Atheendar Venkataramani's avatar

Atheendar Venkataramani

@atheendar.bsky.social

151 followers 149 following 63 posts

Internal medicine doc and health economist at the University of Pennsylvania.

Director, OpportunityForHealth.org.

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

My drawing "Studious" will be displayed as part of the Penn Medicine 2024-2025 Celebration of Art and Life. Feels better than getting a paper accepted!

1 replies 0 reposts 8 likes

Reposted by Atheendar Venkataramani

Julian Reif's avatar Julian Reif @julianreif.bsky.social
[ View ]

“real-time claims adjudication for physical health claims, similar to prescription drug claims adjudication… would reduce fluctuations in consumption decisions, improving household welfare to the extent that unexpected price shocks are eliminated.”

0 replies 3 reposts 3 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Thanks! Can't wait to read.

0 replies 0 reposts 1 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

This looks awesome and very very timely! Do you have the WP up anywhere (sadly can't be at ASHEcon this year).

1 replies 0 reposts 1 likes

Reposted by Atheendar Venkataramani

Reposted by Atheendar Venkataramani

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

My goal is to write an introductory paragraph this good.

0 replies 0 reposts 2 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Thanks so much! Means a lot coming from you!

0 replies 0 reposts 0 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Wondering when to weight your data in regressions?
Want to see a great illustration of collider bias?
Want to learn how to respond to critiques of your paper in a productive way?

All this and more in a paper I never wanted to write, but made worthwhile anyway:
osf.io/preprints/so...

1 replies 2 reposts 6 likes

Reposted by Atheendar Venkataramani

Reposted by Atheendar Venkataramani

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Thank you! I'll check it out!

0 replies 0 reposts 0 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

#econsky I have a DID without staggered adoption but I am worried about heterogeneous tx effects across space + need for conditional parallel trends. Would CSDID or dCdH be better here than a standard TWFE DID?

1 replies 3 reposts 4 likes

Reposted by Atheendar Venkataramani

Demography - the flagship journal of PAA's avatar Demography - the flagship journal of PAA @readdemography.bsky.social
[ View ]

Posted: In “Spatiotemporal Changes in the Slavery-Inequality Relationship: Diffusion of the Legacy of Slavery,” Heather O’Connell et al. identify trends & illustrate “the potential for advanced spatiotemporal modeling techniques.”
read.dukeupress.edu/demography/a...

0 replies 1 reposts 0 likes

Reposted by Atheendar Venkataramani

Chloe N East (she/her)'s avatar Chloe N East (she/her) @chloeneast.bsky.social
[ View ]

ABAWD Work Requirements are center stage in SNAP policy debates

But SNAP's General Work Reqs affect 4X more people including hholds with kids

@jasonbcook.bsky.social & I provide first study of General Reqs and find
-No labor supply effect
-Big Disenrollment effects

www.nber.org/papers/w32441

0 replies 11 reposts 9 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

The "Cricket" episode of Bluey is among the greatest things I've ever seen.

0 replies 0 reposts 0 likes

Reposted by Atheendar Venkataramani

Saloni's avatar Saloni @scientificdiscovery.dev
[ View ]

New article by me:

The rise in reported maternal mortality rates in the US is largely due to a change in measurement. ourworldindata.org/rise-us-mate...

The change was adopted by different states at different times, resulting in what appeared to be a gradual rise in maternal mortality.

3 replies 31 reposts 78 likes

Reposted by Atheendar Venkataramani

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Interstellar soundtrack.

0 replies 0 reposts 1 likes

Reposted by Atheendar Venkataramani

Reposted by Atheendar Venkataramani

Mathew Kiang's avatar Mathew Kiang @mathewkiang.bsky.social
[ View ]

In a new paper lead by Ben Schlüter, we estimated about *1.2 million* youth (<18y) in the US have lost a parent due to drugs or guns since 1999. In 2020, drugs and guns accounted for about one in four parental deaths — about double the proportion in 1999.

jamanetwork.com/journals/jam...

1 replies 7 reposts 11 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Thanks man. I'm sure we'll be doing it again - this was the third attempt at a take down for one of the authors of the critique.

1 replies 0 reposts 0 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

But these are also situations where SDID may not be appropriate.

1 replies 0 reposts 1 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Thanks Seth! Nice thing about SDID is that it is already efficient relative to SC and DID and we have some very precise estimates. But in cases where we don't have as generous a donor pool or noisier pre-period covariates (to a point) can make a difference

1 replies 0 reposts 1 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Completely agree! I was worried a referee would ask but there was no good ex ante rationale for covariates (and most will be the "bad" kind). Thank you so much!

0 replies 0 reposts 1 likes

Reposted by Atheendar Venkataramani

Michael Clemens's avatar Michael Clemens @mclem.org
[ View ]

They can give a closer match between T&C in the pre-period, but the estimated treatment effect is often sensitive to covariate choice—increasing researcher degrees of freedom and thus decreasing transparency. Better I think to just match on outcomes whenever possible (or pre-commit to covariates)

1 replies 1 reposts 0 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Thank you so much! The command allows for them but I couldn't figure out a good reason to use them!

1 replies 0 reposts 0 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Hey #econsky!

Quick question - are covariates needed in synthetic difference-in-differences? My intuition is that if one assumes conditional parallel trends (PT) in a TWFE world, then the answer is no (since SDID imposes PT).

But are there other situations where you might want covariates?

2 replies 1 reposts 1 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

That's a great story. The editor in our case made the same call - but we still spend so much time on it!

0 replies 0 reposts 2 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

One other point here: why did such a flawed critique lead to months of stress+wasted time for us? Critiques like this exist because take-downs, especially of "politicized" topics, are in and are celebrated publicly. This sad scientific culture means extra effort is needed to clear one's name.

1 replies 0 reposts 5 likes

Reposted by Atheendar Venkataramani

Maria Glymour's avatar Maria Glymour @mariaglymour.bsky.social
[ View ]

Listening to BUSPH's Andrew Stokes present on his work on COVID-19 mortality undercounting. I am so impressed with what he & his colleagues have accomplished & proud of my tiny contribution. It seems like such a basic right- to know what is killing people. To have your family know what killed you.

0 replies 1 reposts 2 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Here is our response to a recent critique of our 2018 paper on police killings+mental health.

The critique had obvious problems, but nevertheless instigated a formal investigation (which we passed). This was stressful. I am posting because our reply could be useful for teaching:

osf.io/v72q6

2 replies 3 reposts 9 likes

Reposted by Atheendar Venkataramani

David McKenzie's avatar David McKenzie @dmckenzie.bsky.social
[ View ]

Lee bounds in practice: today's post covers how to use covariates, what to do when your outcome is binary or has ties, issues using Lee bounds with LATE, how to discuss the bounds if they include an opposite sign effect, and other practical issues...

blogs.worldbank.org/en/impacteva...

1 replies 10 reposts 7 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Thank you! I really appreciate that. Even though we were in the right here, I felt a sense of shame about this whole thing (and the paper as a whole) and just wanted it to go away. But we can't have open science without some discomfort I suppose. Hopefully the "take-down" mindset will stop.

0 replies 0 reposts 1 likes

Reposted by Atheendar Venkataramani

Elizabeth Wrigley-Field's avatar Elizabeth Wrigley-Field @wrigleyfield.bsky.social
[ View ]

Firearms are now the #1 cause of US youth deaths (displacing car accidents). But WHO is dying?

We show:

Black youth firearm deaths have been rising in RURAL AREAS for >1 decade;

Since 2018, Black rural youth have matched Black urban youth

This is new

www.nejm.org/doi/full/10....

4 replies 16 reposts 24 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Here is our response to a recent critique of our 2018 paper on police killings+mental health.

The critique had obvious problems, but nevertheless instigated a formal investigation (which we passed). This was stressful. I am posting because our reply could be useful for teaching:

osf.io/v72q6

2 replies 3 reposts 9 likes

Reposted by Atheendar Venkataramani

Reposted by Atheendar Venkataramani

Reposted by Atheendar Venkataramani

Reposted by Atheendar Venkataramani

Don Moynihan's avatar Don Moynihan @donmoyn.bsky.social
[ View ]

New validated measure of administrative burden tolerance w data from 7 countries.
Older adults, better educated and those in poorer health more opposed to burdens. Conservatives and those with high trust in govt more tolerant of burdens.
papers.ssrn.com/sol3/papers....

0 replies 7 reposts 21 likes

Reposted by Atheendar Venkataramani

Reposted by Atheendar Venkataramani

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

In medicine - where the stakes are high - there is an empathetic culture around mistakes, one that supports those who make them while fostering learning for everyone. In social science work - where the stakes are not high - people who make mistakes face hectoring, public take-downs.

0 replies 0 reposts 6 likes

Reposted by Atheendar Venkataramani

Zach Parolin 's avatar Zach Parolin @zparolin.bsky.social
[ View ]

A few studies find that personal earnings inequality (finally) declined in the US from 2010-2022. But, a twist: household income inequality continued to climb. What explains the discordant inequality trends? @lukaslehner.bsky.social, @natewilmers.bsky.social, and I try to answer this Q in a new WP:

0 replies 2 reposts 6 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Glad to hear you are ok! You're a rare one - great researcher with a fantastic sense of humor.

0 replies 0 reposts 1 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

Thanks for sending me this! Such great teaching material!

0 replies 0 reposts 1 likes

Atheendar Venkataramani's avatar Atheendar Venkataramani @atheendar.bsky.social
[ View ]

#econsky - has anyone written out collider bias in simple mathematical terms (e.g., not as a DAG)? This is for a class.

Thanks!

1 replies 3 reposts 3 likes