Blane Lewis
blanedavidlewis.bsky.social
Blane Lewis
@blanedavidlewis.bsky.social

Professor, Crawford School of Public Policy, Australian National University. Research: Fiscal Federalism in Indonesia. Teaching: Quasi-Experimental Methods.

Economics 57%
Political science 33%

stephen a smith "leaving all doors open" to a presidential bid. me too!

also, this just in: chuck schumer is very concerned with developments, more evidence that democrats are fighting the good fight

i’ve lost 15% of my retirement savings so far but at least cory booker broke the record for the longest senate speech ever

Interested in political economy of local elections? We find that Indonesia's local democratic transition had a mostly negative impact on key public spending and service outcomes, especially in districts where clientelistic practices were pronounced.
authors.elsevier.com/sd/article/S...

Now instead of a thesis ‘stretching out over 100 pages’ we have three chapters each ‘stretching out over 100 pages’.

a "potential outcome" (although a very unlikely one)

Confounder and Collider

As good as random

Allman Brother Band

two things:
1. Stict focus on the identification of causal effects.
2. Do something on the US since apparently journal editors and reviewers will automatically consider such research of "general interest:"

Honestly frustrating

How many of you find this set of exchanges totally unintelligible?

impending death that's a fun topic

Schrödinger's welfare gains

people don't really seem thrilled to announce anything over here...

Not all year maybe but recently it's a tie:

Goldsmith-Pinkham et al, 2020. Bartik Instruments: What, When, Why, and How. AER

Borusyak et al. 2022. Quasi-experimental shift-share research designs. RevEconStud

I am evaluating a Aus govt policy to increase school funding. Looks like a Bartik thing

did he head off to the market or just stay at home?

For sure. More good news: you get to keep using that excuse as your kids get older too.

sorry *Vitor

Personally I am not a fan of the matching approach because: (non-random) loss of observations, potential regression to the mean, and it performs no better than conditioning on covariates (in my experience). I'm with Victor on using Callaway and Sant-Anna doubly robust procedures.

I am one of those.

He’s a dick obviously. But is it possible that after that second (or third) drink he was just trying to make a joke? Wait that’s stupid never mind.

Lumayan. Kamu?

#Econsky

I see people asking for intros for the newly arrived.

I am a professor of economics at Crawford School, Australian National University. I study fiscal federalism and teach quasi experimental methods.

Long time US expat with > 30 years living in: Indonesia, Singapore and (now) Australia

So liven things up a bit. But no dunking on Australians (or more importantly US citizens living in Australia)

Only LATE if covariates enter non-parametrically. (Sorry I think I used that same joke on that other site.)