Paddy Carter
@paddycarter.bsky.social
3.2K followers 1.1K following 1.2K posts
Head of development economics at British International Investment, formerly CGD, ODI, lapsed academic economist and macroeconomics hobbyist. https://sites.google.com/site/paddycarter/
Posts Media Videos Starter Packs
Reposted by Paddy Carter
mongabay.com
Natural forest regrowth in the world’s tropical rainforests is expanding.

According to the Forest Declaration Assessment 2025, more than 11 million hectares (27 million acres) of tropical moist forests are under some stage of forest regrowth between 2015 and 2021.
Report finds increased tropical forest regrowth over the last decade
Natural forest regrowth in the world’s tropical rainforests is expanding. According to the Forest Declaration Assessment 2025, more than 11 million hectares (27 million acres) of tropical moist…
news.mongabay.com
paddycarter.bsky.social
he says the AI objective function does not penalise errors, so it's a workaround.

WHY DOESN'T THE AI OBJECTIVE FUNCTION PENALISE ERRORS!!!????
paddycarter.bsky.social
I was talking to someone who (I don't understand any of this) says he has created an AI persona who is punished for making stuff up and (somehow) rewarded for not doing so, and he runs what he does through that and it seems to work. Nuts.
paddycarter.bsky.social
why can't they get the thing to say "I am sorry, I cannot do X"?
paddycarter.bsky.social
why have the designers of AI made things so that it says:

"I sincerely apologise for doing Y, I understand that you asked me to do X not Y. I shall now do X"

[does Y]

over and over again?
Reposted by Paddy Carter
cgdev.org
Country platforms are gaining traction as vehicles to align national priorities, policy reforms, & more.🌍

Alexia Latortue & Jared Goodman draw on interviews to examine if platforms can deliver on their promise, highlighting 4 critical issues and providing recommendations:
https://bit.ly/48qq2q4
Platforms that Perform: Helping the Next Generation of Country Platforms Deliver
There is intense and renewed interest in the not-so-new concept of country platforms. Invoked in conversations from G20 to COP30, these government-led coordination mechanisms that align national prior...
bit.ly
paddycarter.bsky.social
you fail to understand how the almighty was in peril from the marketing tactics of a brewery
paddycarter.bsky.social
I would be very proud of this paragraph, had I wrote it
paddycarter.bsky.social
and I know f-all about whether and if so how Ofwat could have done things differently, to avoid "take out all the money, run it into the ground" behaviour
paddycarter.bsky.social
I won't be able to find it but I swear I read a NAO report that compared the costs of PFI and state-financed projects under a single "everything goes to plan" scenario - a report that is always cited in the press as evidence PFI is bad value.
paddycarter.bsky.social
my mainstrem econ take is that private finance can be good value for relatively simple-to-contract stuff (roads, solar parks) but when you go near complicated (hospitals, water, London Undergrounds) any potential gains will be overwhelmed by transactions costs / getting shafted in negotiations
paddycarter.bsky.social
However, it is a perennial bug bear how many "the government's cost of capital is lower, it should just build everything" takes there are which make no effort to think about the allocation of risks under alternative arrangements.
paddycarter.bsky.social
then you read stories about private providers getting paid to screw in lightbulbs or whatever, which is the exact opposite of the incentives these contracts are supposed to convey (build to a higher quality to reduce maintenance)
paddycarter.bsky.social
I'm just thinking of a simpler story where investors put money in, but then don't get it back if the thing they have financed is obliged to keep delivering the agreed service when i.e. costs overrun. But no doubt that's a fantasy, my grasp on the reality of these things is very weak
paddycarter.bsky.social
but I don't really trust press accounts either, which I don't think are inclined towards balance. I am aware of research that shows the private route does seem better value for some things (simple things like roads, mainly, I think). It's all bit fog of war
paddycarter.bsky.social
the impression I get from reading about these things in the press is that nothing like that occurs, and when things go wrong the state and/or consumer ends up having to cough up more and the private financiers are not taking the risks that might justify their returns (in good states of the world)
paddycarter.bsky.social
it's perfectly OK to pay premia on financing if there are states of the world in which lenders and shareholders lose money, the state/consumer still gets [whatever] at agreed price, and in those same states of the world under the state-owned arrangement the state would be liable for additional costs
paddycarter.bsky.social
I am not sure I understand what you mean by "representative agent" which I thought meant modelling the aggregate economy as if driven the decisions of a single actor, which Aghion & Howitt certainly do not do.
paddycarter.bsky.social
a little more generously, the desire to put into theory things happening as the result of purposeful actions taken by agents (this case firms) who are trying to pursue and objective in a defined environment (taking that from kevinbryanecon.com/mokyraghionh... ).
A Nobel for Innovation: Mokyr, Aghion, and Howitt
kevinbryanecon.com
paddycarter.bsky.social
I still remember Saul Bellow
paddycarter.bsky.social
hopefully the energy regulators can engineer some competition between the storage providers
Reposted by Paddy Carter
pseudoerasmus.bsky.social
Four items by Mokyr: the first is a topic already recognised by the Nobel, but the other three are less well known areas of Mokyr: his study of the Irish famine, his explanation of why the Dutch Republic was not the first, and his artisan theory of the Industrial Revolution.