Stephen Hansen
@StephenEKHansen
Followers
1K
Following
354
Media
14
Statuses
82
Using new data sources to better understand economic activity and behavior. Professor @EconUCL. Texpat.
London, UK
Joined August 2022
Explaining firm-level reactions to macro shocks https://t.co/2FdjMzs5UB via @cepr_org
cepr.org
Understanding differences in firm responses to macroeconomic shocks is key for designing policy actions. This column identifies 67 macro shocks using ‘jump’ days in the US stock market over 2020-20...
0
0
1
📢Full-time predoc or part-time PhD RA position available at @TheIFS
https://t.co/glAKEjpOBA
@ImranRasul3, @fpvinay, and I have a series of projects mapping skills in online job postings to labor market outcomes. Work at intersection of econ/data science/ML/AI. @econ_ra
app.beapplied.com
The Institute for Fiscal Studies (IFS) is hiring a Research Assistant (RA) to support ongoing work that measures the skill content of individual jobs from massive databases of online job postings. ...
0
11
25
More generally, establishing procedures for valid inference in the growing world of AI-generated indicators is a major future challenge. #econtwitter
#AI
#LLM
https://t.co/4cEuy9WFgh
arxiv.org
Researchers now routinely use AI or other machine learning methods to estimate latent variables of economic interest, then plug-in the estimates as covariates in a regression. We show both...
0
1
1
We also show that an IV strategy that uses a human-labeled sample to purge the measurement error in generated variables works poorly when the number of labels is small relative to the unlabeled data.
1
0
0
We provide an illustration of how bias correction increases the estimated impact of remote work on wages across occupations.
1
0
0
We provide a simple bias correction formula that applied researchers can easily use. This restores valid inference and has quantitatively important effects even when AI/LLM are extremely accurate.
1
0
0
We consider the realistic case where algorithms become more precise as the sample size increases. In this setting, **point estimates are biased** but **standard errors are correct**. This is the opposite of the typical generated regressor problem.
1
0
0
Suppose we treat an AI-generated variable as "data" in a regression model. One intuition is that measurement error biases coefficient estimates. Another is that ignoring uncertainty biases standard errors. Which is it?
1
0
1
📢 **new results** LLMs and AI can be used to extract measures from text like sentiment, beliefs, and uncertainty. What can go wrong when plugging these measures into regressions and how to fix the problem? Read more below and check out https://t.co/4cEuy9W7qJ for details
arxiv.org
Researchers now routinely use AI or other machine learning methods to estimate latent variables of economic interest, then plug-in the estimates as covariates in a regression. We show both...
1
26
126
Gracias por haberme invitado!
Hoy tuvimos la oportunidad de participar y compartir opiniones en seminario @BBVAResearch "Fine-Tuning of Large Language Models" by @StephenEKHansen. ¡Panorama Fascinante para la #IA y #LLMs 👇 @Manuj_Hidalgo @rdomenechv @judith_arnal @_Herce @bde_Research @lugaricano
0
0
1
📢 VIMM Seminar 📢 Next week, Anna Cieslak (@DukeU) will present her paper "Policymakers’ uncertainty" @vimacro_org Registration: https://t.co/t8MhknkBoD Info/Paper: https://t.co/Pc7OiO6oN7
#EconTwitter #Research #Communication #Uncertainty #FOMC #MonetaryPolicy #Economics
0
5
13
Check this out for a great application of variational inference to the important problem of estimating nursing home quality. Shows the great promise of VI for large-scale structural models more generally. Congratulations @andrewolenski and @sacher_szymon!
Very excited to see this paper with @sacher_szymon finally accepted! We have lots of results interesting to anyone curious about health care, machine learning, Bayesian econometrics, and especially their intersection! 🧵(1/N)
0
0
4
Thanks for organizing this @ricco_giovanni and Luca Gambetti! Was a great couple of days. #bseforum
@RefetGurkaynak Stephen Hansen (UCL) presenting "Firm-Level Shock Exposures, Stock Returns and Real Outcomes: a Text-Analytic Approach" with Steven J. Davis and Cristhian Seminario-Amez (listening to some deep question)
1
1
27
📢Predoc in Economics + Data Science at @EconUCL I am seeking an RA to implement NLP, machine learning, and AI tools for economic research. Based fully at UCL/London. Funding in place for two years. Apply by May 3 at https://t.co/MTQ2Bkgv21.
@Econ_jobs_ @econ_ra
ucl.ac.uk
UCL is consistently ranked as one of the top ten universities in the world (QS World University Rankings 2010-2022) and is No.2 in the UK for research power (Research Excellence Framework 2021).
1
31
95
Interested in using Large Language Models or unstructured data in economics or business research? Last week to register for this **free** course at IESE's beautiful Madrid campus! https://t.co/KxvcTxZkqU
@iesebschool
iese.edu
1
18
54
Postdoc opportunity in high-dimensional Bayesian statistics with me and David Rossell https://t.co/em4seDoOAn. Great chance to combine modern statistics with economic applications and unstructured data. Located in Barcelona @bse_barcelona
https://t.co/eXrU6V7jxp
sites.google.com
Publications Book and Software I am a faculty member of the Statistics@UPF group at Universitat Pompeu Fabra in Barcelona (Dept. of Economics), and director of the Barcelona School of Economics...
0
3
8
Unstructured data is increasingly used to build measures of economically important variables like uncertainty and competition. But regression analysis using these measures as “data” is prone to a large bias. Read more below! L Battaglia, T Christensen, @sacher_szymon
Inference for Regression with Variables Generated from Unstructured Data
0
7
21
📢We’re hiring! We’re looking for an Assistant Professor (Lecturer) in any field of economics to join our team @ucl Department of Economics. 📅Deadline: 24 November 2023 Find out more and apply➡️ https://t.co/Uno60d5obd
1
71
124
Exciting news! Distinguished economists from renowned universities have joined the Fable Economics Council to enhance our data insights and proprietary economic indicator. #FableInAction #EconomicsExperts #DataForGood #Research #economy #EconomicsCouncil #AcademicResearch
0
1
8