Eva Vivalt @evavivalt.bsky.social · 29/09/2026Very frustrating when a paper is rejected on the basis of a negative review that is not just largely AI-generated (per Pangram) but materially incorrect. If you're going to use AI for your reviews, at least be thoughtful about it. 180
Eva Vivalt @evavivalt.bsky.social · 16/09/2026@elizabethrds.bsky.social @alexbartik.bsky.social @dbroockman.bsky.social @patrickkrause.bsky.social @smiller.bsky.social 020
Eva Vivalt @evavivalt.bsky.social · 16/09/2026OpenResearch released a public-use data set for the OpenResearch Unconditional income Study (ORUS) - a guaranteed income experiment that paid $1,000/month to 1,000 people for 3 years, with a control group receiving $50/month. Test your hypotheses here: data.openresearchlab.org 2122
Reposted by Eva VivaltEva Vivalt @evavivalt.bsky.social · 14/08/2026Have a study, want to collect forecasts, but don't have time? New pilot: the SSPP will draft a forecasting survey for you with AI based on your pre-analysis plan. RA assistance is available to help polish the draft. Contact team@socialscienceprediction.org if interested. 041
Eva Vivalt @evavivalt.bsky.social · 14/08/2026Have a study, want to collect forecasts, but don't have time? New pilot: the SSPP will draft a forecasting survey for you with AI based on your pre-analysis plan. RA assistance is available to help polish the draft. Contact team@socialscienceprediction.org if interested. 041
Eva Vivalt @evavivalt.bsky.social · 14/08/2026So great to see this out! Over ten years in the making. 1192
Eva Vivalt @evavivalt.bsky.social · 10/08/2026Finally, household structure appears to mediate impacts: effects of a cash transfer could be quite different depending on how it is targeted. Read more: nber.org/papers/w35589 @alexbartik.bsky.social @dbroockman.bsky.social @patrickkrause.bsky.social @smiller.bsky.social @elizabethrds.bsky.socialnber.orgIncome Shocks and Intrahousehold Dynamics: Evidence from a Guaranteed Income ExperimentFounded in 1920, the NBER is a private, non-profit, non-partisan organization dedicated to conducting economic research and to disseminating research findings among academics, public policy makers, an... 160
Eva Vivalt @evavivalt.bsky.social · 10/08/2026Our results show that evaluating cash transfers only at the individual level misses part of the story, as there are substantial spillovers to other household members. They also highlight that relationships can change as a result of $ without changes in household structure. 10/ 181
Eva Vivalt @evavivalt.bsky.social · 10/08/2026Several indicators of intimate partner violence declined, especially for women. But reported rates were low, the overall index did not change, and estimates did not survive multiple-testing adjustment. 9/ 160
Eva Vivalt @evavivalt.bsky.social · 10/08/2026Some suggestive heterogeneity masked by aggregate results of no effects on marriage: those without a partner in the household at baseline may have been more likely to stay unpartnered, and those with a partner might be more likely to be in a relationship. Cash enables choice. 8/ 170
Eva Vivalt @evavivalt.bsky.social · 10/08/2026What didn’t change much? Generally, household composition, partnership/cohabitation, decision-making power, or overall division of labor. Among partnered recipients, there was suggestive evidence of greater trust and improved relationship quality. 7/ 161
Eva Vivalt @evavivalt.bsky.social · 10/08/2026One perhaps counterintuitive result: recipients were 2 pp more likely to report that someone in the household had forgone healthcare. This may reflect newly identified needs and is consistent with what we see for recipients in Miller et al. (2026) and Vivalt et al. (2026). 6/ 180
Eva Vivalt @evavivalt.bsky.social · 10/08/2026Household members were also more likely to begin education or job training. This appears to have been driven primarily by effects on partners; children in our sample are generally not old enough to be observed going to college yet. 5/ 181
Eva Vivalt @evavivalt.bsky.social · 10/08/2026There were spillover effects on the employment outcomes of other people in the household. For those who had a partner in the household, partner work hours seemed to have declined by about 2 hours/week, but this is not significant after multiple hypothesis testing correction. 4/ 1101
Eva Vivalt @evavivalt.bsky.social · 10/08/2026Larger households also seemed to have larger increases in expenditures inside the household. These results are consistent with direct expenditures on others such as purchases for children observed in Krause et al. (2026). 3/ 170
Eva Vivalt @evavivalt.bsky.social · 10/08/2026Recipients became more likely to support others. They gave more money, helped more people, and shifted from net recipients to net providers. Net transfers to friends and family rose by about $135/year. 2/ 1111
Eva Vivalt @evavivalt.bsky.social · 10/08/2026What happens when someone receives $1,000/month for 3 years? The effects don't stop with the recipient. In this new NBER working paper, we examine what changed for partners, households, friends & family after 1,000 people received $1,000/month with 2,000 in a control group. 🧵1/ 27627
Reposted by Eva VivaltVoxDev @voxdev.bsky.social · 20/07/2026🆕 Can forecasting social science make experiments better? Today on VoxDev, Stefano DellaVigna (UC Berkeley) &Eva Vivalt (University of Toronto) discuss why research forecasts overestimate effects, and how to correct and reweight them: ow.ly/jfjL50ZpWlvow.lyCan forecasting social science make experiments better?Evidence from 100 Social Science Prediction Platform projects shows that research forecasts are too optimistic on average, but still predictive enough to improve study design when they are combined, shrunk, and weighted carefully. 043
Eva Vivalt @evavivalt.bsky.social · 16/07/2026www.nber.org/conferences/...nber.orgSI 2026 Forecasting & Empirical Methods 010
Eva Vivalt @evavivalt.bsky.social · 16/07/2026How can better forecasts of research results improve power calculations? Presenting in just a few minutes! 110
Eva Vivalt @evavivalt.bsky.social · 16/07/2026It feels like a bit weird of a new way to engage with research but I feel slightly more informed now - and don't mind at all if anyone uses LLMs to pre-engage with my research. :) 001
Eva Vivalt @evavivalt.bsky.social · 16/07/2026In advance of listening to a presentation on FRI's forecasts of the economic effects of AI, I thought to update this tracker: evavivalt.com/writing/fri-...evavivalt.comFRI's AI Forecast Survey vs. Reality - Eva Vivalt 120
Eva Vivalt @evavivalt.bsky.social · 15/07/2026If Spain wins the whole thing, then this World Cup they would have had 0 losses and only 1 tie: Cabo Verde. 41297
Eva Vivalt @evavivalt.bsky.social · 14/07/2026Off to NBER SI, where I'll present on forecasting research results using data from the Social Science Prediction Platform. I'll be at sessions each of the three weeks! (Hey, if jobs are changing to be more based on interaction....) Please connect & catch up if you'll be around. 010
Eva Vivalt @evavivalt.bsky.social · 14/07/2026Basically: we are flying blind and just hoping everything works out, and not only works out, but goes as well as possible. I don't think people are taking that seriously enough yet. I signed this statement because it's important to build common knowledge. 140
Eva Vivalt @evavivalt.bsky.social · 13/07/2026I joined economists and AI researchers in signing “We Must Act Now: A Statement on AI’s Transformation of the Economy.” wemustactnow.aiwemustactnow.aiWe Must Act Now 110
Eva Vivalt @evavivalt.bsky.social · 13/06/2026They're looking to block access to Fable - nevermind any more powerful future models - to non-Americans. archive.ph/vDzFQarchive.ph 021
Eva Vivalt @evavivalt.bsky.social · 12/06/2026Interesting read. I genuinely worry all the time about Canada falling behind with AI. It seems so hard to turn around and the US absolutely has its own incentives. 000
Eva Vivalt @evavivalt.bsky.social · 26/05/2026If you use any of my work in your teaching, please let me know so I can cite it in my tenure package! Thanks! (We have a pretty late deadline here.) 002
Eva Vivalt @evavivalt.bsky.social · 18/05/2026Very happy to be presenting on forecasting research results, based on the Social Science Prediction Platform data, at the IDB's EconNet seminar series. There's a public Zoom which you can find by googling, in case it's of interest (trying to avoid bots).... It starts at 12 ET! 000
Reposted by Eva VivaltEthan Mollick @emollick.bsky.social · 04/05/2026This is from the co-founder of Anthropic, interesting that he refers to public sources when he is also obviously privy to lots of internal sources that he cannot discuss. I assume he sees the same thing at Anthropic. importai.substack.com/p/import-ai-... 88219
Eva Vivalt @evavivalt.bsky.social · 30/04/2026Thanks, good question - yes, it goes beyond asking LLMs to make a forecast. 000
Eva Vivalt @evavivalt.bsky.social · 30/04/2026Thanks, I hadn't thought to but perhaps it's worth running on a larger set to benchmark; might be gathering that data in any case for another project. 010
Eva Vivalt @evavivalt.bsky.social · 30/04/2026Sorry, it changed url with a site redesign: evavivalt.com/writing/pre-...evavivalt.comA Note on Pre-Analysis Plans in Economics - Eva Vivalt 110
Eva Vivalt @evavivalt.bsky.social · 29/04/2026Try it out and see how it does for you. If you're willing to share its report on power, I'll give you free credits. 010
Eva Vivalt @evavivalt.bsky.social · 29/04/2026^ The above table is based on work with the Social Science Prediction Platform. How human forecasts can also improve power calculations is in the paper: www.nber.org/papers/w34493nber.orgForecasting Social Science: Evidence from 100 ProjectsFounded in 1920, the NBER is a private, non-profit, non-partisan organization dedicated to conducting economic research and to disseminating research findings among academics, public policy makers, an... 030
Eva Vivalt @evavivalt.bsky.social · 29/04/2026I recently presented some results that showed that using LLM forecasts in experimental design could dramatically improve power. Studies in our data start severely under-powered (~0.4). Using LLM forecasts would bring it up to ~0.7! And further improvements coming. 120
Eva Vivalt @evavivalt.bsky.social · 29/04/2026Did you ever wish you could get help with your power calculations? Now you can! earlyreview.ai will give you estimates of treatment effects based on your early project documents (e.g., pre-analysis plans, registered reports, grant proposals, etc.). 2226
Eva Vivalt @evavivalt.bsky.social · 21/04/2026Just a note that with all this "LLMs accurately identifying users from text", some free text responses in allegedly de-identified survey data could in principle identify research respondents, if not today then soon. 020
Eva Vivalt @evavivalt.bsky.social · 08/04/2026My secondary school shared a building with the university, so one semester I found myself teaching in the same building as my secondary school. 030
Eva Vivalt @evavivalt.bsky.social · 06/04/2026... Ooh, I guess I'll also need to do this with a script just in case, I didn't check every word or anything. 020
Eva Vivalt @evavivalt.bsky.social · 06/04/2026In drafting up the new site, Claude also of its own volition changed the acceptance/etc. status of some of my papers. Like, um, no, please don't? But that one didn't go live, thankfully. 240
Eva Vivalt @evavivalt.bsky.social · 06/04/2026The best meal of my life was a chickpea salad from a shop after completing the Milford Track. I can't replicate the salad, because the taste was not in the salad but in the hike. 070
Eva Vivalt @evavivalt.bsky.social · 05/04/2026I had deleted and started a new repo in figuring out how I wanted the page to be. It was the same repo, just hiding all my messy changes to get it to that point, since GitHub Pages is public. So I traded one potential embarrassment for another.... 2/2 110
Eva Vivalt @evavivalt.bsky.social · 05/04/2026In principle I could have re-checked it, I just hadn't thought it would be necessary since I had checked it before and the ask seemed trivial. 1/2 110
Eva Vivalt @evavivalt.bsky.social · 05/04/2026I caught it on the non-live version and asked it to fix that before it was pushed. I checked it out, it was fixed, but then later we rebuilt the repo, and in rebuilding the repo it probably overwrote the fix. That’s my best guess, but this was a while back. 2/2 130