Check out our new special issue in the Agricultural Finance Review on 'Financing Resilient Food Systems' including papers from many great authors
editorial with @jisangyu.bsky.social Aleks Schaefer and @yanndemey.bsky.social
doi.org/10.1108/AFR-...
Figure showing the proliferation of methods used in NBER and CEPR working papers. Instrumental variables (IV) peak around 2010, twoway fixed effects (TWFE) five years later.
The rise and fall of instrumental variables and twoway fixed effects. From: arxiv.org/abs/2501.06873
đšToday 4pm CETđš
PCI Webinar: The Drain of Scientific Publishing: why publishing is becoming a burden for science and how to fix it
An overview talk based on three papers on #ScientificPublishing, where it stands & how to fix it. Plus dad jokes and bad acronyms.
peercommunityin.org/pci-webinar-...
Indeed, that is very likely the case. But on a pure principal ground, it feels wrong to accept (free) work when there is no response on the other end (I already contacted the journal about our own submission but got no relevant response)
A journal just sent me a review request. But we have a paper "with editor", i.e. stuck at the first stage after submission for 4 months now. Is it understandable that I request an update on that end first, before agreeing to do work for them? #notaskingforafriend
So â*in a world where there is really no effect*, the result would be produced by chance less than 5% of the timeâ is correct; but without adding that first part lord Bayes would like to have a word. In context: â*in a world where smoking bans actually do not work*â (2/2)
However, one significant comment: your laymanâs definition of a p-value perpetuate a statistical misconception đ it is crucial to add that it is a probability under the null hypothesis, and not in general (1/2)
@planetmoney.bsky.social I love the latest episode on Replication games!! Definitely a most listen for any scientist podcasts.apple.com/be/podcast/p...
Final days to apply!
Gemini even shamed me for allegedly referring to a past labelling incident (I was not) and then shamed me into suffering from Apophenia before admitting the data was fake =) bsky.app/profile/yann...
We are hiring at the Assistant Professor level! The position is open to anyone with a quantitative agri-business research and teaching profile. Feel free to reach out if you have questions đ
www.wur.nl/en/vacancy/a...
Another banger by Paolo Crosetto, Pablo GĂłmez Barreiro, and Mark Austin Hanson! While special issues with guest editors are not a problem in themselves, journals that build a business model around then invites people to start gaming the system...
Very grateful to have received this ERAE Reviewer certificate. But I am passing this on to @khoavuumn.bsky.social in recognition of his unique commitment to our profession. Should have been his in the first place.
And now a podcast segment is out on the Food for Europe podcast frol the EU: podcasts.apple.com/be/podcast/f...
Utrecht University abolishes its English language bachelor's program in economics and replaces it with a Dutch one.
They told staff (~30% international) the week before Christmas. Layoffs seem likely now. Unwarranted lobotomy is the only word I can find for it...
dub.uu.nl/en/news/econ...
A more general overview of the farmers' protests across Europe can be found (open access) in:
onlinelibrary.wiley.com/doi/10.1111/...
The full paper can be openly read here: www.sciencedirect.com/science/arti.... #farmer #protests #Brussels
âPolicy dissatisfactionâ and âUnfair market conditionsâ, were highly mentioned as protest reasons across Germany, France, Belgium and the Netherlands. The paper further focuses on farmersâ own stated grievances and how EU/national policy reactions aligned or failed to align with them.
In our recent Food Policy article, we examined farmersâ own stated grievances during last yearâs protests across four EU countries and how governments responded. Our findings suggested that unresolved and misaligned policy responses would likely contribute to repeated escalation.
The renewed farmersâ protests today highlight the persistence of European agricultural policy tensions. Condemning the associated violence and destructive actions is necessary. At the same time, preventing its recurrence requires understanding why farmers mobilize in the first place.
LinkedIn style would be: "Comment MOAR below and I will share the code via DM"
You can still join Day 3 online: agriculture.ec.europa.eu/eu-agri-food...
#EUAgriFoodDays #Brussels #EuropeanComission #Agriculture #farming #risk #riskmanagement #resilience
Finally, while President von der Leyen highlighted that the CAP aims to provide stability, I highlighted that in my research farmers consistently perceive *policy uncertainty* as a major source of risk. A stable policy framework is therefore essential in my view!
While traditional risk management focuses mainly on robustness to known risks, resilience thinking offers a broader perspective: preventing risks where possible, enabling adaptation, and even allowing for long-term transformation in the face of both known and unknown challenges.
This puts pressure on:
âą Farmers dealing with âeverydayâ on-farm risks
âą Markets handling rarer risks with big consequences
âą Governments stepping in when risks become catastrophic
My main point was: risk has *always* been part of farming, but todayâs risk landscape is changing fast. Risks increasingly interact and compound, and co-occur with big shocks like COVID-19 or the war in Ukraine.
It was amazing to take part in the đœ EU Agri-Food Days đ§âđŸ at the European Commission as a panelist in the policy session on "Building a resilient agri-food sector".
Will you incorporate LLMs and AI prompting into the course in the future? No. Why wonât you incorporate LLMs and AI prompting into the course? These tools are useful for coding (see this for my personal take on this). However, theyâre only useful if you know what youâre doing first. If you skip the learning-the-process-of-writing-code step and just copy/paste output from ChatGPT, you will not learn. You cannot learn. You cannot improve. You will not understand the code.
In that post, it warns that you cannot use it as a beginner: âŠto use Databot effectively and safely, you still need the skills of a data scientist: background and domain knowledge, data analysis expertise, and coding ability. There is no LLM-based shortcut to those skills. You cannot LLM your way into domain knowledge, data analysis expertise, or coding ability. The only way to gain domain knowledge, data analysis expertise, and coding ability is to struggle. To get errors. To google those errors. To look over the documentation. To copy/paste your own code and adapt it for different purposes. To explore messy datasets. To struggle to clean those datasets. To spend an hour looking for a missing comma. This isnât a form of programming hazing, like âI had to walk to school uphill both ways in the snow and now you must too.â Itâs the actual process of learning and growing and developing and improving. Youâve gotta struggle.
This Tumblr post puts it well (itâs about art specifically, but it applies to coding and data analysis too): Contrary to popular belief the biggest beginnerâs roadblock to art isnât even technical skill itâs frustration tolerance, especially in the age of social media. It hurts and the frustration is endless but you must build the frustration tolerance equivalent to a roachâs capacity to survive a nuclear explosion. Thatâs how you build on the technical skill. Throw that âwonât even start because Iâm afraid it wonât be perfectâ shit out the window. Just do it. Just start. Good luck. (The original post has disappeared, but hereâs a reblog.) Itâs hard, but struggling is the only way to learn anything.
You might not enjoy code as much as Williams does (or I do), but thereâs still value in maintaining codings skills as you improve and learn more. You donât want your skills to atrophy. As I discuss here, when I do use LLMs for coding-related tasks, I purposely throw as much friction into the process as possible: To avoid falling into over-reliance on LLM-assisted code help, I add as much friction into my workflow as possible. I only use GitHub Copilot and Claude in the browser, not through the chat sidebar in Positron or Visual Studio Code. I treat the code it generates like random answers from StackOverflow or blog posts and generally rewrite it completely. I disable the inline LLM-based auto complete in text editors. For routine tasks like generating {roxygen2} documentation scaffolding for functions, I use the {chores} package, which requires a bunch of pointing and clicking to use. Even though I use Positron, I purposely do not use either Positron Assistant or Databot. I have them disabled. So in the end, for pedagogical reasons, I donât foresee me incorporating LLMs into this class. Iâm pedagogically opposed to it. Iâm facing all sorts of external pressure to do it, but Iâm resisting. Youâve got to learn first.
Some closing thoughts for my students this semester on LLMs and learning #rstats datavizf25.classes.andrewheiss.com/news/2025-12...
