Hacker News Reader

Comment by moralsupply | Hacker News Reader

Central planning might appear as a "good idea", until your linear programming model has to deal with factors of production that are counted in small quantities (eg. highly specialized machinery), at which point the so called "tragedy of the integers" degenerates any model you throw at the problem.

As it turns out, the "economy" is a distributed intelligence, in which only individual actors know what's best for themselves, who react according to all sorts of information, and the overall effect of their interactions is codified in the prices of products and services. Trying to replace that with a "central brain" inevitably leads to severe misallocation of resources.

View on HN

Well? So does a "distributed" economy. Goods are manufactured, shipped across the planet, to be dumped in a landfill. A significant portion of food goes to waste, all the while some other people starve (incidentally, being forced to plant cash crops e.g. cocoa rather than grow food for themselves). We have given trillions to help pay for the losses of the bankers, who are the people who truly rule us. We are currently in the process of giving trillions in subsidies to industries which have either already doomed us to extinction or soon will. The list goes on.

The trillion dollar bailouts are central planning. The fact that consumer goods are manufactured in countries by people who can't afford to buy them and sold to people in other countries who throw them away is a symptom of the vast disparity in cost of doing business in the respective countries. Subsidies are always and everywhere the product of a central planner.

In order to criticize a distributed system properly it is necessary to identify those things that are the product of the distributed system, rather than those things that are the product of centralized planning.

> The trillion dollar bailouts are central planning.

There's a difference between an measurement that's imposed by a dominant economic actor and having a centrally planned economy.

Bailouts were arguably applied centrally by a central bank, but they are distributed throughout all economic actors that apply for bailouts and meet requirements.

I agree there is a difference. Also “meet requirements” is intended to portray the process as somewhat objective but its not. They write those requirements with an eye towards who they want to get the bailout.

The trillion dollar bailouts are indeed central planning, but they are a reaction to a failure in distributed planners as recognised by the most firm believers in distributed planning. As in, "Oh crap, this free market is in the process of unravelling, the only way to stop it is to throw a bailout at it." So it's still a product of decentralised planning (to use your term).

The reaction following a bailout is usually, "whoops, we didn't do enough central planning beforehand" i.e. the new banking regulations such as tighter loan-to-deposit ratios. So, overall, this supports the parent comment's point of the inefficiencies of decentralised planning.

USSR, as the country, failed because of central planning. Bailout is not a good solution too, because capital flow from competitive players to incompetent, which is bad for evolution. US banking system is far behind other banking systems because of that.

I absolutely was not expressing an opinion either way. All I was saying is that the parent comment was raising a point that supported the opposite argument than they thought, not that I agreed or disagreed with it.

> they are a reaction to a failure in distributed planners as recognised by the most firm believers in distributed planning.

No, they are a reaction to the perceived failures of “distributed planning” as perceived by advocates of central planning. Free marketers have always said that those banks need to be allowed to fail because they made bad decisions and bankruptcy sells their assets so the creditors can get paid and new management can operate them.

With bailout you have this bizarre system where profits are privatized and losses are socialized, which is not a distributed system or a free market.

Bailouts are a necessary resort when the market fails to self regulate. Which spoiler alert, it has always done cyclically since capitalism is a thing.

The question is not if there will be another crisis, is when it will be. Under a free market, economical crisis are a certainty.

Nothing makes them necessary, and the free market regulates by allowing those businesses to fail. Since we don’t have a free market, but a regulated one, those crises you refer to are symptoms of a regulated market.

Cyclic failures are also a result of the banking system and their use of scrip as opposed to specie.

> The trillion dollar bailouts are central planning.

> In order to criticize a distributed system properly it is necessary to identify those things that are the product of the distributed system, rather than those things that are the product of centralized planning.

What you call "central planning" is merely the emergent behaviour of the "distributed" capitalist system. A certain bearded German economist warned that in such an economy, the state is just another apparatus of the bourgeoisie.

Turning the old cliché on its head, pure capitalism only works in theory, never in practice. In the real world, disparities in money are disparities in power, which ensure things such as bailouts and more fundamentally, banking laws that are effectively a license (for some) to print free money for themselves.

> A certain bearded German economist warned that in such an economy, the state is just another apparatus of the bourgeoisie.

That bearded economist was on to something. Alas nowadays policy advocates in his tradition seem to want the bourgeoisie state to write and enforce more rules, forgetting that those rules are unlikely to benefit anyone other than an oligarch.

> pure capitalism only works in theory, never in practice.

Thats dogma and unlikely to be true.

> In the real world, disparities in money are disparities in power, which ensure things such as bailouts and more fundamentally, banking laws that are effectively a license (for some) to print free money for themselves.

Its more accurate to say that "disparities in power lead to disparities in money." Certainly there are other forms of power, like the ones people use to expropriate money and assets from those who have money and assets but less power.

> Goods are manufactured, shipped across the planet, to be dumped in a landfill.

Goods are shipped because someone bought them with their hard-earned money to fulfill a need they personally have. Whatever they do with it afterwards is up to them.

I fail to see how central planning comes into play in this scenario, unless you're trying to claim that you personally know best than anyone else what to do with other people's money or needs.

> A significant portion of food goes to waste

I fail to see the relevance of your comment. Demand is not predicted or specified by a crisp value, and even supply is subjected to stochastic variations. Thus you either underprovision, which causes scarsity, or you over provision, which increases the odds all needs are fulfilled at the risk of increasing losses.

So the problem boils down to either food shortages or "waste".

So, with this in mind, again what's the problem with "waste"?

> all the while some other people starve

I fail to see the rational basis and relevance of your argument. It reads like apuerile appeal to emotion, in fact. I mean, if I forget some leftovers in the back of the fridge and later I decide to just throw it out to avoid any risk of suffering through a bout of food poisoning, am I contributing to a famine somewhere in the world? Would those fictional and hypothetical starved people benefit in any way if I didn't threw out my leftovers? Or would I be contributing to end world hunger if I ate all my food and left no leftovers at the end of a meal? And does the driver of a truck transporting food supplies be found guilty of causing famines if he has an accident which ends up destroying his cargo?

> We have given trillions to help pay for the losses of the bankers (...)

And here comes the rest of the anticapitalist bingo. It's as if centrally planned economies weren't forced to shift resources around during economic crises to avoid collapsing. I mean, al those shipments from the URSS to Cuba were never bailouts, right?

> Goods are shipped because someone bought them with their hard-earned money to fulfill a need they personally have. Whatever they do with it afterwards is up to them.

This sets the tone for any further conversation we might have. You are simply a fundamentalist, a dogmatist. People earn money because they work hard, then they buy stuff with that money, then they do things with that stuff. And that's how our economy works! I don't even know how to reply to such an absurd statement. It's like someone saying the sky is pink. What are you even supposed to argue other than... just look out your window?

> points two and three

Glad you decided to split my sentence in two in order to argue against something that I didn't say. A significant portion of food goes to waste while there are shortages in other places, this is what I said. You can argue all you wan't but you cannot change the basic reality that your system for allocating resources isn't doing a good job at allocating resources.

> We have given trillions to help pay for the losses of the bankers

Bankers (class I people) are given free money that they then lend out, with interest, to the rest of us (class II people) who pay tithe to them for the economy to work at all. Not content with this, if they ever make a serious fuck-up and tank the whole economy, we will be there to compensate them for their losses and make sure that they still get their due profit. This is not right, full stop.

> Central planning might appear as a "good idea", until your linear programming model has to deal with factors of production that are counted in small quantities (eg. highly specialized machinery), at which point the so called "tragedy of the integers" degenerates any model you throw at the problem.

Central planning might appear as a "good idea", until your linear programming model runs into a typical union bureaucrat.

Attributing the failure of union's system to "bad math" is like saying a man dying from old age dying from a heart disease.

The biggest reason, eclipsing all others, for the economic ruin of the union was the extreme, omnipresent corruption of everything economics.

And now we have two good arguments against central planning, just off the cuff. Academically and rhetorically, Hayek demolished the idea some 75 years ago.

I've worked extensively with MIP and never heard the phrase "tragedy of the integers."

Can you tell me where I can go to learn more about this outlook on issues associated with (i suppose) linear relaxations of integer problems?

> Trying to replace that with a "central brain" inevitably leads to severe misallocation of resources.

Now the question is: If a conglomerate has a dominate position in a marketplace, doesn't its administration face many of the same problems and challenges just like a central planner? In this situation, from a technical perspective, which are the similarities and differences here? Of course, one can answer, "this is why we have anti-trust laws", or "the free market still largely works, nobody can fully control an economic sector, even a conglomerate is bounded by the market". But it doesn't respond to the real question. For example, if Amazon or Google is trying to use an algorithm to optimize something globally, intuitively, it seems that they are effectively solving the same type of economic calculation problem, albeit at a much smaller scale. I made this comment because I recently saw an essay that also uses the book Red Plenty as an example to argue that many attempts on "AI", "machine learning" or "big data" by large corporations can potentially fall into the old central planning pitfalls.

What is your opinion?

> If a conglomerate has a dominate position in a marketplace, doesn't its administration face many of the same problems and challenges just like a central planner? In this situation, from a technical perspective, which are the similarities and differences here?

Yes they do. the differences are that they are still dependent on goods that their own firm doesn't produce, so the prices of those goods discipline their budget/plans to some degree, the firm is unable to use violence to compel the behavior they need to survive, and that as they call victim to the same central planning problems as governments, their marginal profit decreases so that the owner sees less profit and the price of a share of ownership decreases because of that.

But they can use violence. By capturing control of the state and shifting state policy to coerce in its favour. Or failing that, employing violence directly themselves as many large monopolies did in the 20s and 30s.

As an aside, I see this as the fundamental point of departure in language between a Marxian socialist and a libertarian-capitalist, they mean two different things by the "state": libertarians believe the state is a parasite on the market and a corrupter and destroyer of the free market. Marxists believe that the state is created _by_ the market for itself ("The executive of the modern state is nothing but a committee for managing the common affairs of the whole bourgeoisie.") Both see the other as hopelessly naive.

> Marxists believe that the state is created _by_ the market for itself

I don't think that is accurate.

Marxists believe that in practice bourgeois control of the State (in the developmental phase of capitalism, bourgeois influence on the aristocracy which controls the State) is the vehicle by which the bourgeois create and maintain the system of private property of which the market is an epiphenomenon. (And, therefore, that control of the State is necessary to abolish that system of property rights and the capitalist system that depends on it.)

On the descriptive points of this libertarian capitalists generally agree (disagreeing on the normative assessment); seeing that the purpose of the state is to enforce the system of property rights that the market relies on, and that the State is defective to the extent that it extends beyond or departs from that function.

> On the descriptive points of this libertarian capitalists generally agree (disagreeing on the normative assessment); seeing that the purpose of the state is to enforce the system of property rights that the market relies on, and that the State is defective to the extent that it extends beyond or departs from that function.

Some see the state as a capture of the dispute resolution process and feel that the market emerges as a consequence of certain norms.

> But they can use violence. By capturing control of the state and shifting state policy to coerce in its favour. Or failing that, employing violence directly themselves as many large monopolies did in the 20s and 30s.

This is true but in functional terms the corporation is then acting with the powers of a government or organized crime, and then is subject to the same arguments that are used to explain the failures/successes of those systems.

> Marxists believe that the state is created _by_ the market for itself

This is somewhat true. Its important to note that it is created or captured by certain market participants, not the market as a whole.

Big companies are little command economies with many of the problems at a smaller scale.

Having run a $100M vertical in a very large organization, it’s pretty easy to see how problems develop. There’s always a trade off between “efficiency”, standards and local/global control.

With central control, a guy in California buys the vehicle fleet, and doesn’t understand why people in Boston need a car that doesn’t get stuck in snow, etc. Another model might be a scenario where semi-autonomous divisions consume shared services as captive market participants.

AI, etc will probably exacerbate these things, as they’ll allow more growth with less people managing things.

There's a reason corporate raiders could make quick money by breaking up conglomerates.

There are hierarchical levels to the "economy" and central planning may work well at some levels. Companies at the lower levels of the hiearchy are centrally planned entities with managers assigning tasks to employees instead of using a free market to distribute them within the organization. At the middle layer, the free market links these centrally planned companies and force them to compete. It's an open question whether having a centrally planned top layer to handle national economic strategy and steer market forces is better than letting the free market handle that as well.

> Companies at the lower levels of the hiearchy are centrally planned entities with managers assigning tasks to employees instead of using a free market to distribute them within the organization.

Not really. I mean, there's a FAANG which is notorious for fostering a competitive environment between development teams even down to it's lower level employees, which includes flexible team configurations where individuals cooperate on separate projects.

As it turns out, means it's your opinion or are you stating it as a fact?

By the way, if you insist that capitalist economies are distributed, how would you explain the extreme concentration of capital in a few hands? The global markets are ultimately controlled by a handful of people.

Isn't that global centralization of market power?

> By the way, if you insist that capitalist economies are distributed, how would you explain the extreme concentration of capital in a few hands? The global markets are ultimately controlled by a handful of people.

You can't equate current-day "capitalism" with "free market capitalism". We live under state-regulated, corporatist economy, which is designed to concentrate power in the hands of big corporations that are associated with the state.

That said, even in a "free market capitalist" system you'd expect concentration of wealth, for the reason that any competitive system presents an exponential distribution of performance in the outcomes. What you wouldn't expect, however, is that the concentration of wealth would be so skewed as we have in a corporatist economy.

On the top of that, inequality on itself is not a problem. I'd argue it's a necessity if you want to be able to create wealth, as it's a side-effect of the incentives associated with competition (as stated above). In other words, inequality is not so much of an issue if the whole of the economy is producing more wealth, which in one way or another becomes distributed as it increases productivity. In the end, the real matter is poverty -- but from all solutions we can think on how to solve that problem, centralization and bureaucracy (here meaning: a state) seem to be the worst of all. In my personal opinion, localism (meaning: acting in the smaller community scale, as opposed to central control) and voluntaryism (meaning: the ethics of voluntary associations, as opposed to coercion) are the keys to set us up to help each other and eradicate poverty.

Reply on news.ycombinator.com