Justin Perry
@justinperry.bsky.social
Reposted by Justin Perry
How did this get past the IRB?!?
April 3, 2025 at 9:40 AM
How did this get past the IRB?!?
In the 50s or 60s an economist built a physical model of the circular flow in macroeconomics. Anyone know who that was and how to find pictures of his model?
February 20, 2025 at 6:57 PM
In the 50s or 60s an economist built a physical model of the circular flow in macroeconomics. Anyone know who that was and how to find pictures of his model?
I like all of my skeets for the same reason that I use the log(x+1) transformation - excessive zeros
November 26, 2024 at 2:25 AM
I like all of my skeets for the same reason that I use the log(x+1) transformation - excessive zeros
Reposted by Justin Perry
Today's edition of "📈📉 papers you should know": a finite asymmetric information game in which *every* sequential equilibrium (in fact, every PBE) has a player convinced of a state and subsequently relinquishing that belief. Let's go through the example (changing language a bit).
November 21, 2024 at 11:27 AM
Today's edition of "📈📉 papers you should know": a finite asymmetric information game in which *every* sequential equilibrium (in fact, every PBE) has a player convinced of a state and subsequently relinquishing that belief. Let's go through the example (changing language a bit).
Reposted by Justin Perry
Today's edition of "📈📉 papers you should know" is Dezsö Szalay's @reveconstudies.bsky.social paper, "The Economics of Clear Advice and Extreme Options". The paper concerns itself with information acquisition choices under delegated decision making. Let's go through an example...
November 20, 2024 at 11:09 AM
Today's edition of "📈📉 papers you should know" is Dezsö Szalay's @reveconstudies.bsky.social paper, "The Economics of Clear Advice and Extreme Options". The paper concerns itself with information acquisition choices under delegated decision making. Let's go through an example...
Reposted by Justin Perry
How it started: "AI is going to replace randomized clinical trials"
How it is going:
ai.nejm.org/doi/10.1056/...
How it is going:
ai.nejm.org/doi/10.1056/...
We Need More Randomized Clinical Trials of AI
In the first prospective clinical trial of artificial intelligence (AI) assistance
in stress echocardiography, there was no difference in diagnostic accuracy between
AI assistance and standard-of-c...
ai.nejm.org
November 20, 2024 at 5:08 PM
How it started: "AI is going to replace randomized clinical trials"
How it is going:
ai.nejm.org/doi/10.1056/...
How it is going:
ai.nejm.org/doi/10.1056/...
Reposted by Justin Perry
A very interesting new working paper by the greats Ashesh Rambachan, Rahul Singh, and @vivianodavide.bsky.social: arxiv.org/pdf/2411.10959
It seems that this is another area where the empirical common practice was "too fast,'' and econometrics is catching up!
Cool and empirically relevant stuff!
It seems that this is another area where the empirical common practice was "too fast,'' and econometrics is catching up!
Cool and empirically relevant stuff!
November 19, 2024 at 3:50 PM
A very interesting new working paper by the greats Ashesh Rambachan, Rahul Singh, and @vivianodavide.bsky.social: arxiv.org/pdf/2411.10959
It seems that this is another area where the empirical common practice was "too fast,'' and econometrics is catching up!
Cool and empirically relevant stuff!
It seems that this is another area where the empirical common practice was "too fast,'' and econometrics is catching up!
Cool and empirically relevant stuff!
Reposted by Justin Perry
Given the #Rstats base vs tidy wars have been reignited, I want to create a new front for #statsky: poisson vs. negative binomial regression.
I firmly believe you should almost never use NB. My reasoning: stats.stackexchange.com/a/653731/116...
And always use a robust/bootstrap SE with poisson!
I firmly believe you should almost never use NB. My reasoning: stats.stackexchange.com/a/653731/116...
And always use a robust/bootstrap SE with poisson!
When to use negative binomial and Poisson regression
When would one use a negative binomial regression and when would one use Poisson regression with respect to the mean and variance?
stats.stackexchange.com
November 18, 2024 at 5:40 PM
Given the #Rstats base vs tidy wars have been reignited, I want to create a new front for #statsky: poisson vs. negative binomial regression.
I firmly believe you should almost never use NB. My reasoning: stats.stackexchange.com/a/653731/116...
And always use a robust/bootstrap SE with poisson!
I firmly believe you should almost never use NB. My reasoning: stats.stackexchange.com/a/653731/116...
And always use a robust/bootstrap SE with poisson!
Reposted by Justin Perry
⁉️ DAG Quiz of the Day ⁉️
Find all conditional independencies between nodes implied by the following directed acyclic graph. (Hint: there is an easy way to rule out many potential dependencies.)
Find all conditional independencies between nodes implied by the following directed acyclic graph. (Hint: there is an easy way to rule out many potential dependencies.)
November 19, 2024 at 7:09 AM
⁉️ DAG Quiz of the Day ⁉️
Find all conditional independencies between nodes implied by the following directed acyclic graph. (Hint: there is an easy way to rule out many potential dependencies.)
Find all conditional independencies between nodes implied by the following directed acyclic graph. (Hint: there is an easy way to rule out many potential dependencies.)
Reposted by Justin Perry
The most obvious answer here is that you're seeing the effect of negative media coverage. But I also wonder if part of what's going on is that people see job growth as as a reflection of their state's economy, but prices as a national issue.
November 18, 2024 at 4:36 PM
The most obvious answer here is that you're seeing the effect of negative media coverage. But I also wonder if part of what's going on is that people see job growth as as a reflection of their state's economy, but prices as a national issue.
Bluesky almost to 20 million users. What’s the tipping point where the network effect starts working against the incumbent networks and begins to tip toward migration?
November 18, 2024 at 11:14 PM
Bluesky almost to 20 million users. What’s the tipping point where the network effect starts working against the incumbent networks and begins to tip toward migration?
Reposted by Justin Perry
This thread is why econtwitter is back.
Much of what we colloquially call “beliefs” are closer to preferences than to Bayesian beliefs
Aumann's agreement theorem says that two Bayesians who share a common prior and have common knowledge of each other's posteriors must share the same posterior --- i.e. they cannot "agree to disagree". But people disagree all the time. What is your favorite hypothesis about how come?
November 17, 2024 at 2:45 AM
This thread is why econtwitter is back.
What’s MGP’s future in the Democratic Party? Really interesting question
Washington 3 being one of the only seats in the country to swing toward Harris, and by more than any other district, is fascinating (and clearly a factor in Marie Gluesenkamp-Perez’s still-impressive 4-point win over Joe Kent).
Per calculations from @drewsav.bsky.social
Per calculations from @drewsav.bsky.social
November 17, 2024 at 12:39 AM
What’s MGP’s future in the Democratic Party? Really interesting question
Reposted by Justin Perry
Washington 3 being one of the only seats in the country to swing toward Harris, and by more than any other district, is fascinating (and clearly a factor in Marie Gluesenkamp-Perez’s still-impressive 4-point win over Joe Kent).
Per calculations from @drewsav.bsky.social
Per calculations from @drewsav.bsky.social
November 17, 2024 at 12:22 AM
Washington 3 being one of the only seats in the country to swing toward Harris, and by more than any other district, is fascinating (and clearly a factor in Marie Gluesenkamp-Perez’s still-impressive 4-point win over Joe Kent).
Per calculations from @drewsav.bsky.social
Per calculations from @drewsav.bsky.social
Reposted by Justin Perry
Paper worth reading: "The Optimality of Majority Rule" by Nina Bobkova.
It offers a new perspective on why simple majority-rule may be best: it not only aggregates information but it also motivates voters to learn about what really matters.
Link: drive.google.com/file/d/1pVVE...
#polecon
It offers a new perspective on why simple majority-rule may be best: it not only aggregates information but it also motivates voters to learn about what really matters.
Link: drive.google.com/file/d/1pVVE...
#polecon
November 16, 2024 at 5:44 PM
Paper worth reading: "The Optimality of Majority Rule" by Nina Bobkova.
It offers a new perspective on why simple majority-rule may be best: it not only aggregates information but it also motivates voters to learn about what really matters.
Link: drive.google.com/file/d/1pVVE...
#polecon
It offers a new perspective on why simple majority-rule may be best: it not only aggregates information but it also motivates voters to learn about what really matters.
Link: drive.google.com/file/d/1pVVE...
#polecon
Reposted by Justin Perry
this would fit into an ecological inference setup, right?
November 17, 2024 at 12:05 AM
this would fit into an ecological inference setup, right?
Does anyone on #econtwitter want to write a paper about optimal harmonization of data across different census boundaries?
November 17, 2024 at 12:15 AM
Does anyone on #econtwitter want to write a paper about optimal harmonization of data across different census boundaries?
Reposted by Justin Perry
I have an #econsky / #polisky question:
Has anyone implemented a partial identification approach to consider the relative vote preference for candidates, considering "no votes"?
E.g. A county votes 55% Trump, 45% harris, turnout is 60%. Bounds on Trump preference is [33% Trump, 73% Trump].
1/
Has anyone implemented a partial identification approach to consider the relative vote preference for candidates, considering "no votes"?
E.g. A county votes 55% Trump, 45% harris, turnout is 60%. Bounds on Trump preference is [33% Trump, 73% Trump].
1/
November 16, 2024 at 11:33 PM
Bluesky just crossed 18m users. Pretty incredible example of networks’ stability and instability in equilibrium
November 16, 2024 at 8:23 PM
Bluesky just crossed 18m users. Pretty incredible example of networks’ stability and instability in equilibrium
This place isn’t the same without shitting on Keith Rabois
November 16, 2024 at 8:19 PM
This place isn’t the same without shitting on Keith Rabois
Anyone have eyes on twitter’s user base right now? Gotta be pretty ugly for them over there in the X factor
November 16, 2024 at 8:00 PM
Anyone have eyes on twitter’s user base right now? Gotta be pretty ugly for them over there in the X factor
Reposted by Justin Perry
I wrote an intro to the literature on simple mechanisms. Out now in the JEP.
It’s a short (14-page) plain-language summary, designed to get students to the research frontier. 📈📉
pubs.aeaweb.org/doi/pdfplus/...
It’s a short (14-page) plain-language summary, designed to get students to the research frontier. 📈📉
pubs.aeaweb.org/doi/pdfplus/...
pubs.aeaweb.org
November 16, 2024 at 3:33 PM
I wrote an intro to the literature on simple mechanisms. Out now in the JEP.
It’s a short (14-page) plain-language summary, designed to get students to the research frontier. 📈📉
pubs.aeaweb.org/doi/pdfplus/...
It’s a short (14-page) plain-language summary, designed to get students to the research frontier. 📈📉
pubs.aeaweb.org/doi/pdfplus/...
Predictions for 2028:
(1) will be millennial vs millennial
(2) DEMs will nominate someone w/ “outsider cred”
(3) republicans will continue to be weirdos and get smoked
(1) will be millennial vs millennial
(2) DEMs will nominate someone w/ “outsider cred”
(3) republicans will continue to be weirdos and get smoked
November 16, 2024 at 2:25 PM
Predictions for 2028:
(1) will be millennial vs millennial
(2) DEMs will nominate someone w/ “outsider cred”
(3) republicans will continue to be weirdos and get smoked
(1) will be millennial vs millennial
(2) DEMs will nominate someone w/ “outsider cred”
(3) republicans will continue to be weirdos and get smoked
Rudy Giuliani, you’ve got terrible taste in watches man. This collection is a slap in the face to those election workers.
Rudy Giuliani’s watch collection being FedExed to GA election workers Ruby Freeman & Shaye Moss: “This is the accumulation of 60 years of hard work. Many of these watches hold great sentimental value!
November 16, 2024 at 3:23 AM
Rudy Giuliani, you’ve got terrible taste in watches man. This collection is a slap in the face to those election workers.