Yesterday I wrote something on facebook that bears repeating here:
A comprehensive understanding of the pencil problem, combined with a thorough understanding of the broken window fallacy (and its inputs and corollaries... Hazlitt for example), makes a pretty good inoculant against socioeconomic lies and stupidities.
Although they are implied by the conditions above, perhaps one should also specifically reference the scale and complexity problems, the perfect information fallacy, the perfect man fallacy, and the law of unintended consequences...Some of my readers were unfamiliar with the pencil problem.
In comments, the novelist Ryk Spoor provided a decent explanation, which I'm going to paraphrase here, with my own edits and revisions (and the addition of the last bit, about planning and control):
No one man, can make a pencil, or at least a pencil which could be sold economically.
In general terms, the pencil problem, is that even simplest and most common objects in our civilization generally require an immense number of people and inputs; to not merely build, but manufacture and sell in sufficient numbers, to make it worthwhile to build them cheaply (or at least so that they can be sold economically).
The applies to everything from cars and computers, to pencils, to paperclips.
If you wanted ONE paperclip, it would be an epic undertaking, from locating the appropriate ores, refining them, turning them into steel, figuring out how to draw the steel into the appropriate size of wire, and then finally producing the paperclip from that wire. The amount of effort involved in it would be months of your labor, assuming you had the talent and resources to do it at all.
Instead, you go to a store and buy a 100ct box of them for a dollar; or even at minimum wage, a few minutes of your time for a hundred of the things.
Multiply that by all the different types of goods and services in a modern civilized society, and it starts to become clear just how many people, in how many different specialties, with how much infrastructure, are needed to keep everything running.
Given that scale and complexity, it should also be clear how impossible it would be to plan, control, and manage, anything approaching a national economy or infrastructure centrally; or in fact in any way other than as devolved and decentralized as possible.The original statement of the problem in this way came from an essay by Milton Friedman (which was a restatement of an earlier essay from Leonard Read, which was a restatement of Hazlitt, which was a restatement of Bastiat and back down the chain).
A video of Friedman explaining the problem:
