[Eran]

The point I was making is that you don't need a market to solve multiobjective optimization problems in computer science. So why do you insist that the market is the only way to solve such problems in economics? There are obviously alternatives. The point is that markets are not the only solution to optimization problems.

Markets do much more than solve multi-objective optimization problem.

To start with, markets effectively "collect" the information required to even pose the problem in the first place. I use "collect" in a qualified manner because, of course, markets do not concentrate the information - that would be as impossible for markets as it would be for any alternative system. But the information gathering problem is already insurmountable for any computer system.

Even that, however, isn't the main point. Markets don't just solve a static economic production problem. Markets are a dynamic system in which not just input parameters change constantly, but even which problem needs solving keeps evolving. Markets allow innovation, and innovation, by its very nature, cannot be programmed into computers.

Computers can solve a well-posed mathematical problem. But posing the problem that markets solve every day is already an impossible task.

It's a rare sort of problem that can't be reduced to smaller parts; economics certainly isn't one of those problems.

It can be reduced to smaller parts - in fact, markets reduce the problem to parts, each of which is "solved" by a single person. However, it is impossible to separate the collection of information required to solve the problem (e.g. how much do individual consumers and, in consequence, consumers in aggregate, relatively value different products) from the solution itself.

You can only find out how much people value (i.e. are willing to pay for) a product by offering them a concrete product for a concrete price, i.e. by creating the market. If you tried to use surveys, for instance, you will get substantially different, in far inferior results.

Yes they are. They're a solution to the multiobjective problem in economics. It's how people figure out how much of X and Y ought to be bought to produce the most profitable amount of Z. You are correct that they do not solve the multimodal optimization problem in economics--but that's not the function of a market.

How or who determines which Z to produce? That is a question entrepreneurs answer every day, and no computer ever did, and probably never could.

Markets are one method by which that theoretical optimum is decided upon. One of many potential ways to solve that problem--as noted before. What they don't do is decide what to optimize for; that's a political question that the people who work in economics go out of their way not to actually study (instead providing a docile service to power).

These are two separate questions. I am assuming we both are talking about solving the economic problem associated with most efficiently meeting the demands (or preferences) of a larger number of people.

Of course if your political system has different ends, e.g. meeting the demands of a much smaller number of political decision-makers, broad, consumer markets aren't the solution.


edit (prematurely submitted, having failed to respond to subsequent comments):

Prove that human nature even exists. No one has ever managed that feat, so I'm continually surprised by people's insistence that they know what is and is not compatible with "it."

This isn't mathematics, so I am not going to attempt to "prove" anything (nor can you, of course). However, I can point to evidence that supports the assertion.

There are certain universals that we observe in every single human society (without exception), both present and past. When those universals relate to characteristics of human behaviour, it makes sense to think of them as part of an immutable human nature.

One such characteristic is the rapidly-declining regard (expressed in materialistic terms) that people have to progressively more remote others. People universally have strong regard to themselves and their immediate family, lower regard to their immediate neighbours (in stationary societies), tribe-members or others belonging to a relatively-small group of well-known and coherently-related individuals, even less regard to strangers who are still considered parts of the same group (e.g. workers in different facilities, members of the same country or large religious group, etc.), and far less regard to complete strangers. Often, "negative regard" (i.e. willingness to invest resources in harming others) can be found applied to "enemies", whether animosity is determined at the tribal, national, religious, ethnic, linguistic or other level.

This universal (and thus presumably universal) aspect of human behaviour is inconsistent with a large-scale gift economy.

Except ultimately accountable to the people they're managing, as opposed to the people who are collecting the profits. If everyone in a group decides they need a manager, then more power to them. But that's a very different situation from some third party imposing a manager on the workforce.

I have much to say about this, but I notice that we are diverging from the topic, and I believe this is an issue we may well be debating in other threads. The relevant issue here is the viability of a gift economy.

Gifts have several problems when compared with payments. First, they aren't simultaneous. A gift, in essence, is a form of credit being extended. That works well when the people exchanging gifts know each other well, or when the gifts are relatively minor in value. It works far less well when what's at issue is shipping a month's supply from your factory to a receiver in China.

The second problem is that gifts do not allow measuring relative value with any precision. That introduces a huge degree of "fuzziness" into the exchange. How can you plan whether to produce product A or product B if you have no idea, in advance, what you will get as gifts in return for gifting A to consumer 1, or B to consumer 2?

Most problematic, perhaps, is that gifts do not allow the recipient to have an effective voice as to the relative priority of the object in question. Let's say you produce different kinds of widgets which have different qualities and, therefore, different costs of production. Other things being equal, I would, of course, prefer the higher quality (and more expensive) widgets.

In the context of a gift economy, how can you tell how important that difference is to me? How can you tell whether it is preferable to produce 1000 high-quality, or 1200 medium quality widgets?

Or we can just drop the false dichotomy and recognize that actual relationships can work in any number of ways not immediately predictable from the outside without context.

I am yet to hear a single coherent description of a national- or international-scale system which doesn't use money to mediate arms-length exchanges between autonomous production units. So far, all you have done is insisted that other solutions exist, without suggesting one. Please feel free to do so.

To be clear, the economic calculation problem can easily be solved in the context of an economy dominated by worker-owned-and-controlled syndicates, provided only that those syndicates trade with each other at an arm's length.

I can easily prove that Marxism can work on a feeble level by example, say North Korea.

No, that doesn't work. Because North Korea doesn't exist in a vacuum. Even if it didn't trade with the outside world (which it does), the North Korean planners (like Soviet planners before them) have access to relative prices in the rest of the world.

It is, of course, possible to (inefficiently) run a small economy without prices, by mimicking, at some level, price-based production decisions made in the wider, market-based global economy.

The Economic Calculation Problem in its ultimate form, however, excludes the possibility of a global priceless economy.

Sometimes, just acting while knowing that the result is wrong and imperfect is good enough.

I guess that really depends on your standards. Going back to being hunters and gatherers might be "good enough" for some people. Certainly, the vast inefficiencies of government-run roads, education and other systems are considered "good enough" by many.

One way of understanding the ECP argument is that it highlights the role of prices. In the complete absence of prices (i.e. in the end-state of global communism), no economic calculation is possible. In more moderate cases, e.g. complete government control of whole industries, or the internal working of a large corporation, some calculation is possible, but the results are far less efficient than under a free market alternative.

It is also possible that the transaction costs of using money are greater than the calculation efficiency achieved with money.

That is certainly the case in many instances. The internal workings of a corporation is a good example. The best solution, of course, is to let the market decide even that. If using money is less efficient, free people acting to promote their interests will choose not to use it.

[Someone5]

Markets do much more than solve multi-objective optimization problem.

No, they don't.

To start with, markets effectively "collect" the information required to even pose the problem in the first place. I use "collect" in a qualified manner because, of course, markets do not concentrate the information - that would be as impossible for markets as it would be for any alternative system.

Markets don't collect one bit of information. Markets are a process that economic actors utilize to make decisions for themselves; those actors are the ones to collect or analyze the information. You're attributing voodoo to markets again. X costs so much, Y costs another something, therefore the optimal amount of Z (made from X and Y) ought to be so much--limited by projections of demand for Z, of course. That's all a market fundamentally does.

But the information gathering problem is already insurmountable for any computer system.

Computers are making these decisions right now. They're what puts the unfathomable and immense amount of economic data into human-sensible reports. Humans need the dumbed down version because we're absolutely horrible at dealing with big sets of data, while computers are amazing at it. Humans can barely work their head around abstract indexed values like the current prices for every commodity on a second-by-second basis--let alone comprehensive econometrics in real time.

Humans would literally be incapable of participating in modern markets without computer assistance above the small local business level. There is literally no way that you could analyze that much data in a time frame that allows meaningful participation without a computer program summarizing it for you. Let me repeat that; computers already do the data processing. Humans just make the decisions that follow from the reports those programs generate.

Modern economic activity is already hopelessly dependent on expert systems and big data processing. The idea that computers have an information gathering problem while humans don't is utterly laughable. Computers have an information comprehension problem, not an information gathering problem. They're way, way better at gathering information than humans ever could be. And given current trends in the industry, I'd say that it won't be too much longer before computers won't even have a comprehension problem for simple problems like microeconomic decisions--and expert systems will turn into oracles that the priests of capital can only pray to wisdom from.

Even that, however, isn't the main point. Markets don't just solve a static economic production problem.

You're the only one talking about static problems.

Markets are a dynamic system in which not just input parameters change constantly, but even which problem needs solving keeps evolving.

Which is why these programs need continual updates from the folks who don't have a comprehension problem. Though, of course, it does not change nearly so much as you imply.

Markets allow innovation, and innovation, by its very nature, cannot be programmed into computers.

Which is the human component of all of this; for a little while yet, anyway.

Computers can solve a well-posed mathematical problem.

And ill-defined ones too, through the use of evolutionary algorithms and neural networks. Genetic algorithms are particularly useful in economics because they're extremely useful in solving optimization problems--and the entire field of economics is the study of an optimization problem (that we call an economic system). That's one of the cool things about programming; you can define an algorithm to create another algorithm that solves a different problem. As long as you can divide the task into programmable steps and have the computational resources to spare, it can be done.

But posing the problem that markets solve every day is already an impossible task.

Hardly.

It can be reduced to smaller parts - in fact, markets reduce the problem to parts, each of which is "solved" by a single person.

So retract your earlier claim, you've already acknowledged that you were wrong with this admission.

However, it is impossible to separate the collection of information required to solve the problem (e.g. how much do individual consumers and, in consequence, consumers in aggregate, relatively value different products) from the solution itself.

Again, you are mixing terms here; as I said, markets can solve the multiobjective problem, but can't solve the multimodal problem. Figuring out what should actually be valued is the multimodal problem. Computers--just like humans--can solve that problem for themselves (I.E. arrive at an algorithmically defined value for a good), but can't solve it for others. I mean, it's entirely possible for a program to determine a value for something simply by adding together every input that went into creating it. That's a valid means of evaluating something; but this doesn't mean that it's the correct method for the humans who might be buying the good. Real value and monetary cost aren't at all the same--no matter how much economists insist. That's just their own little philosophical bent--it's a political objective that they as people want to achieve, not an actual fundamental truth of reality.

Economics has no solution to this problem. Markets don't even touch on it. The capitalist solution is to let elites determine what's important or not. Markets don't actually arrive at the real value of anything; they just let elites negotiate the price at which they'll be convinced to make more of something.

You can only find out how much people value (i.e. are willing to pay for) a product by offering them a concrete product for a concrete price,

And computers are perfectly capable of doing that, and evaluating the consumer responses to prices (I.E. by evaluating changes in consumer purchases in response to price changes), and then adjusting prices to seek greater profit. Nothing there requires some innate human ingenuity. Especially for commodities.

i.e. by creating the market. If you tried to use surveys, for instance, you will get substantially different, in far inferior results.

I'll agree that the response would be different, but I'm not so sure about it being inferior.

How or who determines which Z to produce?

You don't understand what "multimodal" means in this context, do you? You've just restated the problem I said markets didn't solve. Like I said, the capitalist answer is to let elites (the folks who control the capital) make those decisions. With the hope that they'll be guided only by concerns of further profit, not personal power and prestige, or emotional drives.

That is a question entrepreneurs answer every day, and no computer ever did, and probably never could.

Yes, but it is also not something that a market has ever solved either. That is the aforementioned multimodal problem. Entrepreneurs are not markets. You didn't even address my comment.

These are two separate questions. I am assuming we both are talking about solving the economic problem associated with most efficiently meeting the demands (or preferences) of a larger number of people.

I'm talking about solutions to the multiobjective problem--I.E. alternatives to markets. Your tangent about the role of entrepreneurs is a distraction, since markets and entrepreneurs do not solve the same sort of problem and are not actually two words for the same thing. Let me simply point out that even if we did replace markets as a solution to the multiobjective problem, that does not mean that human luminaries would not create demand for new things.

You insist that there is a calculation problem that can only be solved by markets. Demonstrate that this is true. Don't distract from the issue by going off on a tangent about the role of elites in shaping demands.

Of course if your political system has different ends, e.g. meeting the demands of a much smaller number of political decision-makers, broad, consumer markets aren't the solution.

You should know by now that my preferred political system puts decisions in the largest possible number of hands; by distributing political decisions to everyone, not any subset of the population.

Your preferred political system does in fact put every important decision entirely in the hands of the people with the most capital.

This isn't mathematics, so I am not going to attempt to "prove" anything (nor can you, of course). However, I can point to evidence that supports the assertion.

Proof is required in philosophy too. It's a requirement for logic generally. Moreover, as my position is the skeptics' position, it is not my role to prove that something isn't true. Insisting on negative proofs is a classic fallacy. I cannot prove that human nature does not exist, but absent any convincing proof that it does exist, one would be foolish to base their political beliefs on the assumption that it does.

There are certain universals that we observe in every single human society (without exception), both present and past.

Alright, shoot. What are some of those "universals" that every human society in the world has followed. Keep in mind that if I can point out even a handful of counterexamples where those universals were not followed, it would put a lie to your claim about "human nature." Note my responses to your prior "evidence."

When those universals relate to characteristics of human behaviour, it makes sense to think of them as part of an immutable human nature.

Go ahead and name some of these so-called universals. I have my doubts.

One such characteristic is the rapidly-declining regard (expressed in materialistic terms) that people have to progressively more remote others.

Remote in what way? I feel pretty close to my family, even when we live far removed from each other. Do you mean emotional remoteness? Which is really just a tautology--people don't have emotional regard for people they don't have emotional regard for. Not really any useful observation that one might characterize as "human nature."

People universally have strong regard to themselves and their immediate family,

Really, you've never heard of anyone who genuinely hates one or both of their parents? No parent that is so disappointed with their child that they'll cut them off? Because I've got to say that, while that is not the normal condition, it is still distressingly common.

As for self-regard, that's absolutely not true. Tons of people have horrible self-confidence and self-respect. Lots of people think they're worthless ("I could never do something like that!"). Lots of people have such little regard for themselves and their immediate family that they commit suicide. Hardly a human universal constant.

lower regard to their immediate neighbours (in stationary societies), tribe-members or others belonging to a relatively-small group of well-known and coherently-related individuals,

Well now, that's an interesting load of qualifications for a "universal" truth. Because apparently our regard for our neighbors is dependent on whether we live in a stationary society or not--not, in fact, a human universal. And let me also point out that tribal groups are also an ill-defined social construct, where tribe means many different things to many different groups. It's not like there's some universal human definition of what a tribe is--it can be quite a lot larger for some than for others. Some have more exceptions than others. Again, not an actual universal because it varies greatly on culture--some cultures consider tribal associations to be important, others completely disregard them as irrelevant. Moreover, these very concepts are mutable concepts that change over time... which certainly shows them not to be some human constant that all humans share.

even less regard to strangers who are still considered parts of the same group (e.g. workers in different facilities, members of the same country or large religious group, etc.), and far less regard to complete strangers.

Again getting into tautologies; "humans don't care about the people they don't care about." Not really anything like the powerful driving human nature you are attributing political potentials and impossibilities upon.

Often, "negative regard" (i.e. willingness to invest resources in harming others) can be found applied to "enemies", whether animosity is determined at the tribal, national, religious, ethnic, linguistic or other level.

"People don't like the people they don't like." Is all of this supposed human nature comprised of nothing but emotionally tinged tautologies? Because I am severely unimpressed by this, and not nearly convinced that it has any sort of meaningful application when debating what is and is not politically possible.

This universal (and thus presumably universal) aspect of human behaviour is inconsistent with a large-scale gift economy.

Gift economies are logical from a self-interest standpoint alone. Your fundamental mistake here is in assuming that the only motive for a gift economy is altruism and fellowship with other humans. Gift economies work even if you confine them to small groups in interconnected communities--because one does not have to share everything with everyone, only with one's friends. Your circle of friends and family undoubtedly have different circles of friends and family that you do not personally know. Just because your best friend's cousin needs something and you don't care about him does not mean that you do not care for your best friend and that he does not care for his cousin.

As long as people like the people they like, and that they people they like know other people whom you do not, there is a means by which your labor is distributed to people you do not like (using your friends and family as intermediaries).

On another note, simple expectation of reciprocity is enough to justify a gift economy.

[Eran]

Markets don't collect one bit of information. Markets are a process that economic actors utilize to make decisions for themselves; those actors are the ones to collect or analyze the information. You're attributing voodoo to markets again. X costs so much, Y costs another something, therefore the optimal amount of Z (made from X and Y) ought to be so much--limited by projections of demand for Z, of course. That's all a market fundamentally does.

Not quite. Markets tell a producer how much of his products consumers are buying at a given price. Markets tell a mining company what price it can get for its ore. The price for the ore is the result of integrating the demand for the ore's millions of different products, taking into account differences in quality and efficiency of using alternative raw materials at each of the thousands of different intermediate product in the production process which ultimately encompasses thousands of different manufacturers all over the world, creating millions of products to billions of end-customers.

And each of those end-customers can change their mind any day, each of those intermediate producers can come up with new and innovative ideas on new products or new production techniques. And the ore producer is competing with dozens of others all over the world, with demand pressures motivating them to explore for new deposits and new extraction technologies.

That unimaginably complex network (and I have barely touched the surface of its complexity) involving literally billions of human beings, production processes short and long, countless minds coming up with, testing, failing or succeeding with new ideas, all of those are effectively "coordinated" by the market.

Computers are making these decisions right now.

I am not talking about decision-making, but about information-gathering. How can a computer tell how many people would buy a $100 cell phone with a given set of features? Computers can help entrepreneurs create models, but those models, evidently, are less than reliable. That's why so many products actually fail. Because there is no substitute for putting actual product in front of actual consumer at an actual price and letting them decide if it is good enough or not.

What characterizes markets is that the information is being used automatically, without the need (or ability) to collect it to a central computer. You don't need to send questionnaires (of doubtful value) to millions of consumers regarding how much they value this or that feature. Instead, you just observe as they make their own complex, often sub-conscious purchasing decisions.