Challenges in Statistics: A Dozen Challenges in Causality and Causal Inference (Carlos Cinelli, Avi Feller, Guido Imbens, Edward Kennedy, Sara Magliacane & Jose Zubizarreta) arxiv.org/abs/2508.17099

🎉 New article published in the Journal for Labour Market Research with Christoph Kern (@lmumuenchen.bsky.social)! We compare the current rule-based profiling model with various statistical models in terms of model discrimination and calibration labourmarketresearch.springeropen.com/articles/10....

Causal inference needs experiments *and* theory: "Joel Mokyr (...) demonstrated that if innovations are to succeed one another in a self-generating process, we not only need to know that something works, but we also need to have scientific explanations for why" www.nobelprize.org/prizes/econo...

"No ha sido el asombro, sino el fervor social surgido de los sórdidos callejones y la tristeza de las vidas marchitas lo que ha provocado la ciencia económica" (Arthur Pigou). Guau.

📙 Glad to share this new research with @junquera.bsky.social. 📊 We analyse how outsourcing affects time-related dimensions of job quality, focusing on temporary work, part-time contracts and scheduling in the facility management business. 📄 Chapter: lnkd.in/d9nB8Tzj 🔗 Book: lnkd.in/dsfuK6uW

One of the greatest beauties/difficulties of economics is the problem of being a system and its dynamics: multiple pipes (direct, indirect, and induced effects) and different forces (structural causal models) at the long and short run. It makes me crazy

Never ask a guy using/teaching clustering about the robustness of results to slight changes in the distance metric, the pre-specified number of clusters or the geometrical shape of the clusters

The AI infrastructure build-out is so gigantic that in the past 6 months, it contributed more to the growth of the U.S. economy than /all of consumer spending/ The 'magnificent 7' spent more than $100 billion on data centers and the like in the past three months *alone* www.wsj.com/tech/ai/sili...

"En el siglo XXI la divergencia en PTF con respecto a EE. UU. se explica principalmente por la debilidad de la productividad del capital en España, mientras que en el siglo XX la productividad del trabajo era la principal responsable de esta divergencia" www.fbbva.es/publicacione...

Sometimes a simple model is enough to describe the past (@davidautor.bsky.social in doi.org/10.1126/scie...)

OECD Employment Outlook 2025 - Can We Get Through the Demographic Crunch? www.oecd.org/en/publicati...

"El juez considera que el cambio de marco laboral aplicado por Glovo desde diciembre de 2024, optando por la supresión paulatina de los falsos autónomos por otro modelo de contratación laboral no ha supuesto un cambio efectivo en las cuotas de mercado de ambas compañías. elpais.com/economia/202...

Is working from home just a pandemic blip? Not according to the data. We analysed 1 billion job postings in 20 OECD countries. The share of jobs advertising remote/hybrid work quadrupled from 2.5% to 11% (2019–2023) and has been stable since 🧵

Interesting website centralizing European alternatives for digital services european-alternatives.eu

Recordemos que ACCIONA, junto a otras grandes constructoras de este país, han sido sancionadas por haber creado cárteles que amañaron concursos públicos. Resultado: colegios, carreteras, hospitales, asfaltados, etc. costaron más a los contribuyentes por el cartel. Breve hilo⬇️

It's truly pleasant to read The Economy 2.0 - Microeconomics. Although it is still defining market failures as Pareto-inefficiencies, it's cool it reflects Pareto-efficiency *and fairness* as two important criteria to evaluate allocations www.core-econ.org/the-economy/...

🧵 1/9 📊 Dades obertes del MITMS ens permeten saber quanta gent hi havia a cada municipi en cada moment. Mobilitat, esdeveniments, hàbits socials! 💡Periodistes i urbanistes: @elivivas.bsky.social @drxeo.eu @rogertugas.bsky.social @kikollan.llaneras.es 📍Exemples 👇

Nueva Conferencia Sectorial de Empleo y Asuntos Laborales. Lo positivo: financiación plurianual para las políticas activas de empleo. Lo negativo: seguimos con indicadores de desempeño que no dicen casi nada. www.lamoncloa.gob.es/serviciosdep...

Most applications in R run pretty quickly. But these days, many R users are faced with huge datasets and/or long-running applications, making parallel computation attractive. 🧵 1/

📢 Book Alert! "Global Trends in Job Polarisation and Upgrading: A Comparison of Developed and Developing Economies" is out! Published by Palgrave Macmillan/ Springer, this volume examines global patterns of job creation at a global scale | 🔗 link.springer.com/book/10.1007... #EconSky #sociology

He visto el otro país En nuestros hijos mi amor En la tremenda ilusión De creérmelo, de creérmelo open.spotify.com/track/3sLGrh...

Co-director @davidautor.bsky.social explains key findings from his recent NBER working paper, co-authored with our research affiliate @profdaviddorn.bsky.social, Gordon Hanson, Maggie Jones, and Bradley Setzler. Read the paper: shapingwork.mit.edu/research/pla... #EconSky

ineapir: an R package to query statistical data from National Statistics Institute (INE Spain) github.com/es-ine/ineapir

It's bonkers that in academic science we spend hours obsessing over the wording of a paper, but often only one person has seen the code that produced the results! 2/N

Estadística antes de Cristo (del artículo de Wikipedia sobre Moisés)

Data Scientist = Person who is better at statistics than any software engineer and better at software engineering than any statistician (@spite.vc). Fair enough!

Co-director @davidautor.bsky.social explains key insights from his recent QJE paper, which finds that 60% of the work we do today didn't exist in 1940. Read New Frontiers: The Origins and Content of New Work, 1940-2018: shapingwork.mit.edu/research/new... Full clip: youtu.be/A3-nEP1oyeA #EconSky

Nice comment on the state of measurement in psychology. People just keep piling on new constructs and new questionnaires. How can we clean up the mess that this creates? Feat. respectful operationalism and epistemic iteration

After a long wait, the working paper for the Many-Economists Project: The Sources of Researcher Variation in Economics. We had 146 teams perform the same research three times, each time with less freedom. What source of freedom leads to different choices and results? papers.ssrn.com/sol3/papers....

"Relations were extraordinarily tense. “The Common Room was carefully shared,” writes Constance Reid in her biography of Neyman. “[Egon] Pearson’s group had tea at 4; and at 4:30, when they were safely out of the way, Fisher and his group trooped in.”

Out now open access at @ajpseditor.bsky.social. 194 potential exclusion-restriction violations for studies using weather as an instrumental variable onlinelibrary.wiley.com/doi/full/10....

A reminder that {modelsummary}, {tinytable}, and {marginaleffects} are Canadian so if you're an American researcher, you'll have to increase your citations of those #rstats packages by 25%, err, next month.

Qué grande lo que ha hecho el Canijo. Qué grande nuestro carnaval 👏🏼👏🏼 #COAC2025P3

Describing descriptive studies using descriptive language, or the practical virtues of statistical humility statmodeling.stat.columbia.edu/2009/03/02/d...

Trying to compile a list of people with active pedagogic interests in teaching statistics go.bsky.app/Qg6YSq6

La democrática muerte elpais.com/eps/2025-01-...

Sorry for not letting you know in time for your Christmas gifts 🙃 | link.springer.com/book/9783031... #EconSky #sociology

A summary of my thinking on shared prosperity, work and AI in ten bullet points.

Unbundling college: Causal returns to US college majors, minors, and extracurriculars. An audit study of seniors' first-job search, with Carolina Bussotti, John Nunley, and @alanseals.bsky.social. We find: majors matter, zero returns to minors, positive returns to extracurriculars. 1/6 🧵

On teaching and the content/duration ratio of each lesson: "The job is not to cover material but to uncover it" (George Miller). Amazingly clear for statistics teaching

Agree. That's the inter-occupation side of the problem. On the intra-occupation side, computers can remove some alienation from our daily work. It's always about costs *and* benefits.

Guau. Es difícil pensar en una democracia sin capacidad de hablar. Tenemos que seguir construyéndola

My favourite causal inference book has just been updated!

🆕 Nuevo artículo en el que desarrollo una conceptualización de las políticas activas de empleo como subconjunto de las políticas de empleo Muy accesible como introducción a las políticas activas de empleo donde se presentan tipos, mecanismos, fundamentos... revistes.uab.cat/anuarioiet/a...

🚀 #spanishoddata #rstats is on CRAN! 🥳 #gischat Download it as usual: install.packages('spanishoddata') Read our guides in the articles section of the website: ropenspain.github.io/spanishoddata/ Report bugs at: github.com/rOpenSpain/s... Thanks to bsky.app/profile/robi... and Eugeni Vidal-Tortosa

We causal inference folks can still breathe

A reminder to submit your abstract to our workshop in May in Amsterdam! Please reskeet/spread the word.

Es tremendo este dato: en España hay 1.000 registradores y 2.400 notarios embolsándose unos 400.000 pavos (en promedio) al año por tener un chiringuito que debería ser del sector público Más datos en esta radiografía de @dyebra.bsky.social y @yulyjara.bsky.social www.eldiario.es/economia/rad...