Prompts for economists
This page contains example prompts and responses intended to showcase how generative AI, namely LLMs like GPT-4, can benefit economists.
Example prompts are shown from six domains: ideation and feedback; writing; background research; coding; data analysis; and mathematical derivations.
The framework as well as some of the prompts and related notes come from Korinek, A. 2023. “Generative AI for Economic Research: Use Cases and Implications for Economists“, Journal of Economic Literature, 61 (4): 1281–1317.
Each application area includes 1-3 prompts and responses from an LLM, often from the field of development economics, along with brief notes. The prompts will be updated periodically.
Here is the page from Jesse Lastunen.
Is the mortgage interest deduction a good idea?
I usually don’t like arguments like the one that follows, as purely short-run second best considerations tend to rub me the wrong way. Nonetheless I had never thought of it before, so I am happy to present it for our collective enlightenment:
Mortgage interest deductions and other homeownership subsidies are widely believed to be harmful because they redistribute resources from lower-income renters to higher-income homeowners. We argue that renters actually benefit from these policies in general equilibrium for two reasons. First, the rental supply curve is relatively inelastic, which means that rents fall when these policies reduce rental demand. Second, many renters spend most of their income on housing, and these renters gain substantially from rent decreases. We calibrate a quantitative model to match empirical evidence on these factors and show they are strong enough that subsidizing homeownership actually increases welfare.
That is from a newly published article by Shahar Rotberg and Joseph B. Steinberg. Via the excellent Kevin Lewis.
Sunday assorted links
1. Excellent Zach Mazlish tweets on the macro puzzle of our day. He concludes with: “And note that all of 1/2/3 predict negative output effects to come if rates stay high for much longer, unless the (net!) positive supply shock is persistent (AI?), or fiscal demand pressures are persistent.”
2. Very good David Brooks piece on Nikki Haley (NYT). And you all know I don’t like to cover matters electoral.
4. Andreessen and Horowitz on what is wrong in higher education.
5. The Teacher’s Lounge (NYT) is a superb movie, most of all about the collapse of social trust in Germany. It has a chockful of points relevant to social science as well.
6. Data on these links and where they come from (“Straussianism peaked in 2021″…and yes I am always on the lookout for new instances of “Horse Nationalism”).
Comayagua, Honduras
Comayagua is one of the very nicest and most classic of Central American towns. It is safe (yes, this is Honduras), walkable, delightful, and comes to life later in the day in the main square. It is full of colonial buildings and some churches. I didn’t see any North American tourists, and the surrounding countryside is lovely. Population is about 100,000.
A few years ago Honduras switched its main international airport, so a flight to the capital Tegulcigalpa actually brings you much closer to Comayagua — visit there instead! Here is Wikipedia on the city.
As for food, go to Hotel Helechos (central in town, but oddly no one has heard of it), walk out to your right and immediately there is an amazing baleadas stand. The pupuseria on the corner of main square is excellent. In general, Honduras is the country where the quality difference between roadside and street food, compared to the restaurants, perhaps is greatest. And it doesn’t favor the restaurants.
Argentina fact of the day
The country was not quite as rich in the early days as it is sometimes made out to be:
Argentina’s performance on this measure is frequently exaggerated. In 1929, for example, Argentina’s per capita income was less than half of the average of other temperate agrarian societies (such as Canada and Australia) and of European industrialized countries (such as Great Britain, Germany, Belgium, and Sweden). In 1969 and in 1929, it was 38 percent of the U.S. figure…
That is from the very good 1996 Larry Sawers book The Other Argentina: The Interior and National Development. It is related to my earlier post on Salta.
The author also has an excellent explanation of how import substitution strategies run out of steam, even if they produce more growth in a shorter run. The import substitutes usually require subsidies to get started, which puts a squeeze on the government budget, and in fact you can think of import substitution as a kind of deficit spending/borrowing against the future. The import substitution also puts a squeeze on the agricultural sector, which for many countries, Argentina included, had been generating net foreign exchange. The balance of payments then worsens, which also becomes a longer run problem. Over time, in addition, obtaining the needed foreign inputs for the import-substituting sectors becomes yet another problem. In time, tariffs are needed for the nascent domestic sector, and that protectionism lowers living standards and also leads to higher corruption.
As the author notes:
In the early postwar years, ISI [import substitution] was highly recommended by almost every development economist in the world and pursued by virtually every Third World country.
At first it worked, but over time it fared far less well. This is one of the very best and also unheralded books about Argentina, as there are interesting points on almost every page. One point the author makes, for instance, is that the Argentina economy never had great facility in making high fixed investments, even before Peron and various later depredations. Most of all, this is a book that actually tries to answer your questions.
Matt Yglesias on the “soft landing”
I think that’s part of the explanation of the “soft landing.” Back during the inflationary spurt in 2022, nominal wages were growing very rapidly. The public sector struggles in that kind of situation, because it has much less flexibility to offer opportunistic wages or to pay new hires differently from old hires. So a bunch of people probably left their jobs when they realized that they could get raises by going elsewhere, and governments had a hard time replacing them. Then, when the economy cooled in the face of the Fed raising interest rates, there was this long hangover of public sector vacancies that started getting filled even without big pay increases. Even today, though, lots of jurisdictions have outstanding vacancies — you hear about this especially with police officers and teachers, but it’s impacting lots of categories of employment.
That is from Matt’s Friday (gated) column which is about many things, including what the world would have looked like if the American Revolution had not gone forward. This particular point is a good one, and worth further investigation.
Saturday assorted links
1. Why do women travel more than men?
2. Gates and Altman play Desert Island Discs (short YouTube).
3. Try not to let moose lick your car (Canada).
4. Luring top scientists out of universities (NYT).
5. AI will underwhelm, and a six-month pause? And it’s time to learn something serious about private equity. Adam is right throughout.
6. “30% or Portuguese people between the ages of 15 and 39 have left the country.”
7. Ho hum. I’ve been watching the other post-briefing interviews as well.
World’s First Dominant Assurance Contract Platform
In September I alerted you to a crowdfunding campaign to produce a dominant assurance contract/refund bonus platform. Many of you stepped up and it’s now and up and running! The platform is called EnsureDone. It’s starting up small, with just a few projects, but already the projects are quite interesting. MakeSunsets, for example, had a campaign to raise $1000 to fund a test of seeding the atmosphere with sulfur to increase reflection. That campaign failed which meant the people who had agreed to contribute earned a refund bonus! The UX could also use some work. Still, it’s nice to see this idea being tested in the wild and I have inside info that another such platform will launch soon.
The economics of semaglutide
…while one would certainly like to have these drugs available to these patients, if the price of these drugs doesn’t come down, you can make a good argument their cost will literally break Medicare. Why? Well, roughly 35% of Medicare patients are overweight or obese. Roughly 75% of people over 65 have coronary artery disease. (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6616540/#:~:text=The%20American%20Heart%20Association%20(AHA,age%20of%2080%20%5B3%5D.)
Even if you assume these numbers are independent of each other, which is very unlikely, because overweight people are more likely to have cardiovascular disease, this means at least 26% of the roughly 31 million Medicare recipients (and likely more) might benefit from these drugs. We have an eligible population for these drugs therefore of more than 8 million people. At current list prices we would be spending 8mil*1350*12 or about 130 billion dollars a year. Total Medicare spending is about 725 billion a year. So, Medicare spending would go up by more than 17% overnight and stay at that level for a long time to come.
Here is more from Gary Cornell, mostly about the degree of effectiveness.
Economic ornithology
But a hundred years ago, birds were seen as the best remedy for the weeds and insect pests that threatened the country’s food supply and cost farmers hundreds of millions of dollars every year. And in order to identify the precise impact that birds had on agriculture, a field called economic ornithology was born. According to one of its leading practitioners, economic ornithology was “the study of birds from the standpoint of dollars and cents … in short, it is the practical application of the knowledge of birds to the affairs of everyday life.”1 And from the 1880s to the 1930s, birds were widely seen as economic agents, working alongside farmers in the fight against the insect hordes.
By the 1940s, economic ornithology had become discredited and obsolete. Effective and affordable pesticides had entirely replaced the birds’ bug-killing role, while economic ornithologists could never prove that their methods actually increased the number of helpful birds. But before their role in agriculture was dismissed, there was a time when we believed that we depended on birds for our food, and for our very survival.
And here was the method of economic ornithology:
In 1916, Gilbert Trafton summarized the primary approach used by economic ornithologists: “The practical value of birds to man, whether helpful or harmful, depends chiefly on their food habits,” and by examining what they eat, “the exact economic status of a bird is determined.” Sometimes this was done by observing the behaviors of birds in the field, but it usually meant dissecting birds and seeing what they had in their stomachs.
Here is the full Substack, by Robert Francis. Via Philip Wallach.
Friday assorted links
1. Those new service sector jobs, former chess player edition.
2. John O. McGinnis reviews GOAT.
3. Huge ancient city found in the Amazon.
4. Michael Magoon on progress-related Substacks.
5. Elaine Schwartz has been blogging every day for ten years at Econlife.
6. Things you learn dating Cate Hall. And Cate’s essay on how to be more agentic.
*The Economist* on Tatu City, Kenya
Take the pig’s ear first. Unveiled in 2008 as a $15bn smart city project, Konza Technopolis was supposed to be the heart of Kenya’s “silicon savannah” that, by 2020, would create 100,000 jobs and add 2% to gdp. Three years and many missed deadlines later, there is still far more evidence of savannah than silicon.
By contrast, Tatu City, on the northern outskirts of Kenya’s capital, Nairobi, is flourishing. Some 23,750 people already live, study or work there and 78 businesses have made it home. Moderna, an American drugmaker, is opening a $500m vaccine manufacturing facility, its first in Africa. Zhende Medical, a Chinese medical-supplies manufacturer, is also setting up shop.
Tatu and Konza were conceived at the same time. Each, at roughly 5,000 acres, is of a similar size. Both aspire to house populations of more than 200,000 people. And both have been designated Special Economic Zones (sezs), meaning that the businesses they house are eligible for tax benefits and other incentives. Why is one more likely to succeed than other? The answer lies not in their similarities, but in their differences, and these provide lessons for other developments in Africa.
The first is ownership. Konza’s proprietor is the state. Tatu City’s is Rendeavour, a big private urban land developer.
Here is the rest of the article. Here are my three CWT podcasts about Tatu City. I am pleased at how well it is all going!
For the pointer I thank Kurtis Lockhart of Charter Cities Institute.
My Conversation with the excellent Patrick McKenzie
Here is the audio, video, and transcript. Here is part of the episode summary:
Tyler sat down with Patrick to discuss signature fields on the back of credit cards, whether bank tellers or waitstaff are more trustworthy, the gremlins behind spurious credit card declines, how debt collection and maple syrup heists should change your model of the world, Twitter’s continued success as the message bus for government and civil society, crypto vs traditional money transfers, the intended desolation of bank parking lots, why he moved to Japan and how it affected his ambition, why Tether hasn’t collapsed, the internet as a Great Work, how he’s experiencing reverse culture shock after returning to the US, what he’ll learn about next, and more.
Excerpt:
COWEN: How did the maple syrup heists change your overall view of the world and of humanity? What’s the model update?
MCKENZIE: Going back to “evil is an actual thing in the world,” I often assume that large, well-regarded, professionalized industries have a rate of crime associated with them which rounds to zero. I would assume agriculture is relatively heavily regulated. Maple syrup is an agricultural product. The only buyers of maple syrup by the containerful are “real companies.” Therefore, I would expect rounds-to-zero supply-chain fraud in maple syrup.
But it turns out that the amount of supply-chain fraud in maple syrup is actually quite higher than zero. The chain of custody for cars of maple syrup is much less regimented than you would expect that chain of custody to be. There are smaller buyers of maple syrup — not on the scale of you or me, but on the scales like a boutique maple syrup refinery — who just don’t do as many checks as the large producers or suppliers, et cetera, do. So, it is possible to steal millions of dollars of maple syrup and sell it on the black market, which blew my mind.
This rhymes with how Tide, the well-known detergent in the United States, is used as a commodity on the criminal and semi-criminal/system deal, less criminal undergrounds, in that Tide can be resold. Online marketplaces are getting a lot of the attention, but the traditional way to use Tide as a currency for people who have less access to dollars was to resell the Tide to a corner bodega.
Since the corner bodega can easily verify that the Tide hasn’t been tampered with because the Tide seal is on it, and has basically unlimited willingness to do small-dollar transactions that make money, that was a way to recycle Tide which had been stolen — or which had been acquired by means that were not stealing, but also not exactly the official way of getting Tide — and recycle it into the more or less legitimate economy, insofar as we think that the corner bodega is more or less legitimate, which I think is a complicated story.
And:
COWEN: Why are bank parking lots usually so empty?
MCKENZIE: Oh, this is a straightforward result from queuing theory, if you can believe it. Queuing theory/operation science, by the way, tremendously under-understood by many people to whom it is professionally relevant, be that as it may. When you think of your typical stop-in at a bank, you go in, perhaps you deposit a check, perhaps you talk to the teller. You probably think, “I will have three minutes of dwell time.” So, you expect to be in and out. But the thing that the bank really wants to optimize for is new account opening.
New account opening requires 30 to 60 minutes plus of dwell time, depending on what type of account you are opening. Then when you back out that variability, it turns out that gratuitously over-provisioning the parking space almost all of the time maximizes for not losing new account opening at a few very limited windows per year. Those very limited windows can make or break the branch’s ability to contribute to their larger financial institution.
The number of accounts a bank actually opens per year in terms of checking accounts per branch — that is a number that one can have access to in various places. Depending on the bank and the locality, et cetera, it’s between 200 and 500. So, if you turn away a single customer or two customers in a row to open accounts because you had three parking spaces where you could have had seven, you’ve done a very outsized amount of damage to your financial institution, and thus over-provision the parking in all the places.
Definitely recommended, interesting throughout.
The efficiency of counterfeiting in 19th century Peru
“The counterfeiter probably didn’t realize that that coin didn’t exist,” Dr. Ortega said.
He said that an influx of low-value coinage would have been welcomed in Peru at the dawn of the 20th century. The country’s economy was reeling from the recent War of the Pacific, and the government was focusing on printing larger-denomination paper bank notes to pay off international loans; in 1899, the Lima Mint produced roughly one-tenth the number of silver coins it produced just five years earlier.
As a result, people in Peru were using coins from neighboring nations or even cutting their own country’s coins in half to conduct small transactions. “Counterfeiters found a field of opportunity,” Dr. Ortega said.
Here is more from Katherine Kornei at the NYT, mostly about finding an old coin that should not exist and cannot be traced.
Thursday assorted links
1. Claims about the Russian defense sector.
2. South Korea to ban dog meat.
3. Microsoft debates what to do with AI lab in China, perhaps EAers should discuss this more? (NYT) EA is right about many issues, but for far too long foreign policy has been a weak suit of the philosophy. And here is U.S. companies talking to China about AI safety (FT).
4. Johor Bahru metro concept map. And “Malaysia and Singapore agreed to jointly develop a special economic zone and explore a range of measures including passport-free travel to boost trade between the neighbors that each count the other as the No. 2 trading partner.” (Bloomberg, this has been a longstanding prediction of mine).