Professor Jennifer Castle, Sir David Hendry, and Dr Andrew Martinez explain why the OBR’s forecasting model needs an overhaul ⬇️ https://bit.ly/4czvnfa
Andrew B. Martinez
@andrewbmartinez.bsky.social
Working on forecast accuracy and uncertainty. My views are my own. RT ≠ endorsement. https://sites.google.com/view/andrewbmartinez/
In the era of A.I. agents, many Silicon Valley programmers are now barely programming. Instead, what they're doing is deeply, deeply weird. nyti.ms/4bmWQA0
Coding After Coders: The End of Computer Programming as We Know It
In the era of A.I. agents, many Silicon Valley programmers are now barely programming. Instead, what they’re doing is deeply, deeply weird.
nyti.ms
My paper examining the accuracy of hurricane damage nowcasts is now out in the International Journal of Forecasting. Open access link (until Jan 3) is here: tinyurl.com/4eztf7xx
shorturl.at
Britain’s Famous Forecasting Failure. 38 years ago today Britain was blindsided by one of its worst storms. 18 lives lost, 15 million trees destroyed & a billion pounds worth of damage. Almost no one, not even Britain’s trusted weather forecasters, had seen it coming. nyti.ms/4769vVV @nytimes.com
How the UK’s Hurricane-Strength Storm of 1987 Became a Famous Weather Forecasting Failure
The Great Storm of 1987 was a forecasting blunder that left at least 18 people dead, felled 15 million trees and caused a billion pounds worth of damage.
nyti.ms
1. The philosophy of science sometimes gets an unearned reputation as a purely academic exercise that offers little by way of concrete tools for advancing research. This is wrong. And today, as we grapple with how AI is changing the nature of scientific activity, it's desperately wrong.
This is the second largest 2m revision on record. In general they get larger at cyclical turning points. We take more signal that we are at a cyclical turning point than that there is some data collection conspiracy
Powell stressed today that the Fed's decisions will be based on how the economic data performs in coming months. One problem: That data may be becoming less reliable. My story on the BLS's latest cuts to CPI data collection: www.nytimes.com/2025/07/30/b... #EconSky
Cuts to Data Collection May Erode Reliability of Economic Statistics
nytimes.com
EXTREMELY BAD NEWS for economic research, per former BLS Commissioner @ericagroshen.bsky.social on LinkedIn. BLS is suspending access to its restricted data "for the forseeable future." Applies to projects through the Federal Statistical Research Data Centers & onsite projects with BLS. #EconSky
The BLS, which produces the CPI, told economists this week that a hiring freeze led the agency to cut back on the number of businesses where it checks prices. Response rates for one data set increased in the April CPI as fewer businesses were surveyed www.wsj.com/economy/cpi-...
Exclusive | Economists Raise Questions About Quality of U.S. Inflation Data
The Labor Department says staffing shortages reduced its ability to conduct its massive monthly survey, forcing it to turn to less precise guesses.
wsj.com
This isn't good: "BLS is reducing sample in areas across the country. In April, BLS suspended CPI data collection entirely in Lincoln, NE, and Provo, UT. In June, BLS suspended collection entirely in Buffalo, NY." #EconSky www.bls.gov/cpi/notices/...
Notice of CPI collection reductions
Notice of CPI collection reductions
bls.gov
More and more meteorologists are incorporating AI-powered weather models into their forecasting toolbox. This story starts with a new AI weather model built by Microsoft, but also provides a round-up of the recent rapid developments in this field. www.washingtonpost.com/weather/2025...
How an AI weather model by Microsoft produces faster, more accurate forecasts
The model is the latest development in a growing cadre of AI weather models that have emerged in the past few years.
washingtonpost.com
Did the Bank of England’s new scenarios achieve much? I give it a “requires improvement” score, partly because they are just versions of the central scenario on.ft.com/3RZ4SGd A bad scenario for the UK
New @budgetlab.bsky.social report about today's tariff announcement: • The April 2nd policy alone is the equivalent of an 11.5pp increase in the effective tariff rate. When combined with other US tariffs in 2025, we're at 22 1/2%, the highest rate since 1909. 1/7
My chapter together with @ajsw.info on "Forecasting the Macroeconomic Effects of Physical Climate Risk" has now been published in the Handbook of Research Methods and Applications in Macroeconomic Forecasting
Edward Elgar Publishing@elgarpublishing.bsky.social · 2y ago📣 NEW: Handbook of Research Methods and Applications in #Macroeconomic Forecasting by Michael P. Clements & Ana Beatriz Galvão Chapter 14 is #OpenAccess at: doi.org/10.4337/9781... More info: www.e-elgar.com/shop/isbn/97... @mpfarrho.bsky.social @danilocascaldi.bsky.social #Econometrics
Shared some thoughts on the growing challenge of insurability in the United States with the BBC. A growing number of eyeballs outside the US now paying attention to how the insurance market is handling increased natural catastrophe loss costs and growing hazard-related risks.
Climate change is fuelling the US insurance problem
Extreme weather events are making it hard to insure homes in certain parts of the US. What happens when insurance companies simply stop insuring?
bbc.com
Xmas coming early: Really excited to publish a paper in the Journal of Econometrics w Felix Pretis & Xiyu Jiao! Here's a short thread about what we do - and how we show the effects of outliers when estimating damages from climate change.
#EconSky Treasury's Office of Macroeconomic Analysis is hiring this year as well! Please spread the word and apply by 12/22: www.aeaweb.org/joe/listing....
📢 Call for papers! 📅 Deadline 8 Dec In honour of Prof. Sir David Hendry's 80th birthday, the 26th Dynamic Econometrics Conference is coming to Oxford! Mark your 📅 for 3-5 Apr 2024 www.dynamiceconometrics.com
25th Dynamic Econometrics Conference
Submit Your Paper for the 25th Dynamic Econometrics Conference. The conference will take place in London and online on 13 - 14 April 2023.
dynamiceconometrics.com