The problems of induction socialist calculation and altruism

The problem of induction is a philosophical problem about how knowledge is created. The socialist calculation problem is a problem in economics: it is impossible to do economic calculation without a free market. They may sound very different but they are actually very closely related to one another.

The problem of induction

Philosophers like to think that scientific knowledge is created by a process called induction that involves doing observations, using them to come up with an idea about how the world works and then showing that idea is true or probable with more observations. The problem is that induction is impossible.

Observations don’t imply any particular idea about how the world works. Any such idea implies a lot about stuff we don’t observe. Our best idea about how the sun works implies stuff about the core of the sun, which we can’t observe. Nobody has ever observed a dinosaur, only a dinosaur skeleton, but those theories are not primarily about skeletons. As a result of this it is impossible to invent an idea or to prove it is true or probably true.

In addition, it is impossible to do an observation without having some explanation of what you want to observe and why. So ideas are required for observations and cannot be created by doing observations.

Rather, knowledge is created by a process that does not resemble induction in any important respect. First, you look for problems with your current ideas. A problem is just anything that seems worth changing. You then propose guesses about how to solve these problems. You look for criticisms of the proposed solutions and eliminate criticised solutions until only one is left. You then look for new problems with your new set of ideas.

Note that there is no step of trying to show your ideas are good or probable. This is just not possible because all of your ideas about how to solve your problems are guesses. And since all of your ideas about how to test stuff are solutions to problems, all of them are guesses too. So all of your knowledge is guesswork. It is not confirmed or shown to be true or anything like that. Rather, you try to get rid of bad ideas through criticism. This means, in particular, that all of your ideas may be flawed and you should be willing to reconsider any idea.

Another important issue is that it could hardly be the case that your proposals for how to solve problems could be anything other than guesses. If you knew in advance how to solve a problem, then you wouldn’t have that problem in the first place. Sometimes people make the right guess about a problem the first time they say something about it but that must be a result of them having tried and discarded ideas before they said anything about it.

There is nothing about this discussion that limits its conclusions to science. Any process that creates any kind of knowledge (useful or explanatory information) has to proceed by variation of current knowledge and selection among those variations.

The socialist calculation problem

Socialists like the idea that people who are able to produce should give stuff to people who are not able to produce. This idea is morally bad for reasons I will explain later, but let’s leave that aside for the moment and think about whether you could actually run the world like this.

Let’s consider the problem of whether we should make flour and if so how we should make it, and how we should distribute it. According to socialism we’re supposed to do this by considering need, so let’s try to do that.

Let’s start with somebody who is hungry: let’s call him Jack. If you give Jack a bag of flour he might eat it. How much flour should you give Jack? If you drive a dump truck up to his house with a ton of flour and dump it in his grade, then he might not like that too much. He might not be able to eat it before it starts going off and it might attract vermin. So you should give him less than a ton and he won’t want you to dump it in his garden. But exactly how much should you give him? And how should the flour be packaged?

And the problem is worse than that. Jack might want to use the flour to make bread. So then he needs to have the other ingredients of bread and without those ingredients he might not want the flour at all.

But there is more complexity to come. If you sent the flour to a baker who makes bread and the baker gave Jack the bread, Jack would also eat the bread. So should you give any flour to Jack? Maybe you should just give him bread.

Another problem: whether we give Jack flour directly or give it to the baker to make bread the flour has to be made somewhere. In the place where you make it you can’t make many other things. You can’t have a factory that makes computer chips and a flour factory in the same place.

Indeed, you might even want to start making a factory for a product that doesn’t exist yet. You might have an idea for something you could invent that lots of people would want and maybe you should get some space to make it now.

This is starting to look very complicated. It looks like you have to take into account lots of knowledge you can’t have. Is this starting to sound familiar? Doesn’t it sound a bit like trying to come up with a scientific idea that covers lots of stuff you haven’t seen? If you want to make stuff for other people then you need to have lots of knowledge about how those other people will respond to what you’re doing, which is an emergent consequence of the laws of physics, biology chemistry, epistemology and other stuff. The solution to this problem has to be created by variation and selection of current knowledge.

So let’s suppose you know how to make flour. To make the flour you need certain items, like corn, say. So if you’re going to keep making flour you have to get a new supply of the stuff required to make if you want to continue. Unless the person who wants the flour happens to have exactly what you need to make it then he has to give me something you can exchange for stuff you can use to make flour. Now you might imagine he could give you stuff that a particular person wants if that person could supply the stuff you need to make the flour. But all sorts of things could go wrong with that. The person who makes the stuff I need might decide to do something else instead. Or he might have a change of circumstances that means he needs slightly different stuff. So what is really needed is something that he can exchange with other people to get what he wants. What is needed, in short, is a good that can be exchanged for anything – a medium of exchange. We have a name for that good: it is called money.

If you get more money by selling your flour to a baker than to Jack you can make more flour. If you make more money by making the flour in a different way, or with a different variety of corn or whatever then you can make more flour. You can also do other stuff with the money, like buying yourself food or an iPhone or whatever. So if different ways of making flour seem equally attractive in other respects you can choose among them by how much money they make. So the free market solves economic problems, not socialism.

Just like when we’re creating scientific knowledge, economic knowledge has to be created by looking for problems, guessing solutions, selecting among those solutions and then looking for more problems.

A moral flaw of socialism

To create knowledge you have to find problems. Socialism recommends looking at problems other people have and then trying to solve those problems. This is a bad idea shared by many other ideologies: let’s call it altruism. To solve a problem you have to try to understand it. So if you are trying to solve Jack’s problems then who is going to work on the problems that you know more about than anyone else? Nobody. So those problems won’t be solved.

And since you have to spend all of your time catching up to the other person’s problems, you are going to be interfering in that other person’s life in a ham fisted way.

The problems you should try to solve are the problems you know about, the problems you are interested in. You shouldn’t be trying to solve another person’s problems. You can help other people when cooperating with them helps you to solve your problems but that is very different from making it your aim to solve their problems.

Similar problems arise with many other political and moral ideologies that aim at solving another person’s problems. Some conservatives like to think they can solve the problems of poor people by encouraging them to get married. Some libertarians like to claim they can solve everybody else’s problems. Walter Block claims that Nazis can be libertarians if only they are willing to use persuasion rather than force to get Jews into gas chambers. Walter Block ought to have realised that gross irrationality like wanting to murder Jews is incompatible with liberty but he was paying too much attention to their problems and not enough to problems with his own knowledge.

To create knowledge about science or how to live better or anything else you have to start with problems you know something about: problems you are interested in. You propose solutions to those problems, select among the solutions by looking for criticisms and then look for new problems with the surviving solution.

Further reading

On induction: Realism and the Aim of Science and Objective Knowledge by Karl Popper. The Beginning of Infinity and The Fabric of Reality by David Deutsch.

On the socialist calculation problem: Socialism and Human Action by Ludwig von Mises.

On the moral problems of altruism more generally: Atlas Shrugged and The Fountainhead by Ayn Rand.