Congrats! It’s a fantastic paper!

Thanks a ton!

All this matters for how we understand multilateralism.

Global North drives the show and pushes for pet agendas, and the state representation structures within int'l orgs is where all this transpires.

✅ Open access article here: doi.org/10.1080/0969...

Do they get what they want?

YES!

The more rich countries push for market liberalization reforms in a developing country's loan, the more likely it is that IMF staff will include such reforms when they revise conditionality.

Importantly, the IMF's powerful Global North shareholders push for this agenda *collectively*, as they refer and support each other all the time during debates.

🇺🇸 (largest shareholder) and 🇬🇧 & 🇫🇷 (major former colonial powers) dominate and drive the discussions!

But what do they actually talk about?

A lot of the discussion focuses on their preferences for opening up markets in the IMF's borrowers: privatizing SOEs, liberalizing labour markets, etc.

We show that it is bilateral economic relations between the speaker and the country-under-discussion that drives commenting patterns.

Thus, Board members properly represent the interests of their appointing states, rather than drinking the Kool-Aid of IMF staff's worldviews.

Then, we examine three questions:

1. What determines Board members' participation in different debates?

2. How do these powerful Board members coordinate?

3. Are they successful in shaping the design of IMF-mandated reforms?

Initially, we study participation in Board meetings and present 2 stylized findings:

- 🇺🇸 speaks the most

- Middle-income countries receive more attention than low-income countries

Neither is surprising, thus a good indicator that our data is capturing something meaningful

Our analysis focuses on the behavior of the IMF's 5 largest shareholders: 🇺🇸🇬🇧🇯🇵🇩🇪🇫🇷

This was purely due to space constraints for the article - stay tuned for more analyses!

Our dataset contains information on 3,111 Board meetings between 1995-2015: we extracted all comments and matched them to the person speaking.

This added up to 38 million words—about 1.8 million words per year!

Better yet, the data is fully accessible online. You can use it! 🎉

We collected thousands of transcripts from the IMF's archives to figure out how Board members behave in debates vis-à-vis developing countries.

These debates are important because the Board has the ultimate authority to approve IMF loans and their mandated policy reforms

States are represented in the boardrooms of international organizations, but what do they do on these boards?

This remains a black box for scholars, as it's generally hard to access the transcripts.

Thanks for creating the list!

Depends on outside options too.

Powerful countries can reach deals outside multilateral settings or set up alternative arrangements (eg AIIB).

For less powerful countries, multilateralism much more important & may even trump domestic distributional concerns, but don’t know of empirical studies

💯! Also, without deep knowledge, where are the good ideas on stuff that can be causally identified gonna come from? Otherwise, debates -even with adv methods- end up incremental & esoteric to subfields.

Wonder how AI is going to change meth training, given chatGPT can write statistical code etc

It was. But industrial policy is more than infrastructure financing. What the FT article is referring to is something more structural, transformative and state-led.

But haven’t read the underlying WB reports myself