Category: Ec101

Mark Zuckerberg Teaches Us the TMKK of Game Theory

Why is Llama open sourced?

We’ll get to the answer in a bit, but just in case you need help understanding what Llama is:

Llama (Large Language Model Meta AI) is a family of autoregressive large language models (LLMs), released by Meta AI starting in February 2023.
On April 18, 2024, Meta released Llama-3 with two sizes: 8B and 70B parameters. The models have been pre-trained on approximately 15 trillion tokens of text gathered from “publicly available sources” with the instruct models fine-tuned on “publicly available instruction datasets, as well as over 10M human-annotated examples”. Meta plans on releasing multimodal models, models capable of conversing in multiple languages, and models with larger context windows. A version with 400B+ parameters is currently being trained

So what, you might say. There’s the OG ChatGPT, there’s Claude, there’s Gemini… so one more comes along. Ho hum.

Well, if you say that, you’d be very, very wrong.

Why would you be wrong?

Because of this:

In contrast, the most powerful LLMs have generally been accessible only through limited APIs (if at all), Meta released LLaMA’s model weights to the research community under a noncommercial license

Why does this matter? Because, as our good friend Claude explains:

“The release of the Llama model by Meta under a noncommercial license is a significant development in the field of large language models (LLMs) and artificial intelligence more broadly. Here’s why it matters:

  1. Accessibility: Before this, the most powerful LLMs were usually kept secret by the companies that made them. For example, OpenAI’s GPT-3 model could only be used through a paid interface, like a vending machine you put money into to get a result. By releasing Llama’s “model weights” – essentially the knowledge the AI has learned – Meta has allowed researchers and hobbyists to experiment with and build upon a cutting-edge language model. It’s like they’ve given away the recipe for a powerful tool, not just limited access to using the tool itself.
  2. Democratization of AI: Restricting access to top LLMs meant that only a handful of big corporations could really use this powerful technology. Imagine if only a few factories could use electricity. An open-source model changes the game significantly. It empowers way more people to explore creative uses of language models and lowers the barriers to innovation in this space. It’s like the difference between a few people having libraries versus everyone having access to all the books.
  3. Cost: Using LLMs through paid interfaces can get expensive quickly, putting them out of reach for many. It’s like having to rent a supercomputer every time you want to use one. With access to the model weights themselves, people can run the model on their own computers, dramatically reducing costs. This opens up experimentation to students, researchers, startups and others with limited budgets.
  4. Customization: When you can only access a model through an interface, you’re limited to pre-defined uses, like ordering off a set menu at a restaurant. Having the actual model provides much more flexibility to tailor and fine-tune it for specific applications and domains. This could lead to an explosion of niche language models customized for particular industries or use cases – imagine a model specifically trained to understand and generate legal jargon, or one tuned for writing poetry.
  5. Reproducibility and Transparency: In scientific research, it’s crucial to be able to reproduce results. Using an API is like a black box – you can’t see how the model works under the hood, you just get the output. With the model weights, the exact workings of the model can be scrutinized, enabling more robust research and understanding of how these models function. It’s like being able to examine the engine of a car instead of just looking at the exterior.

Model weights are the key to how a language model works. They’re essentially the “knowledge” the model has learned during training. In a neural network (which is what most modern language models are), the weights are the strength of the connections between the neurons. These weights determine how the model responds to a given input, like how a brain’s neural connections determine a person’s response to a question. By releasing the weights, Meta has provided the “source code” of their model, allowing others to understand how it works, modify it, and use it for their own purposes.

While the noncommercial license does place some limits on how Llama can be used (you couldn’t start a company selling access to it, for example), the release of the model is still a major shift in the AI landscape that could have far-reaching effects on research, innovation, and accessibility of this transformative technology. We’re likely to see a proliferation of new applications and rapid progress in natural language AI as a result.”

You don’t just get the dish to eat, as Claude puts it, but you get the recipe so that you can try and recreate (and modify) the recipe at home. Not all of us have specialized cooking equipment at home, but those of us who do can get cooking very quickly indeed.

Speaking of cooking, have you seen this excellent series from Epicurious called 4 Levels? Chefs of varying expertise (home cook to the pros) are invited to cook the same dish, but with varying levels of expertise, ingredients and equipment.

Source

That’s what the 8 billion, 70 billion and 400 billion parameter models are all about. Same idea (recipe), but different capabilities and “equipment”.

But why do this? If Gemini, Claude and ChatGPT are giving away basic versions for free and premium versions for 20 USD per month, then why is Meta not just giving away all versions for free… but also giving away the recipe itself?

Because game theory! (Do read the tweet linked here in its entirety, what follows is a much more concise summarization):

  1. You can get janta to do the debugging of the model for you.
  2. If social debugging and optimization of models makes AI so kickass that AI friends can replace all your friends, then who owns the technology to make these friends “wins” social media. Nobody does, because janta is doing the work for “everybody”. So sure, maybe Mark bhau doesn’t win… but hey, nobody else does either!
  3. The nobody else does point is the really important point here, because by open sourcing these models, he is making sure that Gemini, Claude and ChatGPT compete against everybody out there. In other words, everybody works for Mark bhau for free, but not to help Mark win, but to help make sure the others don’t win.

The economics of AI is a fascinating thing to think about, let alone the technological capabilities of AI. I hope to write more about this in the coming days, but whatay topic, with whatay complexities. Yay!

All this is based on just one tweet sourced from a ridiculously long (and that is a compliment, believe me) blog post by TheZvi on Dwarkesh’s podcast with Mark Zuckerberg. Both are worth spending a lot of time over, and I plan to do just that – and it is my sincere recommendation that you do the same.

Harmunia Mode.

Is an inheritance tax a good idea or a bad idea?

Well, it depends.

What does it depend on?

It depends on whether you are a Congress supporter or a BJP supporter.

What else does it depend on?

It depends on whether you are answering this question in 2024, 2019 or 2014.

What does it not depend on?

Economic analysis.

I have read innumerable tweets/messsages today about the inheritance tax. Or death tax. Or estate tax. These are tweets and messages from economic analysts, politicians, think-tankers, journalists and policy makers.

Every single discussion has been about who has said what about inheritance taxes, and when.

Not a single discussion has been about an economic model that underpins the rationale for or against the inheritance tax.

  • What is it about India’s tax buoyancy that justifies (or doesn’t justify) an inheritance tax?
  • What about her demography? Does that have a role to play?
  • What about the loopholes in the Income Tax Act? Do they have a role to play?
  • What about her direct and indirect tax collections?
  • What about threshold limits for imposing such a tax?
  • How and why have other countries done it? Can we learn from their mistakes? Can we learn from their successes?
  • What are the alternatives?
  • If we do implement it, do we keep it in perpetuity?
  • How much money do we potentially raise if we implement it?
  • What might be the downsides?
  • How do we judge the quality of implementation if we implement it?

But no, let’s see who said what and when, call each other names, and prove that we are right and they are wrong. That’ll show them. Bloody losers. It’s because of folks like them that India is what it is today.

https://help.twitter.com/en/using-x/x-advanced-search

Do you want to play the game yourself, and show to members of your tribe how you were right and they were wrong? Click on that link, learn how to search on Twitter by date and by handle, and prove how Your Side Alone Speaks The Truth.

If, on the other hand, you are interested in, y’know, actual economic analysis, allow me to recommend Atkinson-Stiglitz and Diamond-Mirrlees, both from the 1970’s. It’s been a while since I’ve read them, and it’s not a topic I’ve ever particularly enjoyed, but I assure you that these are good places to begin.

But Twitter takedowns are faster and more fun, so there’s that.

As for me and what I think, I think that not only was Amit Varma right about the IPL, but if you ask me, he’s right about the harmunia too.

What Are You Optimizing For, The MKBHD Edition

There is this thing called the Humane Pin. Maybe you’ve heard of it?

Humane, Inc. (stylized as hu.ma.ne) is an American consumer electronics company founded in 2018 by Imran Chaudhri and Bethany Bongiorno. The company designed and developed the Ai Pin, which started shipping in April 2024.

And by all accounts, if you haven’t heard of it, you aren’t missing much. It’s been panned widely in the press, and for a variety of reasons. In this case, in fact, don’t depend on me to provide a link. Just run a search for a review of the product, and click on pretty much any review that you like. Chances are that the review will say that it isn’t worth buying – for now, at any rate.

But one particular review, by MKBHD, has gone viral. We’ll get into the why of it, but for now, take a look at the review:

This review got the attention of lots of folks on Twitter, including this person:

I usually embed tweets, rather than post screenshots. So why a screenshot today? Because:

It is worth flagging that Vassallo sells a Gumroad course on Twitter growth, so there is a non-zero chance he crafted this post to get engagement whether or not he actually felt strongly about MKBHD’s review (and it worked).

(Note that I copied the screenshot from the same blog this quote is from).

Incentives matter, children. Not every take on Twitter is genuinely held. Some folks just crave the attention. That may or may not have been the case here, but ol’ Thomas really should be your friend here.

MKBHD responded with both a video and a tweet. Now this tweet I don’t mind embedding:

Marques is crystal clear about who he is optimizing for. He isn’t optimizing, he says, for pleasing product companies, or advertisers, or anybody else. He is optimizing for giving his viewers his honest recommendation and feelings about a product. The more he does this, and the better he gets at doing it, he is saying, the more he is likely to be rewarded for it.

And it’s not like I’m an expert at the YouTube game (and that’s an understatement, trust me) – but I would think 18 million subscribers means he’s doing kinda ok. Not bad.

TMKK?

No matter what you choose to do in life, ask yourself this one question ad infinitum:

What Are You Optimizing For?

The clearer you are about the answer to this question, the easier your job is. You don’t even have to trust me on this one – MKBHD says so!

John Gruber put it best:

There’s cool, and then there’s cool.

Opting Out of Opt Ins

I’m conflicted about using Zepto (or its substitutes) for a variety of reasons, but these days, during the summer, I think it is a Most Magical Thing. The ability to order pretty much anything you want, from the comfort of your home, is a wonderful thing – and especially so when the world outside resembles the middle of a 250 degree oven.

And so when Zepto introduced Zepto Pass, I was more than happy to sign up:

Quick commerce firm Zepto on Thursday officially announced the launch of its paid subscription service—Zepto Pass—that will offer greater discounts to consumers, ratcheting up competition in India’s growing quick commerce market.

“The company has piloted Zepto Pass with 5% of its user base for a month and seen rapid adoption—almost a majority of orders came from Zepto Pass subscribers within two weeks during the pilot,” it said in a statement.

According to the company, those who subscribed to the Zepto Pass increased their spending on the app by over 30% and showed a 10% increase in monthly retention. The subscription, priced between 19 and 39 rupees per month, for a majority of customers, offers unlimited free deliveries and up to 20% off on grocery items.

There’s lots to talk about with Zepto, and most of all their pricing. For example, at least in my case, free unlimited deliveries aren’t really free. They are free post a minimum order value (one hundred rupees) and the 20% off offer kicks in only if your order is above six hundred rupees.

But as a student of behavioral economics, I found this to be the most fascinating bit:

For folks who know me (and what I look like), the swimming cap isn’t for me.

But that’s not the most fascinating bit.

The most fascinating bit is the fact that delivery is free, but only if I “apply” the “Free Delivery on this order” coupon. In other words, I first pay what I do to get unlimited free deliveries – this is the Zepto Pass.

But that isn’t enough. I also have to then remember to apply a coupon in order to unlock free delivery. In other words, I have to jump through a hoop to get something that I have already paid for.

Why introduce this friction? Which idiot will choose to not opt for this coupon after explicitly paying for it upfront?

The kind of idiot too busy to notice that one additional button has to be pressed, of course. In other words, a person too busy to sweat over the fine print. What’s an additional thirty rupees off a base value that is around two to three hundred rupees?

This way, not only does Zepto get the up front payment for the Zepto Pass membership, but they also get to collect the delivery fee from at least some of the folks who have paid so as to get… free deliveries.

This is, of course, exactly what opt-in/opt-out and sludge are all about.

But knowing what to call it, and understanding why something is done is a far cry from calling it a good practice. This, I’m sorry to say, does not leave a customer with a good impression.

Don’t get me wrong, I’m not writing this post out of outrage, nor am I demanding my money back, or demanding that Zepto be sued.

Far from it.

But would I recommend Zepto to you, or will I be tempted to pay for Zepto Pass next month?

Far from it.

Recursively. That’s The Only Change I’d Make.

The ability to exercise good judgment is the binding constraint in development is the title of Gulzar Natarajan’s blogpost on Oliver Kim’s essay, which we’ve covered earlier here.

Almost all of doing development is about making non-technical decisions (the technical ones are easier, have limited degrees of freedom, and mostly slot themselves into place). Such decisions are invariably an exercise of judgment by taking into consideration several factors, one of which is the technical aspect (or expert opinion). The most important requirement for the exercise of good judgment is experience or practical knowledge. In the language of quantitative science, it’s about having a rich repository of data points that one can draw on to process a decision.

Source: https://urbanomics.substack.com/p/ability-to-exercise-good-judgment

This applies, Gulzar Natarajan says later on in the essay, to “industrial policy and promotion of industrial growth, macroeconomic policy and inflation targeting, and programs to improve student learning outcomes or skills, increase nutrition levels and health care outcomes, improve agricultural productivity, and so on”.

Let’s take one of these and think about it in slightly greater detail: health care outcomes. Let’s do this in the context of India. Answer these questions, for yourself:

  1. What is the best possible health care outcome you would wish for, in India’s context? Define it however you like – everybody should have excellent healthcare so long as they can pay for it is one option. Everybody should have excellent healthcare regardless of whether or not they can pay for it is another option. Everybody should have free healthcare until we reach a per capita income of x dollars (adjusted for inflation and purchasing power parity, if you prefer) is a third option. You can whistle up a million more, and feel free to let your imagination run wild. You get to define the best possible health care outcome for India – setting the standard is up to you.
  2. Microeconomists will call this the indifference curve, and ask you about the budget line. Mathematicians will call this the objective function, and ask you about the constraints. Humans will say “Haan woh sab to theek hai, magar bhaiya, kaise?”. This is the part where we encounter the bad news – what are you willing to give up in order to achieve your best possible outcome? Best possible health care outcome subject to we spending not more than 20% of our GDP on health might be a constraint you choose to apply, for example. Other people may start to froth at the mouth at the thought we spending 20% of our GDP on health, but you do you (for now). Reduce spending on defense, and pensions, and highways, and on education, you get to say (for now). In my little ivory tower, you get to say, I want India to focus on healthcare outcomes, and healthcare outcomes alone – and I’m ok spending x rupees on it. We get to not spend those x rupees elsewhere, of course – remember, opportunity costs are everywhere, including in imaginary ivory towers.
  3. Here’s another way to think about the same problem. You could also say, I’m optimizing not for healthcare outcomes in the short run, but for free markets. I’m not doing this because of my love for free markets per se, but because of my conviction that this path, and this path alone is the only way to deliver the best possible healthcare outcomes eventually. Sure there will be mistakes along the way, and sure some healthcare services will be denied to people who need it the most for now. But eventually, the market will correct all of these errors, and that in ways we simply cannot know right now. Why can we not know them right now? Well, because we are not omniscient. We don’t know what errors will crop up, and we don’t know what solutions will work best for whatever errors will crop up. If we did know this, we could have avoided those problems in the first place, no?
  4. There are, in other words, unseen consequences to Bastiatian solutions as well. That’s just a fancy way of saying there are opportunity costs everywhere, but let me make the point more explicitly: the opportunity cost of an immediate application of a completely free market solution to healthcare is poor health outcomes for at least some folks today. I might be wrong about this, so please, don’t hesitate to tell me the how and why of it.
    For example, let’s say that government stops spending even a single rupee on healthcare (no CGHS, no PMJAY, no ESIC, no Jan Aushadhi, no government run hospitals, no PHC’s, no government run vaccination programmes, nothing) at 12 pm today. Will health markets be Utopian at 12.01 pm, or will they transition to Utopia eventually? How long is eventually? What problematic outcomes will occur along the way, and do we correct for them? How?
    For example, we may learn that poor families in rural Jharkhand now do not have access to healthcare because all government intervention has stopped. Do we do something about this? If yes, what? If not, why?
  5. This cuts both ways, of course.
    For example, let’s say that government doubles its current expenditures on healthcare, at 12 pm. Will health markets be Utopian at 12.01 pm, or will they transition to Utopia eventually? How long is eventually? What problematic outcomes will occur along the way, and do we correct for them? How?
    For example, we learn that corrupt practices when it comes to invoicing in procurement departments have gone up because government expenditure has gone up. Do we do something about this? If yes, why? If not, why?
  6. Given your ideological bent of mind (and we all have one, learn to live with it), you have an urge to say “Ah hah, exactly!” and “Oh, c’mon!” to pts 4 and 5 – in that order. Or to pts 5 and 4 – in that order – it depends on what your ideology is. But if both of those things was what you ended up saying, that’s just you being bad at elementary economics, because there is no such thing as a free lunch, regardless of what your preferred solution is. You can have inequitable outcomes today and therefore a relatively more efficient outcome tomorrow, or you can have equitable outcomes today and therefore a relatively more inefficient outcome tomorrow.
    Equity today and efficiency tomorrow is like those real estate ads offering you high returns and low risk – it doesn’t happen.
  7. Which brings us back, in a very roundabout fashion, to the point that Gulzar Natarajan was making in his post. When he says that “it is not one decision, but a series of continuing, even interminable, decisions”, I interpret it as two different but very related things.
    One, if you’ve chosen to optimize for equity today, you have to be explicit about the fact that you’ve sacrificed optimized efficiency (today and tomorrow). The worst manifestations of these sacrifices must be adjusted for at the margin. And ditto if you’ve chosen to optimize for efficiency today! You have to be explicit about the fact that you’ve sacrificed optimizing for equity (today and tomorrow). The worst manifestations of these sacrifices must be adjusted for at the margin.
    Two, no matter what your favored path is (and I envy you your conviction if you know that your path is The Best One For All, I really do), there will be errors along the way. That’s just life, there will be unexpected surprises along the way. Call it risk, or uncertainty, or whatever you like (and yes, I know, comparing the two is like comparing Knight and day) – but account for the fact that your battle plan will meet the enemy, and it will not survive.
    You must adapt, and said adaptation will involve a series of continuing, even interminable decisions.
  8. Those adaptations will land you somewhere in the middle of efficiency and equity. At which point, you can adapt your will to your circumstances and say you’ve found the truth, or you can continue with your decision-making. It is, after all, interminable.
  9. “Wait, so there’s no end to this?!”, I hear you ask in righteous indignation. “What is the eventual outcome? Or are we doomed to keep making these interminable decisions forever?”. Kids these days, I tell you. They’re just like kids in those days.

And that’s why I say what I did at the start of today’s post. The only change I’d make is the addition of one word:

The ability to exercise good judgment recursively is the binding constraint in development.

Economists Do It In Tribes

… or at least, economists employed by the governments. So says Amol Agrawal in a searing piece that is at once a lament as well as an indictment.

Why is the nature of discourse today so painfully zero-sum? Why do we have a take-no-prisoner approach to discussions, where disagreement is necessarily proof of the fact that the other person isn’t just wrong, but their motives are suspect? There are plenty of hypotheses for why this is so for all of us at large, but Amol shines a spotlight on my tribe, and my tribe is supposed to internalize for themselves and teach the rest of the world that the world is a non-zero sum game.

Except it isn’t. Not any more, and certainly not in the case of economists talking about the economy in India:

The current government economists discredit any critique of economic policy. Each time any analysis/report comes up critiquing the economic policy, the economists rush in to disagree and discard the criticism. The purpose of these articles is not to engage but rebut/attack the institution/writer of the critiques. Shoot both the message and the messenger seems to be the mantra.

It is also highly fashionable to draw comparisons with earlier eras and say how bad things were back then. They forget it has been ten years of the current government and people are asking questions on the current economic policy. They also forget how they themselves critiqued economic policy and built their own careers. One is also amazed how the media whose job is to critique economic policy, allow so many one-sided articles.

https://mostlyeconomics.wordpress.com/2024/04/11/economists-working-with-the-government-what-has-changed/

Disagreements are not just “fine”, they are the point. When you and I look at a slice of the world and come away with different conclusions, it is because we bring a different perspective, a different methodology, a different set of facts to emphasize and analyze,and a different ideology.

All of these things are true, not just the last one.

For all of us to sweep away the different conclusions, perspectives, methodologies and sets of facts under the carpet, and pin the differences on ideology alone is a tragedy with far reaching consequences.

There are people who will oppose the current government on ideological grounds alone (alone, in this case, is used in this sense: indicating that something is confined to the specified subject or recipient). And likewise, there are people who will defend the current government on ideological grounds alone. That is just the world we live in, and these messages will get amplified and shared more than they should.

But for all of us to behave in only this manner is a society that no longer talks to each other. It is a society that is divided along deeply tribal lines, and with every passing day, those lines get deeper and more permanent.

The hardliners on both sides – on one side are those who critique the government and and on the other those who defend it – will say that the other side started it first, and they had no choice to respond. They will also say that the other side is worthy of this kind of behavior and ostracization, because the other side is evil, and needs to be destroyed for our version of this country to flourish.

Bullshit.

Allow me to labor the point:

When you and I look at a slice of the world and come away with different conclusions, it is because we bring a different perspective, a different methodology, a different set of facts to emphasize and analyze and a different ideology.

We would do well to not ignore all of these points. Hanging the weight of the world on just the one word, regardless of which side does it (or did it first), does nobody any good.

No government in independent India’s history has been uniformly bad. Nor has any government been uniformly good. You and I will (and should!) have opinions about which government was the best, which was the worst, and which lay somewhere in the middle. You and I will try to convince the other of why we say what we do. And you and I will reach some sort of an agreement, or at least an appreciation of why the other person thinks what they do. Disagreements are food for thought, not excuses to launch personal attacks.

That this needs to be said is a matter of shame for everybody, but especially for social scientists, and doubly so for economists. (Yes, I hold my tribe to a higher standard).

Economics is about three things:

  1. What does the world look like?
  2. Why does the world look the way it does?
  3. What can we do to make the world a better place?

“Better” is tricky because better is subjective.

“We” is all of us, those who defend and those who critique the government.

So if I say (and I do) that the census not having been conducted is a problematic thing, I say it because I think it is a problematic thing. The truth value of that statement isn’t only a function of the fact that I am saying it, or that I am saying it in a publication that you don’t like, or what my political affiliations or economic ideology are.

I use the census thing as an example. Replace “census not having been conducted” with “improvements in our airports”, and replace “is a problematic thing” with “is a wonderful thing” for the same take, but from the other side.

If your Pavlovian response to the census thing is whataboutery, or if your Pavlovian response to the airports thing is whataboutery, then you have a problem. Sure, bring up the fact that the pandemic was a factor. And likewise, sure, bring up the fact that oligopolies are a problem. But don’t decide that the statement is wrong as a function who is saying it – decide the truth value on the basis of the statement, not the person behind the statement.

And one final point, to circle back to Amol’s post.

Criticizing the government is not just fine, it’s not just OK, it’s what economists will do. We will do it because we want the world to be a better place.

Economic policy should not be limited to criticizing the previous governments and praising the current government. The policy should lay a framework to improve the economic conditions of the people. It should not just agree to the government decisions but caution the government against missteps. That is how we saw things and admired all the economists who have served the governments all these years.

https://mostlyeconomics.wordpress.com/2024/04/11/economists-working-with-the-government-what-has-changed/

Kudos to Amol for saying what he did, and I look forward to reading more from him about what the government, and its economists, can do better. The fact that he (and I, and so many others) critique their work isn’t proof that the work of the government or its economists is bad. Nor is it proof that we are evil. It is our attempt to help make the world a better place.

Now, please tell me why you think I’m wrong, and let’s have a debate about it.

Very underrated thing to do in 2024!

A Fine Unbalance

“The person we are talking about was born in Germany, in 1915. He took part in anti-Hitler protests in the early 1930’s, and had to flee to Paris as a seventeen year old to escape persecution back home. He later attended lectures by Lionel Robbins and Hayek while in London, and also participated in the Spanish Civil war. In the interim, he also helped thousands of Jewish refugees escape Nazi persecution in France – among them, Hannah Arendt, Marc Chagall and Marcel Duchamp. He managed to do all this at or before the age of twenty-five. He would go on to become one of the most famous development economists of the 20th century.
Who are we talking about?”

You might know the answer to this question if you are a very good quizzer, but you may not know the answer even if you are a very good economist. And that’s because of two reasons.

One, Albert O. Hirschman isn’t as celebrated as he should be. Consistently underrated, you might say.

And second, we in the economics teaching profession don’t like to tell stories about economists. We like to bore people to death with equations and models, but making the economist behind the theories come alive? Doesn’t happen nearly as often as it should, if you ask me.

Watch about a minute or so of this excellent interview from around the 30 minute mark. As always, please watch the whole thing, it’s a great interview – but for the purposes of today’s post, just for about a minute or so from the 30 minute mark.

And Alex is bang on when he says that knowing that the person is French will often tell you more than the fact that the person is an economist. I’d go a step beyond and say that it is not just the location, but also the time that matters. A French person born around 1930 will be a very different person from a French person born in 1980, for example. This isn’t about who is better or worse compared to the other, this is about understanding why those people created the things that they did. And part of this understanding comes from understanding the time and place of their birth. Not just knowing the time and place, mind you, but understanding it.

And to understand why Hirschman was the kind of economist he was, you need to understand where he came from, all of what he experienced in his formative years – and the cultural milieu that surrounded him when we was an economist.

And it is for this reason that reading this blogpost about Hirschman’s work, but also his biography is ever so illuminating. You don’t just get a sense of Hirschman’s central ideas, but you also get a sense of how events in his life formed his worldview, and possibly influenced some of his decisions later on.

He encourages us to see the inevitable pitfalls and stumbles of the growth process not as disappointments, but as opportunities, and gives us a conceptual language to identify them. For randomistas-in-training, steeped in the world of deworming and bed-nets and pre-analysis plans, Hirschman also reminds us that we need to step back from individual interventions more often and think more about development strategies–not just how different projects can complement each other, but also how each project might organically summon market and non-market forces to help growth along.
The Credibility Revolution has yielded, perhaps for the first time, robust evidence for individual program effects. The time is ripe, not to copy Hirschman’s ideas wholesale, but to borrow his clear-eyed approach and think carefully about how projects can be brought together, pressure point by pressure point, into programs for sustained development.

https://www.global-developments.org/p/the-real-development-was-the-friends

This is worth doing for everybody of note, of course, not just Albert O. Hirschman – but if you are a fan of studying the development of the field of development economics, this would be a great place to start.

Lots of homework in today’s post:

  1. Watch Transatlantic (it is available on Netflix)
  2. Read The Worldly Philosopher, by Jeremy Adelman
  3. Read The Strategy of Economic Development, by Hirschman

Yamini Aiyar Asks a Question, and We Try to Answer

First, the question. We’ll get to who the “we” in the title is (or should it be “are” instead of “is”?) in a bit, and also to our answers.

Ideally in a democracy, there ought to be space for evidence-based partnerships with government whilst simultaneously holding the mirror. But when the space for holding the mirror shrinks, when freedoms are trampled upon, what should the public policy professional do? There is a real risk that the pressures of relevance can, and indeed do, push researchers to blunt critique, to inadvertently, perhaps, stop asking difficult questions and refrain from critical public engagement. Is there a need then to redefine our role, to question the narrow prism of relevance and impact that we judge ourselves by?

https://www.deccanherald.com/opinion/on-evidence-policy-making-and-critiquing-it-in-a-polarised-polity-2958920

Say Hello to ArreBhaiWah

Paul Krugman has a textbook on international economics. Standard stuff, and quite a good textbook, running into multiple editions. I may be wrong over here about the specific topic, but I think it is in the context of national accounting with international trade that Paul Krugman asks us to imagine a country called Agraria.

In much the same vein, but for entirely different reasons, I’m going to ask you to imagine a country called ArreBhaiWah.

Because when it comes to India, we will have to spend a significant amount of time having heated debates about whether the space for holding the mirror has shrunk or not. I might (and do!) say that yes, it has shrunk. You, on the other hand, might say that it has expanded instead. And then we will argue and call each other names and get applauded for having reminded the other side of their grandmothers. That is fun to do, but would not be constructive, nor productive.

So let us, instead, focus on ArreBhaiWah.

If you are a public policy professional in ArreBhaiWah, and you see that the space for holding the mirror has shrunk, and you need to analyze what you should do about it – what framework should you use to arrive at your answer?

This isn’t about answering the question for ArreBhaiWah, you see. As with many posts on EFE, it is about supplying you with a framework to think about the problem. Please decide for yourselves whether the question makes sense, is applicable and finally, what your answer (if any) should be.

Which brings us to who the “we” in the title of this post are (or should it be “is” instead of “are”?). It is yours truly, and the late, great A.O. Hirschman. By the way, both Yamini’s piece and this Wikipedia article about Hirschman deserve to be read in their entirety, so please do.

Exit, Loyalty and Voice

Here’s ChatGPT’s summary of one of my favorite books in economics:

“Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States” is a seminal book by Albert O. Hirschman, published in 1970. The work presents a framework for understanding how people respond to dissatisfaction with organizations or states they are part of. Hirschman’s theory is built around three main concepts: exit, voice, and loyalty, which are mechanisms through which individuals can express their discontent and potentially influence change.

Exit: The option of leaving the organization or ceasing to use its products or services. This is a common response in economic markets; for example, if a customer is dissatisfied with a product, they can simply stop buying it and switch to a competitor. Exit is a powerful mechanism in promoting quality and efficiency due to the competitive pressure it creates.

Voice: The option of actively expressing dissatisfaction and seeking to improve conditions from within, rather than leaving. Voice can take many forms, including direct feedback, protests, or any attempt to change the organization’s practices or policies. Voice is particularly relevant in scenarios where exit is not feasible or desirable, such as in monopolies or with state governance.

Loyalty: Loyalty plays a moderating role in the exit and voice framework. It refers to a person’s attachment to an organization, leading them to endure dissatisfaction while trying to improve the organization through voice, rather than exiting. Loyalty can delay exit and give voice a chance to work, as loyal members or customers may seek to solve problems internally rather than abandoning the organization.

Hirschman’s framework is used to analyze a wide range of economic, political, and social phenomena. It provides insights into how organizations and states can deteriorate or improve over time based on the feedback mechanisms available to and utilized by their members or constituents. It also highlights the importance of maintaining the right balance between allowing exit and encouraging voice to ensure the health and adaptability of organizations and societies.

ChatGPT4

And this is an excellent framework with which to answer Yamini’s question: in ArreBhaiWah, what should the public policy professional do?

Should they choose exit, or voice? To what extent should loyalty influence your decision? Remember, loyalty refers to a person’s attachment to the organization (or the entity) in question.

This is where I step in.

What Are you Optimizing For?

  1. Should you be optimizing for what is best for you?
    • Best for you in a professional context, or personal context?
    • Best for your conscience, or best for your career?
  2. Or should you be optimizing for what is best for ArreBhaiWah?
  3. Or should you be optimizing for what is best for the folks who currently run the government in ArreBhaiWah?

Because as Khyati Pathak, Pranay Kotasthane and Anupam Manur point out in their excellent book We, the Citizens:

Source: We, The Citizens, pg 19

So is your loyalty to the government, or to the nation? Or are you of the considered opinion that the two are the same thing? They are manifestly not, by the way, so even if it is your considered opinion that they are the same thing, please do read Chapter 2 from the book, We, The Citizens (and the rest of the book, while you are at it!)

As with everything else in life, there are only trade-offs, and no solutions. There isn’t an easy way to answer this question that Yamini asks, alas. Optimizing for any one of oneself, ArreBhaiWah or its government also implies not optimizing for everything else.

But there you have it: the framework that one should use while thinking about the answer to Yamini’s question in the context of ArreBhaiWah.

What should her answer be, you ask? Why, that is Yamini’s business and no one else’s, surely. Allow me to wish her luck for what lies ahead, and to thank her for her work thus far. You and I may have disagreed with her about some of her conclusions, but that, I would argue, was part of the point.

Voice is currently underrated!

Scalars, Vectors, Incentive Design and McKinsey

… not to mention the horror that is poorly done econometrics. On my bad days – and this may well be one of them – I often end up wondering if the phrase “poorly done” is a redundant one in the context of econometrics. But I’m getting ahead of myself.

For the first course, a tweet:

I came across this tweet via The Zvi’s substack, and based on all the posts that I have read so far, I’m happy to recommend that you subscribe.

For salad, two definitions:

You probably know (or in the case of at least some of you, have blessedly forgotten) what scalars and vectors are – but in either case, here is a quick refresher, courtesy ChatGPT:

“Much like a thermometer measures one thing, and one thing only (the temperature), performance in a hedge fund is measured by one dimension: profit. It doesn’t inherently come with a “good” or “bad” direction; it just shows how much or how little.

This could be contrasted with a vector, which might represent the varied ways an organization or individual measures success. For example, a nonprofit might measure success not just by funds raised (the magnitude) but also by social impact (the direction). Here, performance isn’t just about a number; it’s about a number going in a specific direction towards a specific goal.”

(Note that I have tweaked the original answer a little bit to make it a bit more readable)

For the main course:

…the start of a recently published academic paper (h/t @realChrisBrunet over on Twitter), and two charts (hey, this is the main course! I hope you packed an appetite):

And for dessert:

First, an explanation of reverse causality (courtesy ChatGPT):

“Imagine you notice that when ice cream sales go up, the weather tends to be hotter. If you conclude that “Buying more ice cream causes the weather to get hotter,” you’re assuming that the ice cream sales are causing the hot weather. But actually, it’s the other way around – the hot weather is causing more people to buy ice cream. This mix-up is what we call “reverse causality.” It’s when we mistake the effect (hot weather) for the cause (increased ice cream sales), rather than recognizing the true cause (hot weather) for the effect (increased ice cream sales).

So reverse causality is all about getting the direction of cause and effect backward. Instead of A causing B, it’s actually B that is causing A, just like if someone thought that wearing shorts made the day sunny, when in reality, it’s the sunny day that prompts people to wear shorts.”

Does growth lead to more diversity based hiring, or does more diversity based hiring lead to growth?

And second, Cowen’s First Law:

“Cowen’s First Law: There is something wrong with everything (by which I mean there are few decisive or knockdown articles or arguments, and furthermore until you have found the major flaws in an argument, you do not understand it)”

And surely an espresso is called for after such a big meal:

How does Cowen’s First Law apply to this blogpost?

Top Down Impositions of Cultural Norms and China’s Demography

The policy was both cruel and a blunder. Family-planning officials assumed that birth rates would spring back once controls were abolished. Alas, they re-educated parents too well. One child became the norm, certainly in cities. Consider another figure that should haunt leaders: 1.7. That is the number of children that, on average, Chinese women of child-bearing age call ideal. China’s ideal is one of the world’s lowest, far below the number given in Japan or South Korea. Chinese women born after 1995 want the fewest of all: 48.3% of them told the Chinese General Social Survey of 2021 that they desire one or no children. There is growing evidence that such attitudes are powerfully shaped by how people, and those around them, experienced the one-child policy.

https://www.economist.com/china/2024/03/21/chinas-low-fertility-trap

“The policy” refers to , of course, the infamous one child policy. Please read the entire article (assuming you can, because The Economist articles are often behind a paywall) – it is an eye-opener in many ways.

But my primary takeaway is how difficult it becomes to change culture, once it has been imposed. The last sentence of that excerpt is the worrying one, and not just in the case of China and the one-child policy.

Attitudes are indeed shaped by how experiences one has while growing up. And in the case of China and her demographics, it looks as if attitudes about families and (the number of) children will not be changing in a hurry, if at all.

India is, at best, twenty years away from where China (and most, if not all, of East Asia) is today. All the more reason to worry about India’s growth rates over the next two decades – her current demographic dividend is not a gift that will keep giving forever.