Arthur Turrell

Personal view but it's not a binary thing; there are incredible people who will commute from Cardiff or Bristol (if they can drive–the office is a little way from the train station) but ofc the pool of labour in the centre of Manchester, Leeds, Bristol, Cardiff, etc, is a lot bigger.

"Getting the right skills is the most important part of any endeavour, and where you site your statistical production should reflect this ... A place with a thick labour market for highly educated graduates and post-graduates, but not too high a cost of living, is most likely to deliver this."

Very much agree with @dianecoyle1859.bsky.social; problems are deeper and structural (my personal view!) Won't surprise anyone that I've been thinking about this a lot—put together a hopeful & forward looking vision for what one could do with a blank slate here: aeturrell.com/blog/posts/f...

i like to explore the errors and failure cases of LLMs as much as the next guy, but if you're gonna be serious about assessing whether this stuff is useful you really can't ignore that there are a ton of cases where 40-50% as good for 1% of the cost (or even 10-20% with QA) is extremely exciting

Or any time

@bestofecontwitter.bsky.social

I'm looking for good pedagogical materials on causal inference in (observational) longitudinal data (also just 2-4 time points), e.g. interrupted design, pre-post design. I have a feeling economics has written lots to be imported to cogsci but I see only pieces & no overview. @p-hunermund.com ?

Does anyone know of a blog post somewhere that lays out best practices for ChatGPT prompts having to do with classification of text (in my case, open-ended answers in a survey)? Thanks!

No10 Innovation Fellowship - No10 Innovation Fellowship The No10 Innovation Fellowship Programme is 10 Downing Streets flagship initiative for bringing world class technologists into government for high impact tours of duty.

If you're interested, you can find details of the Fellowship here:
no10innovationfellows.campaign.gov.uk

Great evening launching this year's No10 Innovation Fellowship @designmuseum.bsky.social! It was incredible to meet so many talented people enthused about doing the hard yards needed to make the public sector even better. The appetite for reforms and improvements was genuinely very inspiring.

Currently revamping the default output view for the Gelbach Decomposition. Goal is to help both academics but also data scientists in industry to communicate results easily. Are the column and panel headers immediately clear? Any better way to summarize decomp results?

Careers Picture a safer economy for all – then help make it happen

I'll say more about the new gig in due course. In the meantime, if you're a data scientist and you want to make a contribution to the public sector, the Bank is increasingly an exciting and excellent place to do it. Check out jobs here: www.bankofengland.co.uk/careers

Data governance to set us free − speech by James Benford Given at the Gartner Data & Analytics Summit

If you want the wider picture, our departing Exec Director for Data has spoken a lot about what we're doing in speeches too:

www.bankofengland.co.uk/speech/2025/...

www.bankofengland.co.uk/speech/2024/...

Cutting through Complexity: How Data Science Can Help Policymakers Understand the World Economies are fundamentally complex and becoming more so, but the new discipline of data science-which combines programming, statistics, and domain knowledge-can help cut through that complexity, pote...

And I can share some of what I've written about how data + AI can help the public sector, which has informed a lot of the work I've done internally at the Bank over the last couple of years

arxiv.org/abs/2502.03010

Shining a light on private equity backed corporates in four findings Neha Bora, Sarah Burkinshaw, Alice Crundwell and Tuli Saha Private equity (PE) has rapidly become an important source of financing for UK businesses. Funds use pools of capital, largely from instit…

But I can share some links to published work by our wider team that reflect the advancing use of data:
bankunderground.co.uk/2024/09/25/s...

bankunderground.co.uk/2025/07/24/w...
bankunderground.co.uk/2025/06/05/m...
bankunderground.co.uk/2024/10/17/3...
bankunderground.co.uk/2024/04/18/t...

I've been genuinely very lucky to have such a phenomenal team + colleagues, both locally & across the Bank, and we have collectively really pushed forward the way the org does data, data science, & data talent. It's not job done but there's real momentum. Many of our successes have been internal...

Some personal news!

I am temporarily leaving the Bank of England to go on secondment to No10 Downing Street as an Innovation Fellow. It's been an amazing two-ish years working on the links between the real economy & financial system, & leading data strategy in financial stability strategy & risk.

We're all Jerome Powell now

Warming stripes by Ed Hawkins Source: https://en.wikipedia.org/wiki/Warming_stripes?wprov=sfti1#

Climate spiral by Ed Hawkins Source: https://en.wikipedia.org/wiki/Climate_spiral?wprov=sfti1#Critical_response

Hey #rstats, #pydata, #julialang & #dataviz folks! 😊

What are some iconic data visualizations, similar to @edhawkins.org‘s warming strips & climate spiral?

Hi #datascience #pydata world! We have an ad up for a senior data science manager in London or Leeds. You will have a major role in setting how data science is used at one of the UK's most important policy institutions! £108,800-£122,400.

www.datascientistjobs.co.uk/job/senior-m...

please repost

My mum always favoured a long interview, sometimes with the offer of a bible to take home thrown in.

Hi Ellie, fan of the podcast! This tweet really stuck with me and I was wondering if could point me at any of that discussion, if it's available? I'm interested in how EBM understand (or do not understand, in this case) causality / stats.

My award, engraved on a glass box

Receiving the award

I'm very proud to have won the Royal Statistical Society's award for best statistical commentary by a non-journalist this year 🥹

My home venue, very exciting! 100% of things I ever saw there involved the audience shouting "he's behind you" and I truly hope that your show is no different

Pyfixest continues to hit it out of the park. Huge congratulations @s3alfisc.bsky.social and team!!

Coding for Economists will be updated with the quantile regression methods in due course.

Is it a coincidence that this roughly correlates with the rise of the celebrity-authored children's book!? 👀😋

I have a detailed (Python) example of using the new API here that you could port to R fairly easily using an LLM. aeturrell.github.io/coding-for-e...

GitHub - aeturrell/cookiecutter-research-project: A cookiecutter for creating research projects A cookiecutter for creating research projects. Contribute to aeturrell/cookiecutter-research-project development by creating an account on GitHub.

See also github.com/aeturrell/co...

monroeclinton.com/counting-all...

Lovely example of a very pure and fun data science project

Weather / Meteorology folks -- how might get information on simulated paths of weather / hurricanes? E.g. what is the probability of a place getting rain, a place getting hit by a hurricane, etc.?

Basically what's the place to get the models that make these: