Michael Pollard | Lightnews

Michael S. Pollard

@sociologian.bsky.social

1.8K followers 470 following 200 posts

Sociologist and Demographer at RAND. Research mostly on social connection, substance use (alcohol and cannabis), and migration, but with many tangents. Opinions = my own, RTs ≠ endorsements.
https://scholar.google.com/citations?user=iRep51wAAAAJ&hl=en .. more

Videos

Psychology 30%

Public Health 24%

Posts Replies Media Videos

Pinned

Michael S. Pollard @sociologian.bsky.social · Nov 11

13 months ago I refused to promise "no cats" on this account. Here, finally, is a shameless appeal to new followers: my cat in a box.

A calico cat sitting in a box that is far too small, surrounded by cat toys that cost money but are being ignored at the expense of the box.

Michael S. Pollard @sociologian.bsky.social · 3d

Watched it and generally enjoyed it last night, other than a scene where it is nighttime in France and California at the same time. (Not a spoiler).

Michael S. Pollard @sociologian.bsky.social · 9d

One data source we used was all available tweets from the 1% Twitter Firehose API between January 5, 2020 and August 15, 2022. Related: sadly, last night I accidentally updated my Twitter phone app for the first time since it became X (July 2023) and I lost the Twitter Bird icon.

Reposted by Michael Pollard

Kevin Collins @kwcollins.bsky.social · 9d

I hear they rely now on machine learning models on the text of letters to the North Pole to evaluate whether kids are naughty or nice. Calling it "Sentiment And Naughtiness Text Analysis" or SANTA for short

Michael S. Pollard @sociologian.bsky.social · 9d

More findings: 2) tuning leads to notable improvements over pretrained models, 3) cross-target tuning can provide a viable alternative to in-target tuning in some settings, and 4) complex prompting strategies lead to improvements over pretrained models but underperform tuning approaches. 2/2

Michael S. Pollard @sociologian.bsky.social · 9d

New publication alert! In this study, we evaluate lexicon-based models, supervised models, and LLMs for stance detection using two corpuses of social media data. Our findings demonstrate that: 1) LLMs can distinguish stance on a specific target even when multiple subjects are mentioned...1/2

Cambridge University Press Political Science & IR @cambup-polsci.cambridge.org · 9d

#OpenAccess from @polanalysis.bsky.social -

Stay Tuned: Improving Sentiment Analysis and Stance Detection Using Large Language Models - https://cup.org/3XZutSs

- Max Griswold, Michael W. Robbins & @sociologian.bsky.social

#FirstView

Reposted by Michael Pollard

Cambridge University Press Political Science & IR @cambup-polsci.cambridge.org · 9d

Michael S. Pollard @sociologian.bsky.social · 14d

Sociologists! Were any of you involved with the nomination/development of the CIP Code: 45.0102 Social Science, Research Methodology and Quantitative Methods? Would love to get in touch. The nomination apparently included supporting materials from @asanews.bsky.social

nces.ed.gov/ipeds/cipcod...

CIP user site

Classification of Instructional Programs

nces.ed.gov

Michael S. Pollard @sociologian.bsky.social · 15d

Another daily reminder that other than the Baby Boom cohort, these other cohorts are all just loosely defined without real agreement. They didn't even get the Boomer age correct here.

Michael S. Pollard @sociologian.bsky.social · 17d

I like that only the opossum is singled out for peculiar flavor here. Did any pages look like well-used recipes with spills/stains?

Reposted by Michael A. Messner, Rebecca Sear, Cameron Campbell , and 2 more Michael A. Messner, Rebecca Sear, Cameron Campbell, Karen Benjamin Guzzo, Michael Pollard

Philip N Cohen @philipncohen.com · 18d

New from me: In the New Republic @newrepublic.com, I wrote that we should refuse to choose between the easy way and the hard way to boost birth rates.

What If Both Sides in the Birth Rate Debate Are Wrong?

The right wants to steer women away from work and toward motherhood. The left wants to provide them with more state support so it’s easier to have children. We should refuse this choice.

newrepublic.com

Michael S. Pollard @sociologian.bsky.social · 22d

This new commentary over at the Council on Foreign Relations, based on work with my colleagues at RAND, highlights a few ways women in China have been impacted by the legacy of the One Child Policy and China's hukou (household registration) system.
www.cfr.org/blog/gendere...

Gendered Legacies of Control: How China’s Population Policies Reshaped Women’s Lives

Guest Contributors are Tahina Montoya, a defense and policy researcher at RAND, and Kelly Atkinson, a political scientist at RAND.

www.cfr.org

Michael S. Pollard @sociologian.bsky.social · 23d

SMALLER BRAINS

Michael S. Pollard @sociologian.bsky.social · 23d

"Raccoons have adapted to living in urban areas to such an extent that they are now showing physical changes that resemble early signs of domestication...Their snouts have become shorter... smaller teeth, curlier tails, smaller brains and floppier ears."

www.theguardian.com/us-news/2025...

Drunk raccoon found passed out in Virginia liquor store

Store employee found masked bandit sleeping off a bender after invading booze store and tippling a tad too much

www.theguardian.com

Michael S. Pollard @sociologian.bsky.social · 23d

All the under-foot floof attests to a best lived life.

Michael S. Pollard @sociologian.bsky.social · 24d

One of the cool things in this report is an analysis of Chinese of public sentiment towards the 2- and 3-Child Policies. Changes in net sentiment align with two phases of pronatalist policy: introduction (people are optimistic) and implementation (people grow frustrated).

A chart showing public sentiment towards the Two- and Three Child Policies in Chinese social media posts from 2017 to 2025.
Based on post volume and net sentiment trends in the Figure, we identify four distinct periods:
1. From January 2018 to January 2020, social media users posted at high volumes, with sentiment generally negative.
2. From January 2020 to January 2023, post volume decreased (and remained at low levels) and net sentiment turned from negative to positive, strengthening in positivity.
3. From January 2023 to July 2024, net sentiment decreased until leveling out at zero, reflecting neutral sentiment.
4. From July 2024 to July 2025, net sentiment briefly turned positive before decreasing.
We assess that the changes in net sentiment align with two phases of pronatalist policy:
introduction and implementation. Net sentiment after the two-child policy was implemented is consistently negative, with many QQ users expressing frustration over the recklessness and lack of empathy of Chinese policymakers who overlooked public opinion and ignored their basic needs for building a livable, fulfilling, and sustainable life.
With the launch of the three-child policy in 2021, we assess that the steadily increasing net sentiment score reflects social media users’ optimism about the policy change. Yet after local governments began implementing the policy in 2023, net sentiment turned negative. We assess that this change reflects the mismatch between social media users’ expectation for promising policy and disappointment at the reality of implementation.
Social media users reflect positive sentiments again in 2024, which aligns with the government’s issuing of 13 targeted and comprehensive measures to promote fertility. As the government continued to struggle to reverse declining fertility and implemented cash subsidies in 2025, net sentiment again turned negative.

Reposted by Mike Brewer, Ben Barr, Michelle Kelly‐Irving , and 10 more Mike Brewer, Ben Barr, Michelle Kelly‐Irving, Pierre Bataille, Nils Gustafsson, Joshua Goodman, Michael Pollard, Alessandro Nai, Todd Pugatch, Sebastian Karcher, Jens Rommel, Patrick Präg, Xavier Molénat

Per Engzell @pengzell.bsky.social · 28d

A thread of directed acyclic graphs (DAGs) that look like record covers... because that's EXACTLY what the world needs

1. Huey Lewis and the News: link.springer.com/chapter/10.1...

Michael S. Pollard @sociologian.bsky.social · Nov 25

They are way ahead of you. It's on tour now. www.babybjorn.com/north-americ...

North American Bouncer Tour

Join us at store events as we bring our beloved giant bouncer on tour around North America!

www.babybjorn.com

Michael S. Pollard @sociologian.bsky.social · Nov 24

As was the style at the time.

Michael S. Pollard @sociologian.bsky.social · Nov 24

@genebaxter.bsky.social Good old Utah over-delivering on the Oak- names!

Michael S. Pollard @sociologian.bsky.social · Nov 24

This whole thread is a wild ride. Get a cup of coffee and enjoy!

TristanBridges.bsky.social @tristanbridges.bsky.social · Nov 24

Baby names aren’t just personal choices; they reflect culture, geography, gender, identity, and as i've been investigating, politics too. I’ve been analyzing 50 years of U.S. baby naming to see how they map onto political polarization in the U.S.
Here’s what I found 🧵

Michael S. Pollard @sociologian.bsky.social · Nov 24

@genebaxter.bsky.social this whole thread might be relevant to your interests.

Michael S. Pollard @sociologian.bsky.social · Nov 24

What do you make of this?: "Applications voted by the committee to be in the middle third will be designated as 'competitive but not discussed' ... Applications in the middle third will be considered for funding, along with the discussed applications."

Reposted by Philip N. Cohen, Karen Benjamin Guzzo, Michael Pollard , and 1 more Philip N. Cohen, Karen Benjamin Guzzo, Michael Pollard, Rense Corten

TristanBridges.bsky.social @tristanbridges.bsky.social · Nov 24

Reposted by Michael Pollard

Karen Benjamin Guzzo @karenguzzo.bsky.social · Nov 24

New NIH guidance related to the shutdown. One of the changes is that "the percent of applications discussed in most meetings will be reduced to 30-35%, instead of the current ~50%."

NOT-OD-26-012: Updated Guidance on Reopening of NIH Extramural Activities Following the October 1, 2025 - Lapse in Appropriations

NIH Funding Opportunities and Notices in the NIH Guide for Grants and Contracts: Updated Guidance on Reopening of NIH Extramural Activities Following the October 1, 2025 - Lapse in Appropriations NOT-...

grants.nih.gov

Michael S. Pollard @sociologian.bsky.social · Nov 19

[cue "Just What I Needed" by The Cars]

Michael S. Pollard @sociologian.bsky.social · Nov 17

"Twenty years ago, the U.S. spent almost four times as much as China on research and development. But by 2023, China had almost closed the gap."

American Sociological Association @asanews.bsky.social · Nov 17

Research from ASA members Yu Xie and Junming Huang @princeton.edu illustrates how China is attracting scholars, especially in STEM fields, in the wake of the administration’s funding cuts and immigration restrictions in this @washingtonpost.com article.

Why Trump’s cuts to scientific research are a big win for China

China is attracting American scientific talent, especially in STEM fields, partly due to funding cuts and immigration restrictions under President Donald Trump.

bit.ly