got their payment.

131

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

If you don't have emergency savings, you might be able to use a credit card. But we also see that low-income families had income disruptions for almost two months after the storm, making it harder to pay back any credit used.

120

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

When Hurricane Ida hit Louisiana in 2021, just over half of low-income families in the area had enough emergency savings to cover a $500 expense. Direct relief payments made a huge difference! Full report here:

120

Daniel M. Sullivan @dsullivan.bsky.social · 27/05/2026

New Report from my team at the JPMorganChase Institute! Targeting disaster relief: Who needs aid to weather the storm? Lots of people don't have an emergency fund to cover disaster expenses AND many have their incomes disrupted! Who needs aid? Report: www.jpmorganchase.com/institute/al...

3166

Daniel M. Sullivan @dsullivan.bsky.social · 14/04/2026

Very excited to see the results! I know a lot of people (me) who want to know more about how people think about long term IDR forgiveness and the spectre of growing balances.

210

Daniel M. Sullivan @dsullivan.bsky.social · 08/12/2025

🚨 We're hiring at the JPMorganChase Institute! My Consumer Finance team is looking for a research associate. It's a flexible role, and we're considering both experienced RA's/pre-docs and rookie PhDs/ABDs. Please share widely! jpmc.fa.oraclecloud.com/hcmUI/Candid...

jpmc.fa.oraclecloud.com

JPMorganChase Institute Research Associate

Collaborate with a dedicated team of data scientists, economists, and industry and policy experts to conduct unique research using proprietary data.

020

Daniel M. Sullivan @dsullivan.bsky.social · 10/11/2024

I took a class in ugrad freshman year and I still think about it regularly. Allusions and wordplay people have been unraveling for hundreds of years, an absolute next level work.

010

Daniel M. Sullivan @dsullivan.bsky.social · 03/10/2024

I guess I'm gonna have to write something about auto debt...

110

Daniel M. Sullivan @dsullivan.bsky.social · 03/10/2024

I would think so. Cursory search for what fraction of auto loans are floating didn't find anything. My prior is that it's mostly fixed but that's a guess.

210

Daniel M. Sullivan @dsullivan.bsky.social · 03/10/2024

I happened on this today while looking for something else, from the Other Institute. Looks like the spike in prices subsided as rates rose (probably some causality there). But median and average payments are not up much more than inflation. NB: some of this throws out people who miss any payments

111

Daniel M. Sullivan @dsullivan.bsky.social · 03/10/2024

Which it was, until mid '22 when auto rates started to rise. But IDK if enough people are savvy enough to drive a trend like this, though.

110

Daniel M. Sullivan @dsullivan.bsky.social · 03/10/2024

The "rolling negative equity" Q is interesting, and something we might be able to do with our data. there was also about a year between inflation spiking (partly b/c car buying) and rates rising. You could get 1.99% until ~Jan 2022. I would (and did) take every dollar possible under those terms.

110

Daniel M. Sullivan @dsullivan.bsky.social · 03/10/2024

Yes, that too. Unfortunately, idk if we'll ever have admin data to answer this. I'm pretty sure make/model/price isn't in bureau data, and no bank is going to reveal how underwater its own loans are.

210

Daniel M. Sullivan @dsullivan.bsky.social · 03/10/2024

Delinquencies aren't too crazy yet. Credit got pretty loose in '21-'22, so the overall credit pool is more risky and it's an open question how much of recent credit trends (delinq., underwater) are from that.

140

Daniel M. Sullivan @dsullivan.bsky.social · 03/10/2024

I got to a point once with a very good RA where we communicated almost exclusively via Trello, and the occasional phone call as needed to talk through stuff. I could track exactly what he was doing real time, and he knew exactly what to do next. It was amazing for both of us.

040

Daniel M. Sullivan @dsullivan.bsky.social · 27/09/2024

it's still the shared version and allows simultaneous editing, but in the standard Word UI. But it's possible it's not a universal feature

010

Daniel M. Sullivan @dsullivan.bsky.social · 27/09/2024

We have a "open in desktop app by default" option. Don't know if that's standard.

110

Daniel M. Sullivan @dsullivan.bsky.social · 23/08/2024

🚨We're hiring at the JPMorganChase Institute!🚨 Roles are similar to "pre-docs" but our analysts and associates tend to stay longer and you get to pick up a lot of very valuable Big Data experience. The pay is also Quite Good. Please share! jpmc.fa.oraclecloud.com/hcmUI/Candid...

jpmc.fa.oraclecloud.com

JP Morgan Institute - Research Associate

Collaborate with a team of scientists to implement robust data pipeline that realizes the full potential of the firm's data.

195

Reposted by Daniel M. Sullivan

Eric Roston @eroston.bsky.social · 07/03/2024

As of next week, the oceans will have set a rather preposterous heat record every day for a year.

ClimateReanalyzer daily sea surface temperature chart for all years, showing absolutely bonkers daily records beginning in mid-March 2023.

39472

Daniel M. Sullivan @dsullivan.bsky.social · 12/02/2024

We're hiring at the JPMorgan Chase Institute! We're looking for someone to lead our research on local economies/urban econ and small businesses. For background, this is the same type role I had when I joined the Institute in 2019. Please share! jpmc.fa.oraclecloud.com/hcmUI/Candid...

jpmc.fa.oraclecloud.com

JP Morgan Chase Institute Research Lead - Small Business and Local Economies

Lead research related to small businesses and local economies in which they operate, while leading a small team of Data Scientists.

031

Reposted by Daniel M. Sullivan

Ed Hawkins @edhawkins.org · 09/01/2024

The C3S/ECMWF team announce that 2023 was hottest on record, about 1.48°C above pre-industrial levels. "2023 was an exceptional year, with climate records tumbling like dominoes," says @oceanterra.bsky.social Great graphics in this BBC article about the data: www.bbc.co.uk/news/science...

59070

Reposted by Daniel M. Sullivan

Jonathan A. Parker @profjaparker.bsky.social · 14/11/2023

Joanne Im on the PhD job market from MIT, has important results on finance and the environment (and a more recent abstract that I last sent out) mitsloan.mit.edu/phd/students...

1106

Reposted by Daniel M. Sullivan

Daniel M. Sullivan @dsullivan.bsky.social · 06/11/2023

Hello, #EconSky, we're hiring at the JPMC Institute! Come lead our research on mid-size business! Early career PhD's, including rookie JMC's, are a great fit, no prior middle market exp required. Please reach out w/ Q's about the job, Institute, etc! jpmc.fa.oraclecloud.com/hcmUI/Candid...

086

Daniel M. Sullivan @dsullivan.bsky.social · 06/11/2023

086

Reposted by Daniel M. Sullivan

Jean-François Bonnefon @jfbonnefon.bsky.social · 01/10/2023

When the abstract says ❝We exploit the quasi-exogenous assignment of 1.5 million patients to physicians in US military, using the difference in ranks to measure their power differential❞ you know you're in for something great @vinisingh.bsky.social stschwab.github.io/images/power...

Abstract of the paper 'How Power Shapes Behavior: Evidence from Physicians'

26739