Jessica Leight

12.5K posts

Jessica Leight banner
Jessica Leight

Jessica Leight

@leightjessica

Development economist, @Yale / @UniofOxford / @MIT. Asso editor, REHO / CER / PloS One / BMC Public Health. Mom of 3. Views my own

Connecticut, USA Katılım Eylül 2013
316 Takip Edilen10.8K Takipçiler
Sabitlenmiş Tweet
Jessica Leight
Jessica Leight@leightjessica·
What countries are overrepresented in dev research? Using new database of published dev papers, a scatterplot of papers published vs. log extremely poor population (<PPP $2.15 in 2021); left-truncated for < 100K ppl. Curve captures proportional representation (given log scale)
Jessica Leight tweet media
English
1
17
67
9.8K
Jessica Leight
Jessica Leight@leightjessica·
This morning I opened the laptop to check email, realized I needed some additional (simple) robustness checks for a paper, opened Claude and entered my prompt; then my toddler climbed on my lap + I cuddled himwhile answering simple Qs so the agent could do tedious coding
English
6
6
68
8.5K
Jessica Leight
Jessica Leight@leightjessica·
@ho_ben I haven't tried that yet but I'll probably be there soon. Also if I take out my phone someone usually tries to grab it to watch cocomelon (youngest), play 80s one hit wonders (eldest), or play who let the dogs out (middle). Good thing is they can't yet operate a laptop
English
0
0
0
284
Ben Ho
Ben Ho@ho_ben·
@leightjessica a nice new feature of gpt codex is easy access to the agent from your phone. so could check in on agent progress and give new instructions while waiting for kid to fall asleep
English
1
0
1
453
Pedro H. C. Sant'Anna
Pedro H. C. Sant'Anna@pedrohcgs·
@leightjessica I participated in a chat about AI at the Atlanta Fed and my main point was this all along: without AI, I would have not been able to do any serious work this year. Having 5 small kids, teaching obligations, and doing research is quite a challenge! I am very thankful to AI
English
4
8
114
5.2K
Jessica Leight
Jessica Leight@leightjessica·
@auyonomous That's what I meant, but actually the toddler was chiming in too with questions like is that an elephant? With an eye toward the office shelf. (Yes it was - my office is full of elephants)
English
0
0
4
270
Auyon Siddiq
Auyon Siddiq@auyonomous·
@leightjessica I'm assuming here the "simple Qs" were from the agent and not the toddler
English
1
0
2
373
Jessica Leight
Jessica Leight@leightjessica·
I really think the AI discourse is sleeping on implications for balancing work with care work! Probably bc most people involved don't care much (+ are mostly men), but there's lots to unpack. Some good (like my morning), some bad (like constant pressure to upskill + do more)
English
0
5
20
913
yasemin dildar
yasemin dildar@yasemindildar·
I just got back from Turkey to find this letter waiting for me. I'm happy to share that I've been promoted to Full Professor. It's been a pretty great summer so far!
yasemin dildar tweet media
English
52
7
426
27.7K
Jessica Leight
Jessica Leight@leightjessica·
Saturday R&R is the best kind of R&R because it somehow redeems all the weekens ruined by checking email + getting rejections
English
2
0
44
3K
Jessica Leight
Jessica Leight@leightjessica·
@rglenner @JPAL @FCDOGovUK Should be feasible, if the acknowledgments note what initiative provided the funding + I can match it to the sources for each initiative
English
0
0
0
32
Jessica Leight
Jessica Leight@leightjessica·
Who funds development RCTs? Using my new database of dev papers 2021-2025, I pulled information from funding acknowledgments. (This requires, generally, full-text, and is not feasible using metadata; thus I did not do this for non-RCTs.) #econtwitter #econsky
English
2
25
125
14.1K
Jessica Leight
Jessica Leight@leightjessica·
But given the high level of dispersion, smaller funders dominate in all three categories. Other ideas for analysis welcome!
Jessica Leight tweet media
English
1
0
6
719
Jessica Leight
Jessica Leight@leightjessica·
If we look across funder tiers for the top 22 funders, J-PAL is even more dominant in the top 5, with IGC also overperforming. The World Bank, USAID, and Gates are overrepresented in the Journal of Development Economics.
Jessica Leight tweet media
English
1
0
6
837
Jessica Leight
Jessica Leight@leightjessica·
@jonasdovern I was thinking the premium even in where the paper is published (editors / referees viewing papers differently based on presentations)
English
0
0
0
13
Jessica Leight
Jessica Leight@leightjessica·
I don't subscribe to the view of AI as plagiarism (nor would I like to do a video chat referee report); but I can easily believe that the rise of AI tools will lead to more of a premium being placed on high-quality conference or seminar presentations
Branko Milanovic@BrankoMilan

There is no easy solution for AI plagiarism. The modus operandi of the journals will have to change, One possibility is that refereeing not be done as now by reading and commenting on the article but by arranging a video conversation between the referees and the author(s). In a conversation, it would become clear if the article is an AI plagiarism or is genuine. Generally speaking, we have to move to oral examinations.

English
2
0
14
3K
Jessica Leight
Jessica Leight@leightjessica·
Being able to articulately explain + defend your work + its importance is an important skill and one that shouldn't be (and can't be, really) outsourced to AI. If other research inputs become more abundant + this remains scarce, it will be more valuable
English
0
0
3
314
Jessica Leight
Jessica Leight@leightjessica·
@akib_kn Thanks! It's just five years going backward from last year. I would doubt the time period makes a big difference, but I could certainly extend it at some point.
English
0
0
1
12
Akib Khan
Akib Khan@akib_kn·
@leightjessica Amazing public good, as always. Thanks! Tangential q: Is there any particular reason to start in 2021? I feel like 20-22 saw a lot of backlog clearing, esp for RCTs as people didn't get to launch new ones. Shouldn't make a lot of difference, though, I suppose.
English
1
0
0
12
Jessica Leight
Jessica Leight@leightjessica·
What are the most common topics in development economics? Following on the analysis of what % are RCTs, I developed a Python script to assign topic fields. (All the code + results are public here - anyone is welcome to use it.) #econtwitter #econsky github.com/leightjessica/…
Jessica Leight tweet media
English
1
14
32
3.5K
Jessica Leight
Jessica Leight@leightjessica·
But, I do find it interesting that the subfields that are often the focus of the criticism for being just "optimization of charity projects" are also the most female, with the opposite being true of the subfields that are often highlighted as being "what we all should be doing"
English
0
0
0
92
Jessica Leight
Jessica Leight@leightjessica·
Dev is well-known to be more heavily female (+ sample also includes more journals). By topic, we see the highest concentration of women in gender, health, and education, + lowest representation in trade / macro, political economy, and conflict / crime. No big surprises here
Jessica Leight tweet media
English
1
0
1
529
Jessica Leight
Jessica Leight@leightjessica·
How many authors of development economics papers are women? Around 30% #econtwitter #econsky. Updated calc using my trusty new database of recently published (2021-2025) economics papers, and using the functions gender_guesser and Namsor to infer gender.
Jessica Leight tweet media
English
1
2
12
1.4K

Keşfet

@pedrohcgs @Ni_Zakharov @ho_ben @auyonomous @yasemindildar @rglenner @JPAL @FCDOGovUK