The Use of Knowledge in Society (1945) [pdf]
kysq.org
kysq.org
- Just because there are restoring forces doesn't mean stability is reached. This is basic control theory, but it escaped economists for a long time. There's always nonzero lag. Often there's a lot of lag. This can move things out of the stable region.
- There's an assumption here that a market economy is a competitive market economy. This breaks down when the number of major players in a market is small. Or somebody has a "moat". Read Thiel's "Zero to One".
- Hayek was writing in an era when manufacturing dominated, production cost exceeded marketing cost, and the size of companies was limited by inefficiencies in coordinating a really large organization. Those constraints favor a market economy. Today, services dominate, marketing often costs more than production (which means most of the cost is advertising), and computing has made it possible to scale companies to planetary scale without organizational collapse. Those factors favor sheer scale.
Aside, there's also a fundamental trade-off been a competitive market and a market where individual actors have freedom.
For example, one model may assume that everyone has perfect price information and therefore the market is very efficient, while another assumes that everyone can make secret deals and keep prices hidden and therefore cartels are unstable.
And again just last year everyone got fully bailed out when SVB collapsed.
This type of stuff encourages all kinds of halucinations.
I wonder. Say we had various factories for producing goods around the world, and different factories requested the same inputs in different amounts. An algorithm could probably calculate how much of each input it should distribute to those factories. However, the structure of this economy still relies on the commodity form, since those inputs would almost certainly be worked in a certain, standardized way to be properly organized by the algorithm properly. Therefore, the computerization of the economy only leads to an image of the kind of unrestrained productivity you might imagine technology would bring, but lacks the kind of creativity necessary for it.
This is all to say that if not the market, then whatever else could've managed an incomprehensibly complex economy certainly existed before contemporary computers (and certainly, the structure of the economy in even the 19th century in early high capitalism itself was the "computer," the quantitative management of social forces). Its not as if our era, today, is particularly unique, or that Hayek was right then or wrong now, but rather that he was fundamentally wrong to believe that the market itself was the path to freedom, even then.
That's a manufacturing planning system. Each product has a bill of materials, and all those parts have to be present before the product is assembled. Factories which make many similar products either have to have very good systems to get all the parts in the right place at the right time, or carry excessive inventories. It's not a market-based system. The main goal is minimizing stalls, where production stops due to lack of a part. The cost of stalls usually dominates the cost of the parts, because revenue is zero until the product is complete.
It's especially unfortunate that writers assume everyone is on the same page about what "knowledge" means because the meaning of the term seems to have changed a lot over the last few hundred years.
What I personally find very interesting is the age of that paper and the fact that, the discrepancy between model most commonly presented in school ( in my case, variant of perfectly rational individual ) and real life has been voiced nearly eight decades ago and is not seriously discussed as foundational knowledge when it comes to economics.
But yes the definitions of several other terms seem slippery and to change from one use to the next, including knowledge.
Look up "logical positivism" as a philosophical system. It became the dominant academic philosophy for a while (including around 1945). It asserted almost exactly that - that there was no knowledge apart from scientific knowledge (although I believe it did accept direct sensory input as knowledge, and also the operations of pure logic).
It's difficult now to see the water that he was swimming in then.
His conclusion is that knowledge is dispersed, represented in prices that arise from markets which is simply understood as cooperation. Knowledge may frankly be expressed as whatever enables and motivates someone to offer something on the market. If that's too vague for your taste, too bad.
The wonder is that uncoordinated, independent cooperation gives rise to such abundant and sophisticated products. Not only that it's in a system that expresses everyone's individual preferences and competing interests. Central planning fails spectacularly to do this, and there's no reason to believe that it ever will even with fantastic computational power.
Knowledge in economic terms is much less definite because prices reflect subjective value.
Its definition being difficult to pin down is a feature--not a single knowledge, but a multitude of (in)compatible knowledges and messy knowledge-making processes, and the many different interests that arise from it.
I'd strongly recommend to folks interested in this to pick up the Stiglitz, et al 2001 Nobel Prize on information asymmetries-research https://www.nobelprize.org/uploads/2018/06/advanced-economic...
Are you suggesting that I'm trying to curry political favor with...Stiglitz (or someone else)?
On the contrary, his point is that the equilibrium price in a decentralized market is a good sufficient statistic that aggregates the current demand and supply situation.
Building on his example on page 525, if more screws of a particular size are suddenly in higher demand, then the price will increase, as it should!
The goal is not to stabilize the price but to have the price reflect the marginal opportunity cost.
Most famously interest rates without some government (or other, in the case of Crypto) hand in distribution and projected distribution, the market can fail (everyone is encouraged to hoard).
In Stiglitz' case (not looking it up, but from memory), used car markets fail. While the marginal net opportunity cost is what the price yields, it creates a negative feedback loop where people that have a used car that's more valuable than is verifiable exit the market, and then you get .... all the more 'lemons' -- i.e. only bad cars). Dealerships are one way to correct for that information loss, but markets don't always value sufficiently the information that will solve it.
We can be a bit more smug/hopeful nowadays, because information is a lot more easily aggregated/hosted. But we have to recognize the .... value of those components.
This is basically the "prices are all we need" of economics. It's written in historical context when some still economists thought that a centrally planned economy could work. Hayek writes about the price system and how it enables an economy to function in a decentralized manner, and why it can't function without it. Hayek argues that it's essential that the decisions are made with local knowledge, because every individual possesses private and unique knowledge, which is not available to central planners.
On the other hand, all the information which an individual needs from other individuals is transmitted through prices, i.e. everyone only needs to know how to make best use of the prices they see. Thus, there's no need for any kind of oracle or central entity which knows what's going on in the economy to make it function.
This is still relevant of course, in the way that most people don't realize how magical the price system is, and how humans basically just stumbled upon it without anyone understanding it.
If you believe the Fed/ability-of-the-Fed to smooth the boom/bust cycle, then you disagree with Hayek -- he wasn't (just) arguing for a generally free market -- he believed that all markets were perfect (especially/including the price value of Money).
It turned out Keynes was right.
I think it's "too soon" to say that Keynes was right. Afaik, Hayek predicted the Great Depression based on the Austrian business cycle theory. I think that ABCT is mostly right, but it's probably imperfect. There's so much going on in the real world that it's almost impossible to say whether a policy or a theory or whatever actually caused something or didn't cause, and what would have happened otherwise.
This is for the simple reason that prices more or less only communicate information about the amount of labor required to produce a thing[0].
Therefore prices on their own are, for example, incapable of transmitting information about what action needs to be taken to correct the relationship to the biosphere. Information about the state of the biosphere will only enter into prices to the extent that things start taking more labor to produce. But there’s no market mechanism that would then cause that to direct action towards stabilizing the climate.
[0]: This is because cost resolves into business owner’s cut + labor cost + cost of inputs, and the inputs can recursively be split into the same until you’re left with the amount owners take, the amount paid to workers, and the amount paid to owners of natural resources.
The business owner’s cut and the amount paid to owners of natural resources are socially determined and bear almost no relationship to the physical world or reproduction of society.
Compare and contrast the repost today of The Singularity - would Hayek believe that a computer (with the goal of being an Economy Maximizer) would even theoretically be able to have enough knowledge to implement economic equilibrium (and maintain that equilibrium over time)?