Economic calculation problem - how serious is it?

[Otto Kretschmer]

(Non economist here)

In short - it's a theory developed by Mises and Hayek saying that a centrally planned economy can never match the efficiency of a market one because due to lack of price signals, the central planners will never be able to correctly determine the demand for each product, and they'll not be able to react with the same speed and elasticity as the market.

Do you guys have an opinion on this? My knowledge of economics is limited and I can neither refute nor confirm it myself.

[Raphael]

On the one hand, I'm generally sceptical about anything asserted by people like Hayek and von Mises.

On the other hand, the history of the 20th century seems to prove them right on this.

On the third hand, a modern economy is an incredibly complicated and complex thing, so getting, let alone processing, all the necessary information to manage all of it sounds like a gigantic task.

On the fourth hand, in order for anything like that to even stand a chance of working, it would have to be run by people who both expect and deliver complete honesty - higher ranking people who never give subordinates trouble for telling them things they don't want to hear, and subordinates who never try to curry favor with higher ranking people by telling them what they want to hear. That's difficult enough to achieve under the most open, transparent, and accountable social structures, and impossible under any others.

I'm not an economist myself, but then again, I'm kind of sceptical about much of the economics profession.

[Otto Kretschmer]

@Raphael

Note that in 20th century command economies the planning was done by hand while nowdays we have powerful computers with stuff like AI and Big Data

[Raphael]

@Raphael

Note that in 20th century command economies the planning was done by hand while nowdays we have powerful computers with stuff like AI and Big Data

Yes, but the economy keeps getting more complex, too. And this doesn't address my "fourth hand" point - the computers would have to get information that's actually accurate, not inputs written by opportunists who want to make a good impression on their bosses. And then, the instructions of the computers would have to be actually carried out, without opportunistic careerists simply telling the computers that they carried out their instructions, while not actually having done so.

Serious question: does anyone have information of how the World War 2 era wartime economic planning of countries other than the Soviet Union worked? How were the necessary calculations made, and the results carried out? Might be interesting in this context.

[alice]

@Raphael

Note that in 20th century command economies the planning was done by hand while nowdays we have powerful computers with stuff like AI and Big Data

See also malloc's AI thread Don't forget that Mises and Hayek were ultimately about finding reasons for not paying taxes.

[zompist]

In short - it's a theory developed by Mises and Hayek saying that a centrally planned economy can never match the efficiency of a market one because due to lack of price signals, the central planners will never be able to correctly determine the demand for each product, and they'll not be able to react with the same speed and elasticity as the market.

The argument goes back to an article by von Mises in 1920, which should be a warning sign right there: it's based on no real-world data, and its ideas about the market are a hundred years out of date, and ideologically suspect even then. (Austria wasn't exactly a poster child for capitalism in 1920.)

Actual experience is more complex. Soviet and Chinese experience both suggest that a period of communism works pretty well to turn a premodern society into an industrial one— the worry in the 1930s was that capitalism was in third place behind the totalitarian systems. But neither system could keep up the gains, in part because that transition can only be done once. The Soviet system certainly was not impossible, but it was notoriously bad at feeding people and producing consumer goods. The Chinese system ran aground on Mao's hubris, which is a political not an economic problem.

Neither author had access to any data on mixed systems... which is pretty much all of them, since WWII. By now it seems pretty clear that governments are better at providing some things (health care, utilities), and regulated markets are better at others (consumer goods, art). For big items like cars, who knows? Renault seemed to do OK under the French government.... it tripped up in the 90s, but a lot of private car companies failed too. I wonder what von Mises would make of "too big to fail" payouts.

The authors did not, I think, account for failures of capitalism. The Great Depression is widely considered to make laissez-faire capitalism impossible too. (The US, that bastion of market fundamentalists, got out of it by government intervention, including that huge experiment in central planning called WWII.) Then there's the monopoly/cartel problem, which the authors should have been aware of. Due to founder effects, many inventions aren't really subject to market pressures, and become rent-seeking parasites instead. And then there's externalities, which neither capitalism nor socialism can inherently handle. And then the problem of bubbles, a hugely distorting problem with investor classes.

All this means that conworlders are pretty free to imagine all sorts of systems. rotting bones has argued for, if I understand it, centralized ownership guided by democratic votes. This could be an alternative to the market: if everybody votes for more avocados, we get more avocados; once there's too many, people vote for something else. As the authors never contemplated such a thing, they can't be said to have refuted it.

I'm skeptical that computers somehow solve the problem... in part because I know software projects too well; they are not perfect soulless machines, but cobbled together according to the often contradictory whims of managers, programmers, and users. But also because the raw data of economics— people's preferences— are not accessible. We can guess at them in a market, but history is full of bad guesses and unexpected visions. To give just one example, no one in 1980 could have calculated that a huge industry would develop selling people coffee. People didn't know they wanted that till they saw it.

[Otto Kretschmer]

@zompist

Thanks for your answer.

Was that avocado reference intentional?

[Ares Land]

Neither author had access to any data on mixed systems... which is pretty much all of them, since WWII. By now it seems pretty clear that governments are better at providing some things (health care, utilities), and regulated markets are better at others (consumer goods, art). For big items like cars, who knows? Renault seemed to do OK under the French government.... it tripped up in the 90s, but a lot of private car companies failed too. I wonder what von Mises would make of "too big to fail" payouts.

Renault was sold in the 1990s -- though the French government still owns 15% of it. I think it had a bad spell in the 80s. The company has an interesting and eventful story but all in all it was no better and no worse than its non-state owned competitors so it's hard to draw conclusions either way.
Post-war France is an interesting case, because the left-wing was basically the communist party and the right-wing, while capitalist, really wasn't sold on the free market thing. So we had five years plans and a high commissioner to the plan (the position was reinstated, though as a sinecure.) It worked until the 70s, or the 80s. For a certain value of 'work': people can be nostalgic about les trente glorieuses but I'd doubt they would have liked it very much.

Personally, I think we hit a better formula in the 1990s with both more socialism and a freer market.

My own view is that the economic calculation problem is a real and that Mises and Hayek's conclusions are generally correct; if you don't take them to ridiculous levels bordering on fundamentalism (which is, alas, the present right wing view.) I'm just an armchair economist, and probably not a very good one, but it seems mainstream economics (even on the left), tend to agree.

Thét being said, there are plenty of cases where socialism is just better. Healthcare is one of these. Part of it is that there's no question of free market or free choice in the matter: on the 'consumer' end, if you don't have healthcare you die, on the 'producer' end, letting people die is usually frowned upon. There's no great question of personal preference either; either treatments work or they don't and personal preferences are uniform: people universally prefer being healthy and alive over being sick and dead.

When it comes to socialism, I'm more interested in sharing value. I'm interested in how people get as fair a share of possible in their work and what they produce; I'm pretty much agnostic when it comes to how it's produced, as long as it works.

[Travis B.]

Neither author had access to any data on mixed systems... which is pretty much all of them, since WWII. By now it seems pretty clear that governments are better at providing some things (health care, utilities), and regulated markets are better at others (consumer goods, art). For big items like cars, who knows? Renault seemed to do OK under the French government.... it tripped up in the 90s, but a lot of private car companies failed too. I wonder what von Mises would make of "too big to fail" payouts.

Renault was sold in the 1990s -- though the French government still owns 15% of it. I think it had a bad spell in the 80s. The company has an interesting and eventful story but all in all it was no better and no worse than its non-state owned competitors so it's hard to draw conclusions either way.
Post-war France is an interesting case, because the left-wing was basically the communist party and the right-wing, while capitalist, really wasn't sold on the free market thing. So we had five years plans and a high commissioner to the plan (the position was reinstated, though as a sinecure.) It worked until the 70s, or the 80s. For a certain value of 'work': people can be nostalgic about les trente glorieuses but I'd doubt they would have liked it very much.

Personally, I think we hit a better formula in the 1990s with both more socialism and a freer market.

My own view is that the economic calculation problem is a real and that Mises and Hayek's conclusions are generally correct; if you don't take them to ridiculous levels bordering on fundamentalism (which is, alas, the present right wing view.) I'm just an armchair economist, and probably not a very good one, but it seems mainstream economics (even on the left), tend to agree.

Thét being said, there are plenty of cases where socialism is just better. Healthcare is one of these. Part of it is that there's no question of free market or free choice in the matter: on the 'consumer' end, if you don't have healthcare you die, on the 'producer' end, letting people die is usually frowned upon. There's no great question of personal preference either; either treatments work or they don't and personal preferences are uniform: people universally prefer being healthy and alive over being sick and dead.

When it comes to socialism, I'm more interested in sharing value. I'm interested in how people get as fair a share of possible in their work and what they produce; I'm pretty much agnostic when it comes to how it's produced, as long as it works.

My personal view of things is that the economy can be split up into three things:

1. Things of very low marginal cost (but often very high total cost regardless of number of units produced); examples are software, movies and TV shows, and musical recordings
2. Things that behave like normal goods with non-trivial marginal costs
3. Essential services that everyone needs which are often very expensive (housing, healthcare, education, etc.)

1 and 3 would be paid for through taxes -- with there solely being nation-wide income and wealth taxes applied to all entities, both individuals and enterprises, that would be distributed down equally on a per capita basis to local bodies to avoid the problem whereby some locales get more money than others due to the nature of property taxes -- by the government, and 1 in particular should follow a model similar to what rotting bones advocates (except that for things such as software not just people but enterprises would be able to vote because many pieces of essential software, such as the Linux kernel, are highly necessary to enterprises but have low visibility to the general public). Note that all software funded through the government would be required to be FLOSS. 2 would behave like under a normal market economy, except that some essential things like food would be subsidized by the government.

As for enterprises themselves, I am for all enterprises either being worker cooperatives or being hybrid worker-consumer cooperatives (which are owned by both their workers and their consumers). I am against the idea of direct government ownership of enterprises to avoid the problems of state capitalism, but I am for very robust regulation of enterprises by the government. For instance, in the healthcare sector, all healthcare enterprises would be organized as cooperatives but would be solely funded by the government and heavily regulated by the government.

[Raphael]

I wouldn't really have a problem with Travis's proposals, although I don't mind "state capitalism" as much as he does. That said, if this is now a thread for our economic policy ideas, I'll post something that has been developing a bit inside my head for a while:

How about having a kind of limited explicit economic neo-feudalism?

When people these days talk about "feudalism" in economic contexts, they usually mean it as something bad - for instance, someone might say that many workers in the current stage of capitalism are effectively living under a new feudalism, and that's a very bad thing. But at the higher levels of the hierarchy, a new form of vassalage might be better than what we have now.

The idea would be that the public owns all the, for lack of a better word, businesses, but instead of some central planning authority telling them all what to do, they would be given "in fief" to specific businesspeople, who would get to manage them, but always under the rule that they don't get to own them, and can lose them. The crucial thing would be that keeping your "fief" would always come with conditions, such as creating jobs, and not just any jobs, but jobs that people can live off.

Compared to regular capitalism, this would have the advantage that it would remove the investor class, and all the problems they bring us. And compared to more conventional forms of socialism, it would have the advantage that there would still be some space for "entrepreneurial" impulses, just in case that they might actually be useful.

[WeepingElf]

That is an interesting idea for sure, Raphael. Would you mind me adopting it for my classical Elves, whom I have always envisioned to have an economic system with both liberal and socialist traits, though I never found a viable solution? (If you want me to acknowledge you by your real name in the planned web site about the Elves and future novels, please tell me by PM.)

[Raphael]

That is an interesting idea for sure, Raphael. Would you mind me adopting it for my classical Elves, whom I have always envisioned to have an economic system with both liberal and socialist traits, though I never found a viable solution? (If you want me to acknowledge you by your real name in the planned web site about the Elves and future novels, please tell me by PM.)

I don't mind you adapting it, and you don't need to acknowledge me. Thank you!

[WeepingElf]

That is an interesting idea for sure, Raphael. Would you mind me adopting it for my classical Elves, whom I have always envisioned to have an economic system with both liberal and socialist traits, though I never found a viable solution? (If you want me to acknowledge you by your real name in the planned web site about the Elves and future novels, please tell me by PM.)

I don't mind you adapting it, and you don't need to acknowledge me. Thank you!

Thank you, too.

[zompist]

Post-war France is an interesting case, because the left-wing was basically the communist party and the right-wing, while capitalist, really wasn't sold on the free market thing. So we had five years plans and a high commissioner to the plan (the position was reinstated, though as a sinecure.) It worked until the 70s, or the 80s. For a certain value of 'work': people can be nostalgic about les trente glorieuses but I'd doubt they would have liked it very much. [...]

My own view is that the economic calculation problem is a real and that Mises and Hayek's conclusions are generally correct; if you don't take them to ridiculous levels bordering on fundamentalism (which is, alas, the present right wing view.) I'm just an armchair economist, and probably not a very good one, but it seems mainstream economics (even on the left), tend to agree.

These kind of contradict, don't they? You basically say that some central planning worked, and then that Mises & von Hayek are right that it's impossible.

But then, I think the problem with M&H is they're absolutists. They weren't even looking at the Soviet experience, much less the French; they were applying pure ideology to an imaginary form of central planning. I think the only reason the idea isn't simply thrown out is that a) it supports economists' pro-market bias, and b) at a rushed glance, it fits Soviet history.

We should be very skeptical of any theory that says "well actually the way we do things now is the only rational way." Economists tend to be very prone to this, e.g. continuing to promote Ricardo's comparative advantage model, which completely fails to address development economics.

(Why did the Soviets do so badly? That central planning is difficult is part of the answer. But there's also the problem of information: totalitarianism inhibits honest numbers. And Stalin's hubris. And little incentive for increasing productivity. And above all, politics: the decision to concentrate on industry rather than agriculture or consumer goods was a political decision, not an economic one.)

I'd also note that absolutely no one is pursuing laissez-faire economics with no government interference, and that includes Elon Musk's goons. I think you only have to look at how techbros are cowering before the Republicans. They want to use state power to prop up their favorites rather than those of the last-century consensus.

[Raphael]

And above all, politics: the decision to concentrate on industry rather than agriculture or consumer goods was a political decision, not an economic one.)

Well, that's part of the whole problem, right? If, in a centrally planned economy, a very bad decision is made at the very top, it can crash everything. If, in an economy that is decentralized in the one or other way, a very bad decision is made at the top of one business, harm might be limited to that one business.

Oh, and zompist, a bit of pedantry: the "von" belongs to Mises, not Hayek.

[zompist]

And above all, politics: the decision to concentrate on industry rather than agriculture or consumer goods was a political decision, not an economic one.)

Well, that's part of the whole problem, right? If, in a centrally planned economy, a very bad decision is made at the very top, it can crash everything. If, in an economy that is decentralized in the one or other way, a very bad decision is made at the top of one business, harm might be limited to that one business.

Sure, I'm not arguing that centralized decisionmaking is good, only that M&H didn't prove it was impossible.

Spreading out power is, in general, a good thing. Sure isn't preventing very bad decisions being made in this country, though.

[Ares Land]

These kind of contradict, don't they? You basically say that some central planning worked, and then that Mises & von Hayek are right that it's impossible.
But then, I think the problem with M&H is they're absolutists.

I probably could've been a lot clearer . The problem is indeed with the absolutism.
The point that can be taken from it is that markets tend to be more efficient, there's even as I recall explanations as to why it is so, not that central planning is impossible. I don't think they ever managed to prove such a thing.

It's perhaps worth mentioning that Hayek started out as a reasonable economist. Certainly leaning to the right, but otherwise reasonable. I actually read The Road to Serfdom, though that was ages ago, and it's as I recall a fairly sober work. It's pretty far away from libertarianism, and I think even some distance away from Thatcherism. I think Hayek turned increasingly crankish with age, and of course his Austrian school or Chicago school successors went even crazier.

To say nothing of Trump/Elon!

I'd also note that absolutely no one is pursuing laissez-faire economics with no government interference, and that includes Elon Musk's goons. I think you only have to look at how techbros are cowering before the Republicans. They want to use state power to prop up their favorites rather than those of the last-century consensus.

I must say it feels a bit weird discussing libertarians -- because they don't exist anymore. They're all fascists now.

[Raphael]

It's perhaps worth mentioning that Hayek started out as a reasonable economist. Certainly leaning to the right, but otherwise reasonable. I actually read The Road to Serfdom, though that was ages ago, and it's as I recall a fairly sober work.

I don't find the idea that having a welfare state inevitably leads to Nazism all that reasonable or sober, even if it is written down in a reasonable and sober tone.

[Ares Land]

It's perhaps worth mentioning that Hayek started out as a reasonable economist. Certainly leaning to the right, but otherwise reasonable. I actually read The Road to Serfdom, though that was ages ago, and it's as I recall a fairly sober work.

I don't find the idea that having a welfare state inevitably leads to Nazism all that reasonable or sober, even if it is written down in a reasonable and sober tone.

It's really not about the welfare state at all, but about central planning, with state control over the whole economy.

So really it's arguing against economic proposals that were possibly very relevant in 1944 these days but decidedly a minority view these days.

[Raphael]

It's really not about the welfare state at all, but about central planning, with state control over the whole economy.

So really it's arguing against economic proposals that were possibly very relevant in 1944 these days but decidedly a minority view these days.

Even then, I don't agree, and as you can see, some people in this thread have the minority view you describe.