Lots of talk on X today about rethinking the results produced by close elections RDDs. So, I thought I'd drop a few recent articles about issues that arise in this space so that folks can have them all in one place. 1/N - ThreadSky

johnholbein1.bsky.social • 22 days ago

Lots of talk on X today about rethinking the results produced by close elections RDDs.

So, I thought I'd drop a few recent articles about issues that arise in this space so that folks can have them all in one place. 1/N

Comments

maibennett.com•21 days ago

cc @wsandholtz.bsky.social if you hadn’t seen this!

johnholbein1.bsky.social•22 days ago

#1. Hartman (2021) shows that most tests of covariate balance around close election cutoffs fail to rule out large imbalances. Remember to equivalence test when testing for covariate balance!

https://www.cambridge.org/core/journals/political-analysis/article/equivalence-testing-for-regression-discontinuity-designs/43F77CDC6337A63AE0A5E6DC3EE01A41

johnholbein1.bsky.social•22 days ago

#2 Marshall (2022) shows that it is far from clear what effects you are estimating with close election RDD estimates of candidate characteristics. There are compound treatments at these cutoffs!

https://onlinelibrary.wiley.com/doi/10.1111/ajps.12741

mikedenly.com•21 days ago

Very nice thread! Bertoli and Hazlett (forthcoming, Political Analysis) have a response to Marshall (2022) that is also relevant:

johnholbein1.bsky.social•22 days ago

#3 Stommes et al (2023) show that RDDs in political science, many of which are close election RDDs, are "underpowered to detect all but large effects." As with all low power tests, we should be wary not only about large standard errors, but also type M/S errors!

https://journals.sagepub.com/doi/10.1177/20531680231166457

johnholbein1.bsky.social•22 days ago

#4 de la Cuest and Imai (2016) discuss the limited external validity of close election RDDs. Generalize beyond close elections with care!

https://imai.fas.harvard.edu/research/files/RD.pdf

johnholbein1.bsky.social•21 days ago

#5 Here's another recent critique of close elections RDDs.

It's unclear whether there is actually good evidence against potentially fatal precise manipulation around the cutoff in these designs.

johnholbein1.bsky.social•21 days ago

The McCrary (2008) DCdensity and Cattaneo, Jansson, & Ma (2018; 2020) rddensity approaches are widely used as tests for precise manipulation. But virtually everyone interprets statistically insignificant results of these as evidence of no manipulation.

noahmckbraun.bsky.social•21 days ago

These threads are why I need a new bookmarks folder, thanks!

reillysteel.bsky.social•21 days ago

Good resource, thank you! Also, maybe this isn't "recent" anymore, but also Eggers et al. (2015)? https://onlinelibrary.wiley.com/doi/10.1111/ajps.12127

jblabonne.bsky.social•21 days ago

📌

Posting Rules

Be respectful to others
No spam or self-promotion
Stay on topic
Follow Bluesky's terms of service

Comments

Posting Rules

Reply