iiinova — where your digital self keeps evolving

Reposted by Thiago Moreira

Political Analysis @polanalysis.bsky.social · 13/08/2026

Currently in FirstView: In “Let Them Eat Pie: Addressing Sample Selection in Multiparty Elections,” Ali Kagalwala, @tmqm.bsky.social, Guy Whitten, and Yongzhi Xu introduce a novel maximum likelihood approach that accounts for the biases that come from sample selection in a compositional setting.

111

Thiago Moreira @tmqm.bsky.social · 09/08/2026

Many thanks to Guy, Ali, and Yongzhi for the collaboration. We are putting the finishing touches on the SelectPie R package. In the meantime, practitioners can use the function we provide in the supplementary material.

010

Thiago Moreira @tmqm.bsky.social · 09/08/2026

We develop a maximum-likelihood approach that addresses sample selection, while preserving the compositional nature of vote shares. In a series of Monte Carlo simulations, our model outperforms conventional approaches, including the standard Heckman correction.

110

Thiago Moreira @tmqm.bsky.social · 09/08/2026

The paper tackles a common but overlooked problem in multiparty elections: many parties do not compete in every district, and the districts they do contest are not random. This partial contestation creates a sample-selection problem that standard compositional models do not account for.

110

Thiago Moreira @tmqm.bsky.social · 09/08/2026

This interest brought me together with Guy Whitten, Ali Kagalwala, and Yongzhi Xu on this project, “Let Them Eat Pie: Addressing Sample Selection in Multiparty Elections,” which is now out in Political Analysis.

110

Thiago Moreira @tmqm.bsky.social · 09/08/2026

Studying elections with compositional models was one of the reasons I went to Texas A&M for my PhD. I was surprised by how few scholars treat election outcomes as compositions, even though vote shares range from 0 to 1 and changes in votes for one party are offset by changes in votes for the others.

230

Thiago Moreira @tmqm.bsky.social · 07/11/2025

In this new paper with Spencer Goidel and @barmstrong.bsky.social, we show that the drug crisis boosts Republican vote, especially among independents, by lowering both Democratic support and abstention.

042

Reposted by Thiago Moreira

Adam Przeworski @adamprz.bsky.social · 24/03/2025

Here is the update through Week 6, early because I am traveling. adamprzeworski.substack.com/p/diary

adamprzeworski.substack.com

Diary

I decided to keep a record of my thoughts as events transpire, a diary.

1277

Thiago Moreira @tmqm.bsky.social · 24/03/2025

With survey data, I show that as the perceived polarization increases, partisans are less likely to have a negative view of the economy under their party’s administration, and those who still hold a negative economic evaluation are less likely to punish their party for it. 3/3

000

Thiago Moreira @tmqm.bsky.social · 24/03/2025

Using county-level data, I show that increasing polarization over time is associated with a decrease in the number of voters who cast a ballot for the opposition or abstain in response to negative economic conditions. The results are robust for the LDV and a time trend. 2/3

100

Thiago Moreira @tmqm.bsky.social · 24/03/2025

What happens to economic voting under high polarization? In this @bjpols.bsky.social paper, I argue that voters stick with their preferred party as the ideological distance between parties increases, reducing their propensity to hold incumbents accountable for economic outcomes. 1/3

1143

Reposted by Thiago Moreira

Victor Araújo @victoraraujo.bsky.social · 14/02/2025

New (open access) publication with @philipp-kerler.bsky.social. We draw on fieldwork to build a theoretical model that outlines the conditions when the expectation of clientelism can drive selective political engagement, something we call asymmetric participation. #distributivepolitics #clientelism

0135

Reposted by Thiago Moreira

Adam Przeworski @adamprz.bsky.social · 18/11/2024

Why democracy, backsliding, or autocracy? Unabashedly, I think this model has exceptionally rich results. The introductory overview is intended to be self-explanatory without delving into the math. See papers.ssrn.com/sol3/papers.cf

25918

Thiago Moreira @tmqm.bsky.social · 05/09/2024

My 6th APSA. Crazy how fast time flies.

120

Thiago Moreira @tmqm.bsky.social · 04/09/2024

Heading to APSA in Philly. If you will be there and still don't have plans for Saturday afternoon, drop by to see our presentation. José Cheibub, Andrea Junqueira, and I study effective participation across income groups with granular electoral data from Brazil.

020