5 ms·

A nice read but I don't agree completely with the final paragraph. The invisible hand uses a greedy algorithm. It does take a mastermind to get close to any k

A nice read but I don't agree completely with the final paragraph. The invisible hand uses a greedy algorithm. It does take a mastermind to get close to any kind of optimal solution in many cases. We can't trust the invisible hand to provide healthcare to everyone, nor can we trust it to take into account external costs like pollution etc.

baddox 13y ago
What is the economic mechanism by which we can expect a government, particularly one guided by direct voting or representation, to do any better? After all, the production of a government that can solve market failures isn't an easier task for a society than solving the market failures themselves.
logicchains 13y ago
The invisible hand isn't in the business of making moral judgements external to those of the society in which it exists. If people wish to support each other's healthcare, then they're quite capable of doing so without any external assistance. Regarding pollution, it's interesting to note that in America at least it was originally considered a tort (if someone harmed you by polluting, you could sue them). But various judges have, in the name of the 'common good' of industrialisation, eroded this right.
- lifeisstillgood 13y ago
  Now that is interesting. THank you - I had forgotten my torts. What a nice way of putting externalities back. Which court however enfgorces torts where a polluting country pours noxious gas into another country's borders
  - logicchains 13y ago
    I believe such things are covered under international law, although I'm not sure how successfully it's enforced.
- nhaehnle 13y ago
  I think it's important to keep in mind that people acting to support each other's healthcare without external assistance is one of those things that government is. After all, European societies with solid health systems did not require external intervention by aliens or anything to do so. They just agreed that having such a system is a good thing, and the government happens to be a useful framework for implementing it (and keep in mind that a lot of healthcare-related stuff is actually private in Europe, depending on which country you're talking about). Overall, "government vs. markets" is a false dilemma. Markets need rules, and to a large extent, the historical evolution of governments goes hand in hand with the evolution of markets. Your pollution example fits this observation nicely. Another important point in the context of the original submission: A lot of discussion in society in this realm is confused about the centralization-decentralization and class/power structure axes. Those are largely independent, but people from all political backgrounds tend to muddle them up. [0] The pencil story is a nice illustration for why completely centralized control of the economy would be a disaster. However, it is often brought into (mostly laypeople's) political economy discussions where it really has no place, because most such discussions are not about centralization vs. decentralization. Most political economy discussions are (albeit usually in a hidden way) about class structure and power. Should there be a (necessarily tiny) class of capital owners who own and control the means of production and who, using this control, will be able to wrest for themselves a disproportionally large share of the economy's output? Or should power be distributed more evenly, along democratic lines? Ownership, control, and (de-)centralization are largely orthogonal. To give a concrete example, one could easily imagine a functioning, decentralized economy in which companies compete in the market place much like they do in the Western world today; but with the important difference that companies are (as a rule) not anybody's property, and that decisions within the company are made through a (probably representative) democratic process in which every employees has one vote. Such an economy would have vastly different distributive outcomes, but it would still be a decentralized economy that takes all the lessons of "I, Pencil" to heart. [0] This is particularly annoying when people on the left do it. At least people on the right have a motive for muddling these things up.
  - logicchains 13y ago
    There is however a link between centralisation and distributive outcomes in the sense that the former is necessary to ensure certain forms of the latter. If for instance the democratic companies in your thought experiment were somehow structurally less competitive in the market than companies led by a single CEO (I'm not suggesting this is the case, just using it to illustrate), then in a decentralised system the CEO-led companies would outcompete the democratic ones and naturally come to dominate the market. It would hence not be possible to bring about the distributive outcomes of a market dominated by democratic companies without some centralised force capable of granting the democratic companies some advantage over the CEO-led ones. In short, if a decentralised economy necessarily leads to greater concentration of wealth/power (again, I'm not suggesting it does), then people arguing for a more even distribution of wealth/power must necessarily also argue for a more centralised economy.
cynicalkane 13y ago
A "greedy algorithm" means an algorithm follows local optima. It does not mean the actors are greedy. Human beings are capable of jumping over local maxima, their alleged greediness notwithstanding.
- Retric 13y ago
  Free markets don't care about which company or approach scales better they just react to price. So, it's a greedy algorithm. That said we free markets are extremely uncommon and decisions are often made based on information other than price.
  - nickff 13y ago
    Even in free markets, the individuals may act in non financially-optimized ways. If people only used money to advance their financial interests, no one would found a charity, or volunteer, or even go to a theme park. Even for-profit companies may be founded for reasons other than accruing wealth for the shareholders.[1] >"[D]ecisions are often made based on information other than price" Price is one of the most important sources of information, as Mises made clear.[2] [1] https://www.youtube.com/watch?v=uegOUmgKB4E https://www.youtube.com/watch?v=uegOUmgKB4E [2] http://mises.org/books/socialism.pdf http://mises.org/books/socialism.pdf
pdabbadabba 13y ago
I don't know whether we can find an economic system that gives better results. I'm willing to assume for now that we cannot. But what I am sure we can improve is the incentive system upon which our current economic model operates -- for me it is struggle enough to ensure that the "greedy" system we have seeks to maximize actual human flourishing, and not something else. I do not dispute that the system of incentives in place now has worked and continues to work miracles. But the poorly-handled corner cases have become pressing. The problem is that there are important goods that we have not figured out how to treat economically. Human happiness has value, even when the human in question has no means to pay for it. The long-term health and beauty of our planet has value, but we lack the concepts to list it as a gain or loss on a balance sheet. The same is true for human health, peace, friendship, solitude, ... the list goes on. Our current approach is to consider how much a person would pay for any of these things -- and this is a fine start. But this approach ensures that we overlook ways to create value for the poor. And it overlooks the fact that economic culture may create blind spots: goods that we value but that we will not pay for economically in proportion to their value.
washedup 13y ago
Life itself is a greedy algorithm.
bcoates 13y ago
This essay was written in 1958, before the 'Invisible Hand' was replaced by the Coase theorem which actually explains why externalities don't happen in free markets as often as you would think they would (instead of just observing it). "It does take a mastermind to get close to any kind of optimal solution in many cases" isn't correct. Relaxing the optimality constraint doesn't actually make central planning any easier, difficulty is dominated by the worse-than-cubic scaling on the number of moving parts of the system (and look at how complex even pencil manufacturing is). Nations actually tried this, a whole lot of useful math for optimization problems was discovered, but economic optimization is still algorithmically intractable.
- czr80 13y ago
  >This essay was written in 1958, before the 'Invisible Hand' was replaced by the Coase theorem which actually explains why externalities don't happen in free markets as often as you would think they would I'm struggling to understand what you mean by this - Coase's theorem shows why externalities are often not addressed by market solutions, even when an apparent market solution exists (spoiler: transaction costs).
- dragonwriter 13y ago
  > This essay was written in 1958, before the 'Invisible Hand' was replaced by the Coase theorem which actually explains why externalities don't happen in free markets as often as you would think they would (instead of just observing it). Coase's theorem explains why externalities occur even if you make the idealized assumptions of rational choice theory (perfect knowledge and utility-optimizing behavior); it does directly state that (in those idealized assumptions) you wouldn't have externalities without transaction costs, because those who would otherwise suffer externalities would be motivated to buy off those who would inflict them. However, transaction costs are a very real thing, and are particularly significant in the case of externalities with diffuse impacts (whether that's harms or benefits that are diffuse.) Coase himself criticized misapplication of the theorem to claim that externalities were unlikely, since such applications necessarily discounted the real occurrence of transaction costs. Such misapplication essentially inverts the whole point of the theorem.

baddox 13y ago

What is the economic mechanism by which we can expect a government, particularly one guided by direct voting or representation, to do any better? After all, the production of a government that can solve market failures isn't an easier task for a society than solving the market failures themselves.

logicchains 13y ago

The invisible hand isn't in the business of making moral judgements external to those of the society in which it exists. If people wish to support each other's healthcare, then they're quite capable of doing so without any external assistance. Regarding pollution, it's interesting to note that in America at least it was originally considered a tort (if someone harmed you by polluting, you could sue them). But various judges have, in the name of the 'common good' of industrialisation, eroded this right.

lifeisstillgood 13y ago
Now that is interesting. THank you - I had forgotten my torts. What a nice way of putting externalities back. Which court however enfgorces torts where a polluting country pours noxious gas into another country's borders
- logicchains 13y ago
  I believe such things are covered under international law, although I'm not sure how successfully it's enforced.
nhaehnle 13y ago
I think it's important to keep in mind that people acting to support each other's healthcare without external assistance is one of those things that government is. After all, European societies with solid health systems did not require external intervention by aliens or anything to do so. They just agreed that having such a system is a good thing, and the government happens to be a useful framework for implementing it (and keep in mind that a lot of healthcare-related stuff is actually private in Europe, depending on which country you're talking about). Overall, "government vs. markets" is a false dilemma. Markets need rules, and to a large extent, the historical evolution of governments goes hand in hand with the evolution of markets. Your pollution example fits this observation nicely. Another important point in the context of the original submission: A lot of discussion in society in this realm is confused about the centralization-decentralization and class/power structure axes. Those are largely independent, but people from all political backgrounds tend to muddle them up. [0] The pencil story is a nice illustration for why completely centralized control of the economy would be a disaster. However, it is often brought into (mostly laypeople's) political economy discussions where it really has no place, because most such discussions are not about centralization vs. decentralization. Most political economy discussions are (albeit usually in a hidden way) about class structure and power. Should there be a (necessarily tiny) class of capital owners who own and control the means of production and who, using this control, will be able to wrest for themselves a disproportionally large share of the economy's output? Or should power be distributed more evenly, along democratic lines? Ownership, control, and (de-)centralization are largely orthogonal. To give a concrete example, one could easily imagine a functioning, decentralized economy in which companies compete in the market place much like they do in the Western world today; but with the important difference that companies are (as a rule) not anybody's property, and that decisions within the company are made through a (probably representative) democratic process in which every employees has one vote. Such an economy would have vastly different distributive outcomes, but it would still be a decentralized economy that takes all the lessons of "I, Pencil" to heart. [0] This is particularly annoying when people on the left do it. At least people on the right have a motive for muddling these things up.
- logicchains 13y ago
  There is however a link between centralisation and distributive outcomes in the sense that the former is necessary to ensure certain forms of the latter. If for instance the democratic companies in your thought experiment were somehow structurally less competitive in the market than companies led by a single CEO (I'm not suggesting this is the case, just using it to illustrate), then in a decentralised system the CEO-led companies would outcompete the democratic ones and naturally come to dominate the market. It would hence not be possible to bring about the distributive outcomes of a market dominated by democratic companies without some centralised force capable of granting the democratic companies some advantage over the CEO-led ones. In short, if a decentralised economy necessarily leads to greater concentration of wealth/power (again, I'm not suggesting it does), then people arguing for a more even distribution of wealth/power must necessarily also argue for a more centralised economy.

cynicalkane 13y ago

A "greedy algorithm" means an algorithm follows local optima. It does not mean the actors are greedy. Human beings are capable of jumping over local maxima, their alleged greediness notwithstanding.

Retric 13y ago
Free markets don't care about which company or approach scales better they just react to price. So, it's a greedy algorithm. That said we free markets are extremely uncommon and decisions are often made based on information other than price.
- nickff 13y ago
  Even in free markets, the individuals may act in non financially-optimized ways. If people only used money to advance their financial interests, no one would found a charity, or volunteer, or even go to a theme park. Even for-profit companies may be founded for reasons other than accruing wealth for the shareholders.[1] >"[D]ecisions are often made based on information other than price" Price is one of the most important sources of information, as Mises made clear.[2] [1] https://www.youtube.com/watch?v=uegOUmgKB4E https://www.youtube.com/watch?v=uegOUmgKB4E [2] http://mises.org/books/socialism.pdf http://mises.org/books/socialism.pdf

pdabbadabba 13y ago

I don't know whether we can find an economic system that gives better results. I'm willing to assume for now that we cannot. But what I am sure we can improve is the incentive system upon which our current economic model operates -- for me it is struggle enough to ensure that the "greedy" system we have seeks to maximize actual human flourishing, and not something else. I do not dispute that the system of incentives in place now has worked and continues to work miracles. But the poorly-handled corner cases have become pressing. The problem is that there are important goods that we have not figured out how to treat economically. Human happiness has value, even when the human in question has no means to pay for it. The long-term health and beauty of our planet has value, but we lack the concepts to list it as a gain or loss on a balance sheet. The same is true for human health, peace, friendship, solitude, ... the list goes on. Our current approach is to consider how much a person would pay for any of these things -- and this is a fine start. But this approach ensures that we overlook ways to create value for the poor. And it overlooks the fact that economic culture may create blind spots: goods that we value but that we will not pay for economically in proportion to their value.

washedup 13y ago

Life itself is a greedy algorithm.

bcoates 13y ago

This essay was written in 1958, before the 'Invisible Hand' was replaced by the Coase theorem which actually explains why externalities don't happen in free markets as often as you would think they would (instead of just observing it). "It does take a mastermind to get close to any kind of optimal solution in many cases" isn't correct. Relaxing the optimality constraint doesn't actually make central planning any easier, difficulty is dominated by the worse-than-cubic scaling on the number of moving parts of the system (and look at how complex even pencil manufacturing is). Nations actually tried this, a whole lot of useful math for optimization problems was discovered, but economic optimization is still algorithmically intractable.

czr80 13y ago
>This essay was written in 1958, before the 'Invisible Hand' was replaced by the Coase theorem which actually explains why externalities don't happen in free markets as often as you would think they would I'm struggling to understand what you mean by this - Coase's theorem shows why externalities are often not addressed by market solutions, even when an apparent market solution exists (spoiler: transaction costs).
dragonwriter 13y ago
> This essay was written in 1958, before the 'Invisible Hand' was replaced by the Coase theorem which actually explains why externalities don't happen in free markets as often as you would think they would (instead of just observing it). Coase's theorem explains why externalities occur even if you make the idealized assumptions of rational choice theory (perfect knowledge and utility-optimizing behavior); it does directly state that (in those idealized assumptions) you wouldn't have externalities without transaction costs, because those who would otherwise suffer externalities would be motivated to buy off those who would inflict them. However, transaction costs are a very real thing, and are particularly significant in the case of externalities with diffuse impacts (whether that's harms or benefits that are diffuse.) Coase himself criticized misapplication of the theorem to claim that externalities were unlikely, since such applications necessarily discounted the real occurrence of transaction costs. Such misapplication essentially inverts the whole point of the theorem.