In response to footnote 4, a land value tax that was assessed against the market value of the land would not unfairly punish landowners banned from building, since the ban on building would proportionately diminish the market value of the land.
Land value taxes are meant to be assessed on the market value of the land itself, independent of whether there are buildings on it. So that value should stay the same whether or not there's a skyscraper on the land. But then the people with the skyscraper pay the same as their neighbor who only has a single-family house on equally-sized land.
In other words, you're right that the building regulations diminish the market value of unimproved land, but LVT is meant to estimate the value of *all* land as if it's unimproved land, so this doesn't help distinguish them.
I don't think this is right. The unimproved land value will depend on the property rights one obtains alongside it.
To illustrate, consider one of the Georgist schemes for estimating the market value of the land: the lot owner declares a "land value" to be taxed against; any member of the public can buy the land from him above the declaration price (if they promise to tear down any improvements).
Suppose a lot owner lives in a single-family house in an unzoned lot inside a busy urban area. Now suppose mayor NIMBY replaces mayor YIMBY and zones his lot as single-family: the lot owner can reduce his "land value" declaration, since previously, the potential buyers would have valued his land using the sum of future skyscraper rents; now the buyers can only earn single-family house rents.
Yes, that all seems right. But if we're using this mechanism, the owner of a skyscraper on an equivalent land plot would be able to declare the same land value, since the buyer would have to tear down the skyscraper and start from scratch, which they wouldn't want to do. Yet for that tax bill they can get all the benefits of having a skyscraper on that land. So my point remains that the tax burden is unfair to people who didn't manage to get skyscrapers built before burdensome regulation, compared with those that did.
Maybe I initially misunderstood your point; would it be fair to characterize it as:
"There is some inherent unfairness in taxing a skyscraper-owner the same amount as an adjacent single-family homeowner, perhaps in the same way it might be unfair to tax a CEO the same amount as a janitor. Relative to a property tax, an LVT imposes this form of unfairness, but has the benefit of not disincentivizing land improvements. However, if in some contexts land improvements are de jure / de facto banned, then the relative benefit of the LVT is negated; in those contexts a property tax and an LVT are equally efficient, so the property tax should be preferred on fairness grounds."
I think a Caplan-style libertarian econ-brained person would simply dispute the premise that taxing the skyscraper owner the same amount is unfair, and an Alexander-style utilitarian econ-brained person would dispute the idea that fairness matters intrinsically.
Brought to mind Adam Bulter's recent Substack, The Strip Mining of Trust. Shared insight: Economic optimization systematically consumes the non-economic substrate on which economic activity depends.
The convergence is striking given your different backgrounds (institutional finance vs decision theory/AI safety). Butler calls it "strip-mining trust." You call it "undermining sociopolitical autonomy." Both describe the same phenomenon—the systematic conversion of accumulated social capital into short-term legible gains.
Suggest you two might want to compare notes as you develop your theses.
I'd consider these people as applying non-economic frameworks to understand economics. I think virtue is not yet an economic concept in the same way as a concept like surge pricing is.
I should also clarify that economists themselves are often not very econ-brained when thinking about other disciplines (and even when thinking about parts of their own discipline—hence why so few economists are libertarians). Econ-brain involves an attempt to be intellectually consistent across domains which is rare (but disproportionately common amongst the most influential intellectuals).
To me this sounds like saying "may economics burn in hell". Basic economics (and rational choice theory) is good and useful, as long as you apply it appropriately (and in particular don't try to make it into an eternalist ideology, as per https://meaningness.com/eternalism).
LVT has always, since Henry George himself, been one half of a policy package, with the other part being what Henry George called the "citizen's dividend", most people today would call UBI (though that wouldn't be precise), and I call Universal Natural Income (UNI).
Georgist policy proposals are rooted in two ideas. One, that land is a public resource and everyone is entitled to a slice of it. And second, that if you exclude the community from a piece of land, that's fine, but you have to compensate them for being excluded. If you take all the money paid as compensation and distribute it equally between everybody, then everyone gets cash equivalent to the value of the slice of land they are entitled to, and can use it to pay compensation on a piece of land without ever worrying they will run out of money to pay for land (they might have to move to a different plot if the plot they're on rises a lot in value, but they will always be able to find a plot with a value below or equal to the country's land value per capita).
Georgist will differ in their views on how much LVT revenue should be distributed as UNI and how much should be spent by the government on public goods, but as long as a significant amount is distributed as UNI, the problem with LVT you pointed to in this post vanishes.
This blog post is correct in the sense that naive Economics 101 doesn’t work in the real world. But real economics doesn’t end at 101, it goes all the way to 999 or whatever the final course is these days. The problem isn’t with Economics as a discipline, the problem is with laymen (including many EAs/rationalists) having a shallow understanding of economics.
1. Re: prediction markets. The primary objection isn’t about bribery or assassinations, rather it’s about them being defacto a replacement for online gambling for many people. See: Whales vs Minnows market on Manifold for an early example of what can go wrong.
2. Re: bribery via prediction markets. If you’re going to bribe someone… why not just send them Monero instead of doing complicated market shenanigans? I buy that crypto introduces a new corruption vector, but don’t think prediction markets move the needle meaningfully.
3. Re: Amish and fertility. I’ve never seen anyone seriously suggest we should learn from the Amish. People instead talk about Israel because it’s a developed capitalist country where even secular families have 2+ kids.
4. Re: LVT and ownership. Texas of all places already has a 2% real estate tax. Does this make property rights in Texas less meaningful compared to Germany which will tax 55% of your salary but only charge you 0.1% in property tax?
5. Re: signaling. Arguing against it is like trying to tell the peacock that there’s gotta be cheaper ways to show off his reproductive capabilities :-) I don’t think that’s a failure of economics as a science.
6. Re: dependency on China. That one is a political failure, economists have long warned that this would happen once you let China into the WTO (among other trade treaties).
7. Re: coastal elites concentration. This would be the case even sans China due to automation. How many people does Tesla employ today per car made vs Ford in 1960?
8. Re: AGI. The section sort of reads as “tails I win, heads you lose”. If “Econ brains” have gotten many things right, maybe they’re not so bad after all? :-)
Note that I agree with you that AGI doom is under appreciated in discussions of the future. But this has more to do with people not understanding how modern “AI” works vs a failure of economics in particular.
What you criticize as excessive "econ-brainedness" is an instance of what David Chapman calls "rationalism", as opposed to having good "meta-rational" judgment of being in the "sweet spot". Unfortunately despite his decades-long crusade these notions still have little purchase even in the sci-fi-nerd-sphere, while the lack of the common language to discuss, and even conceptualize this stuff is glaring, once you see it...
Isn't section 1 a strong argument that (admittedly burdensome) know-your-customer/beneficial-owner regulations need to be applied to prediction market companies at least as much as to banks?
You'd also need to ban prediction market runners that didn't provide a mechanism for anyone with a plausible concern to query the real identity of any market participant (i.e. basically banning markets with genuinely anonymous accounts).
This overhead and hassle in setting up a prediction market account would also reduce the severity of prediction markets' "make gambling more accessible" problem
I don't buy this criticism of LVT. If building densely on parcels is legally easy, and governments would have much reason to keep land use regulations rather liberal with land taxes as a revenue stream, community hubs could affordably rent space from property owners who manage several units on the parcel of land. They wouldn't have to bear the entire burden of LVT.
Furthermore, in practice for most of these artist groups, churches, community spaces... They don't own the land. They rent a unit to then operate from. The impact of an LVT on them would be rather attenuated, especially if the LVT replaces other kinds of taxes and promotes investment into things like transit that makes operation easier.
And the thing is, governments need tax revenue to provide services. The actual harms here to autonomy by an LVT compared to even our current tax regime, which has welfare cliffs and other pathologies, seems misguided to me to emphasize. It seems much more respectful of people's liberty than the alternatives, and aligns the government's incentives with maximizing the kinds of uses that people can get from the land-- which is in of itself a kind of liberty.
It seems to me that this kind of market pressure in practice is so much more benign than other pressures on people that more directly limit their autonomy and what they can do in a much more categorical fashion, such as governments, religion, community pressure and the like. Markets also provide much needed counters to these pressures on people's autonomy: the reason why the Amish are able to survive is in large part because of markets, that they sell things.
An LVT becoming a risk to churches or other non-profit organizations is a good insight. I think that in many cases it may be avoidable as you could just have your property on the outskirts of a city where the LVT should drop off very rapidly, but I may be thinking in terms of a medium-sized city rather than a million plus city.
I feel more positive about an LVT than about straightforward YIMYism because the incentives for restricting development remain in place without an LVT. It just seems like a treadmill to try and to improve the ability to permit construction when there's a powerful financial and social incentive to prefer lower density residential areas. I think the social reasons are valid - another reason I like it is that it removes and even reverses the financial incentive but if people want to pay to preserve a single family neighborhood character that is something that they should have the option to do as opposed to blanket upzoning.
Sufficient deferall options seem important, especially for seniors.
I had always assumed that it would be necessary to have exemptions to a Land value tax to make it politically viable (e. g. churches, parks in cities). Exemptions obviously bring their own issues but if there were limits on how much land in a city / borough could be exempted then it might not be abused too much.
I found this to be an interesting post. My initial reaction is that the problem with the "econ brained" examples is usually not with the use of quantitative modeling per se, but rather with not using models that consider a wide enough set of variables and the complex possible interactions between agents and second-order effects.
Interesting post and find the concept of sociopolitical autonomy compelling and important.
This is maybe an excessive amount of soldier mindset on my part, but I was trying to imagine how EA-style charity might actually increase sociopolitical autonomy.
Here's an example: health interventions like malaria bednet donation. They create dependency in the short term and therefore undermine sociopolitical autonomy in the short-term.
They may also decrease the number of deaths and increase the population. They almost certainly free up individuals from fighting disease (on an individual biological level, and on a collective level as a society) to making other economically productive (or even sociopolitically productive!) contributions to their countries. These two effects, over a timescale of 2-4 decades, could increase those individuals' and their countries' sociopolitical autonomy.
Also: any reduction in sociopolitical autonomy that results from bednet donations might not concern the donors (who are most interested in generating QALYs as an intrinsic good) and could even be a bonus!
As you describe in your post, US has an active rivalry with China. It is a long-standing practice of the US to create relationships with other nations that increase their dependency on the US in order to gain allies during their rivalries with other great powers. Historically the primary example is the cold war with the USSR. The same logic could apply to the current competition with China.
I do not think donating bednets is the most effective way to generate US allies in the great power rivalry, but if it has that side effect, it's not clear that's a bad thing!
I found the post very insightful and good food for thought. In the last years I've come to a better understanding and appreciation of markets and capitalism, which would mean I'd be in a right spot at the moment to be a bit biased in favor of econ views, but I find the opposite to be happening, and at the core of this is what I feel as an unease about taking economic toy models too generally and too seriously, and i feel this post of yours puts a finger on this kind of overexertion. Examples that just grate on me are EA strong-pilled consequentialist aggregate and impartial calculation of +EV, or Caplan himself. Most of his arguments feel to me like a combination of reductionism, pigheadedness and willful blindness but tbf, we all have are axioms, and both Caplan and EAs are just very strongly consistent with them and at bullet biting, which is... kind of a virtue? Still, I retain the intuition that this leads them to the short of blindness, flight from common sense and stupidity that only very smart people can fall for.
In response to footnote 4, a land value tax that was assessed against the market value of the land would not unfairly punish landowners banned from building, since the ban on building would proportionately diminish the market value of the land.
Land value taxes are meant to be assessed on the market value of the land itself, independent of whether there are buildings on it. So that value should stay the same whether or not there's a skyscraper on the land. But then the people with the skyscraper pay the same as their neighbor who only has a single-family house on equally-sized land.
In other words, you're right that the building regulations diminish the market value of unimproved land, but LVT is meant to estimate the value of *all* land as if it's unimproved land, so this doesn't help distinguish them.
I don't think this is right. The unimproved land value will depend on the property rights one obtains alongside it.
To illustrate, consider one of the Georgist schemes for estimating the market value of the land: the lot owner declares a "land value" to be taxed against; any member of the public can buy the land from him above the declaration price (if they promise to tear down any improvements).
Suppose a lot owner lives in a single-family house in an unzoned lot inside a busy urban area. Now suppose mayor NIMBY replaces mayor YIMBY and zones his lot as single-family: the lot owner can reduce his "land value" declaration, since previously, the potential buyers would have valued his land using the sum of future skyscraper rents; now the buyers can only earn single-family house rents.
Yes, that all seems right. But if we're using this mechanism, the owner of a skyscraper on an equivalent land plot would be able to declare the same land value, since the buyer would have to tear down the skyscraper and start from scratch, which they wouldn't want to do. Yet for that tax bill they can get all the benefits of having a skyscraper on that land. So my point remains that the tax burden is unfair to people who didn't manage to get skyscrapers built before burdensome regulation, compared with those that did.
Maybe I initially misunderstood your point; would it be fair to characterize it as:
"There is some inherent unfairness in taxing a skyscraper-owner the same amount as an adjacent single-family homeowner, perhaps in the same way it might be unfair to tax a CEO the same amount as a janitor. Relative to a property tax, an LVT imposes this form of unfairness, but has the benefit of not disincentivizing land improvements. However, if in some contexts land improvements are de jure / de facto banned, then the relative benefit of the LVT is negated; in those contexts a property tax and an LVT are equally efficient, so the property tax should be preferred on fairness grounds."
I think a Caplan-style libertarian econ-brained person would simply dispute the premise that taxing the skyscraper owner the same amount is unfair, and an Alexander-style utilitarian econ-brained person would dispute the idea that fairness matters intrinsically.
Brought to mind Adam Bulter's recent Substack, The Strip Mining of Trust. Shared insight: Economic optimization systematically consumes the non-economic substrate on which economic activity depends.
The convergence is striking given your different backgrounds (institutional finance vs decision theory/AI safety). Butler calls it "strip-mining trust." You call it "undermining sociopolitical autonomy." Both describe the same phenomenon—the systematic conversion of accumulated social capital into short-term legible gains.
Suggest you two might want to compare notes as you develop your theses.
A lot (most) of actual economists don’t think in terms of standard rationality and incentives (econ-brained).
I'd consider these people as applying non-economic frameworks to understand economics. I think virtue is not yet an economic concept in the same way as a concept like surge pricing is.
I should also clarify that economists themselves are often not very econ-brained when thinking about other disciplines (and even when thinking about parts of their own discipline—hence why so few economists are libertarians). Econ-brain involves an attempt to be intellectually consistent across domains which is rare (but disproportionately common amongst the most influential intellectuals).
Agree. May rational choice theory burn in hell, and salt be strewn in its grave so nothing ever grows there again.
To me this sounds like saying "may economics burn in hell". Basic economics (and rational choice theory) is good and useful, as long as you apply it appropriately (and in particular don't try to make it into an eternalist ideology, as per https://meaningness.com/eternalism).
LVT has always, since Henry George himself, been one half of a policy package, with the other part being what Henry George called the "citizen's dividend", most people today would call UBI (though that wouldn't be precise), and I call Universal Natural Income (UNI).
Georgist policy proposals are rooted in two ideas. One, that land is a public resource and everyone is entitled to a slice of it. And second, that if you exclude the community from a piece of land, that's fine, but you have to compensate them for being excluded. If you take all the money paid as compensation and distribute it equally between everybody, then everyone gets cash equivalent to the value of the slice of land they are entitled to, and can use it to pay compensation on a piece of land without ever worrying they will run out of money to pay for land (they might have to move to a different plot if the plot they're on rises a lot in value, but they will always be able to find a plot with a value below or equal to the country's land value per capita).
Georgist will differ in their views on how much LVT revenue should be distributed as UNI and how much should be spent by the government on public goods, but as long as a significant amount is distributed as UNI, the problem with LVT you pointed to in this post vanishes.
This blog post is correct in the sense that naive Economics 101 doesn’t work in the real world. But real economics doesn’t end at 101, it goes all the way to 999 or whatever the final course is these days. The problem isn’t with Economics as a discipline, the problem is with laymen (including many EAs/rationalists) having a shallow understanding of economics.
1. Re: prediction markets. The primary objection isn’t about bribery or assassinations, rather it’s about them being defacto a replacement for online gambling for many people. See: Whales vs Minnows market on Manifold for an early example of what can go wrong.
2. Re: bribery via prediction markets. If you’re going to bribe someone… why not just send them Monero instead of doing complicated market shenanigans? I buy that crypto introduces a new corruption vector, but don’t think prediction markets move the needle meaningfully.
3. Re: Amish and fertility. I’ve never seen anyone seriously suggest we should learn from the Amish. People instead talk about Israel because it’s a developed capitalist country where even secular families have 2+ kids.
4. Re: LVT and ownership. Texas of all places already has a 2% real estate tax. Does this make property rights in Texas less meaningful compared to Germany which will tax 55% of your salary but only charge you 0.1% in property tax?
5. Re: signaling. Arguing against it is like trying to tell the peacock that there’s gotta be cheaper ways to show off his reproductive capabilities :-) I don’t think that’s a failure of economics as a science.
6. Re: dependency on China. That one is a political failure, economists have long warned that this would happen once you let China into the WTO (among other trade treaties).
7. Re: coastal elites concentration. This would be the case even sans China due to automation. How many people does Tesla employ today per car made vs Ford in 1960?
8. Re: AGI. The section sort of reads as “tails I win, heads you lose”. If “Econ brains” have gotten many things right, maybe they’re not so bad after all? :-)
Note that I agree with you that AGI doom is under appreciated in discussions of the future. But this has more to do with people not understanding how modern “AI” works vs a failure of economics in particular.
What you criticize as excessive "econ-brainedness" is an instance of what David Chapman calls "rationalism", as opposed to having good "meta-rational" judgment of being in the "sweet spot". Unfortunately despite his decades-long crusade these notions still have little purchase even in the sci-fi-nerd-sphere, while the lack of the common language to discuss, and even conceptualize this stuff is glaring, once you see it...
Isn't section 1 a strong argument that (admittedly burdensome) know-your-customer/beneficial-owner regulations need to be applied to prediction market companies at least as much as to banks?
You'd also need to ban prediction market runners that didn't provide a mechanism for anyone with a plausible concern to query the real identity of any market participant (i.e. basically banning markets with genuinely anonymous accounts).
This overhead and hassle in setting up a prediction market account would also reduce the severity of prediction markets' "make gambling more accessible" problem
I don't buy this criticism of LVT. If building densely on parcels is legally easy, and governments would have much reason to keep land use regulations rather liberal with land taxes as a revenue stream, community hubs could affordably rent space from property owners who manage several units on the parcel of land. They wouldn't have to bear the entire burden of LVT.
Furthermore, in practice for most of these artist groups, churches, community spaces... They don't own the land. They rent a unit to then operate from. The impact of an LVT on them would be rather attenuated, especially if the LVT replaces other kinds of taxes and promotes investment into things like transit that makes operation easier.
And the thing is, governments need tax revenue to provide services. The actual harms here to autonomy by an LVT compared to even our current tax regime, which has welfare cliffs and other pathologies, seems misguided to me to emphasize. It seems much more respectful of people's liberty than the alternatives, and aligns the government's incentives with maximizing the kinds of uses that people can get from the land-- which is in of itself a kind of liberty.
It seems to me that this kind of market pressure in practice is so much more benign than other pressures on people that more directly limit their autonomy and what they can do in a much more categorical fashion, such as governments, religion, community pressure and the like. Markets also provide much needed counters to these pressures on people's autonomy: the reason why the Amish are able to survive is in large part because of markets, that they sell things.
An LVT becoming a risk to churches or other non-profit organizations is a good insight. I think that in many cases it may be avoidable as you could just have your property on the outskirts of a city where the LVT should drop off very rapidly, but I may be thinking in terms of a medium-sized city rather than a million plus city.
I feel more positive about an LVT than about straightforward YIMYism because the incentives for restricting development remain in place without an LVT. It just seems like a treadmill to try and to improve the ability to permit construction when there's a powerful financial and social incentive to prefer lower density residential areas. I think the social reasons are valid - another reason I like it is that it removes and even reverses the financial incentive but if people want to pay to preserve a single family neighborhood character that is something that they should have the option to do as opposed to blanket upzoning.
Sufficient deferall options seem important, especially for seniors.
I had always assumed that it would be necessary to have exemptions to a Land value tax to make it politically viable (e. g. churches, parks in cities). Exemptions obviously bring their own issues but if there were limits on how much land in a city / borough could be exempted then it might not be abused too much.
I found this to be an interesting post. My initial reaction is that the problem with the "econ brained" examples is usually not with the use of quantitative modeling per se, but rather with not using models that consider a wide enough set of variables and the complex possible interactions between agents and second-order effects.
Interesting post and find the concept of sociopolitical autonomy compelling and important.
This is maybe an excessive amount of soldier mindset on my part, but I was trying to imagine how EA-style charity might actually increase sociopolitical autonomy.
Here's an example: health interventions like malaria bednet donation. They create dependency in the short term and therefore undermine sociopolitical autonomy in the short-term.
They may also decrease the number of deaths and increase the population. They almost certainly free up individuals from fighting disease (on an individual biological level, and on a collective level as a society) to making other economically productive (or even sociopolitically productive!) contributions to their countries. These two effects, over a timescale of 2-4 decades, could increase those individuals' and their countries' sociopolitical autonomy.
Also: any reduction in sociopolitical autonomy that results from bednet donations might not concern the donors (who are most interested in generating QALYs as an intrinsic good) and could even be a bonus!
As you describe in your post, US has an active rivalry with China. It is a long-standing practice of the US to create relationships with other nations that increase their dependency on the US in order to gain allies during their rivalries with other great powers. Historically the primary example is the cold war with the USSR. The same logic could apply to the current competition with China.
I do not think donating bednets is the most effective way to generate US allies in the great power rivalry, but if it has that side effect, it's not clear that's a bad thing!
I found the post very insightful and good food for thought. In the last years I've come to a better understanding and appreciation of markets and capitalism, which would mean I'd be in a right spot at the moment to be a bit biased in favor of econ views, but I find the opposite to be happening, and at the core of this is what I feel as an unease about taking economic toy models too generally and too seriously, and i feel this post of yours puts a finger on this kind of overexertion. Examples that just grate on me are EA strong-pilled consequentialist aggregate and impartial calculation of +EV, or Caplan himself. Most of his arguments feel to me like a combination of reductionism, pigheadedness and willful blindness but tbf, we all have are axioms, and both Caplan and EAs are just very strongly consistent with them and at bullet biting, which is... kind of a virtue? Still, I retain the intuition that this leads them to the short of blindness, flight from common sense and stupidity that only very smart people can fall for.