Thread by @notEricMin

I am elated to say that my #APSR article with Azusa Katagiri is now available online! In it, we digitize 18,000 declassified documents from the Berlin Crisis of 1958-63 to test the relative credibility of public and private diplomatic signals.

dx.doi.org/10.1017/S00030…

We argue that in the real-world diplomatic environment, seemingly costly public signals are very noisy, making them easy to misinterpret or not perceive at all. Meanwhile, ostensibly costless private statements are rarer, direct, and focused, making them easier to perceive.

Our archival data includes 18,000 documents from the Department of State (private statements), the Foreign Broadcast Information Service (public statements), and the White House (elite perceptions). We digitized these and applied supervised learning methods to code for resolve.

Using these document data, we find that: (1) cost-sinking material actions have greater impacts on elite beliefs than any statement; (2) public statements are much noisier than private statements; and (3) private statements affect elite perceptions far more than private ones.

Our study challenges key facets of costly signaling theory and provides another reason why private diplomacy matters. Our process of digitizing archival documents can be applied to many topics — especially those historically limited to formal models and qualitative studies.