Hacker News Reader

Comment by webnrrd2k | Hacker News Reader

The feature that stands out the most to me is that the black "boom or bust" line that shows prosperity/recession is effectively never at 100. It crosses the line (going up or down) a bunch of times, but is almost never on the line. I have to wonder if more balance with this would result in better outcomes, or worse.

Also, and this is just an eyeball analysis from the pdf, the booms seem roughly balanced with the busts until the great depression. The prosperity of WWI seems to have set up the cycle of bust for the great depression.

View on HN

> I have to wonder if more balance with this would result in better outcomes, or worse.

Iirc that is exactly the problem that the Soviets claimed they solved with state-owned and centrally planned production, no more boom/bust cycle. And early on it did have some in the West worried about it. Turned out not to work, probably b/c you can't stablize an inherently unstable complex system.

Well, it worked in the sense that you can avoid a boom/bust cycle, by just having it be bust all the time...

I seem to remember that due to political infighting and ass-covering, they never really tried it. There were competing proposals that they hybridized in broken ways, and nobody in power wanted to commit to have their name on something that might fail or become unpopular.

I recommend the excellent novelization of the period “Red Plenty”. My understanding is that it stays fairly true to the history.

> I seem to remember that due to political infighting and ass-covering, they never really tried it.

What do you mean by "never really tried it"? This was a state that set production quotas, dictated prices, and set priorities in 5 year intervals. What more could they have done? The fact that even a regime that was willing to starve and enslave its own people could not make a planned economy work, just shows the futility of the approach.

The point is not about the basic principles, but how these decisions were made. This is like being a higher up in a company making a call about something, which goes from placing the bet on the wrong horse, political infighting between different departments, up to decisions which are based on incentives that go counter to the well-being of the company. We are talking about humans with all their flaws.

Especially large corpos developed more and more similarities to how eastern bloc countries were operated, believe it or not. Very top down decision making (with all the problems), large bureaucracies, people who are employed but effectively don't do anything ("bullshit jobs" / David Graeber), company propaganda (all hail to the great company!), bothersome people getting "mistreated", and so on. I've experienced the eastern bloc from inside and I'm not sure whether I should laugh about or be terribly afraid of the things that are still to come. Edit: Forgot one big thing: Metrics / quotas.

The other thing to keep in mind that "responsibilities" were split between members of the eastern bloc and the decisions who took over what that were often influenced by all kinds of things, just not what made most sense.

There was also huge trade embargo, CoCom, in place.

All your points are correct.

I think the commenter was referring to a supposedly accurate historical fiction book that describes how the system was never truly working in an honest fashion. I have a colleague that worked in optimization in the Soviet Union and he explained that factories would lie about their data in order to look better and essentially you'd have garbage in and garbage out for the models. So we don't know if the linear programming tools were truly broken or if it was because all the input data was corrupted. I think the book said the same thing.

Like you, I've become fairly suspicious of economic models that attempt to explain something complex and unstable like an economy. Equilibrium is an exception in the real world...not an inevitability and the caveats of that model are often discarded.

>What do you mean by "never really tried it"? This was a state that set production quotas, dictated prices, and set priorities in 5 year intervals. What more could they have done?

Obviously: not have different political factions and concerns from individual officials to "look good" affect the content of those plans, nor have the same factions and concerns distort the reports about the state of production and the results of said plans.

And instead to try what they purported to be doing but didn't do: plan solely based on optimization concerns, and get back non purposefully-distorted reports so that they can re-plan and course-correct as needed.

Btw, those "five year plans" are not what people think, which usually involves a mental picture of someone calculating the amount of desired production of X or Y product for the next five years and setting some prices in stone.

They rather were sets of goals and associated organized efforts on multiple fronts to meet them. Like "let's industrialize that province" or "let's build transport infrastructure", etc.

Like current multi-year "initiatives" or often still call "plans" like "The Biden- Harris Plan to Revitalize American Manufacturing and Secure Critical Supply Chains in 2022"

https://www.whitehouse.gov/briefing-room/statements-releases...

or those regarding broadband access, or the federal plan to "land a man on the moon" back in 1960.

Just make it so that the people in your system don't do politics. Obviously.

Workable at tiny scale in religious communes, or maybe for a different species.

I think it's much simpler than that. Instead of factories self-reporting, you have rivals report on each other, as well as client and suppliers report on what went in and out of the factory in a given time period.

It wouldn't be a perfect system, but like capitalism the competition would force everyone into relative honesty. It is pretty weird they didn't do that at the time though, considering the KGB kinda had that sort of setup with respect to everyone individually being required to snitch on each other.

Or you could punish people for misreporting. Would sort things out very fast!

Or automate the reporting (totally feasible in 2024).

> The fact that even a regime that was willing to starve and enslave its own people could not make a planned economy work, just shows the futility of the approach.

I disagree with this way of thinking. A willingness to apply lots of cruelty to an attempt doesn’t make up for a lack of skill or capabilities. Cruelty is often a failure-mode, not a recipe for success.

>Turned out not to work, probably b/c you can't stablize an inherently unstable complex system.

Or because they didn't have enough computing power at the time

https://chris-said.io/2016/05/11/optimizing-things-in-the-us....

The best critique I heard about the central planning system is that it’s just that, planning.

You might be able to plan, but what about the advancements that nobody would plan for.

“Don’t build a faster horse”

Would a communist system ever advance beyond the planned scientific and enginnering goals? Is there someone thinking “what people need doesn’t exist?” and actually direct resources to it?

That’s the beauty of the the free market system - you can come up with a crazy idea and just do it. You often fail, but if you succeed, the risk you took is rewarded.

I can’t ever see a communist system coming up with an iPhone. They might copy an existing product, but that’s not innovation.

> I can’t ever see a communist system coming up with an iPhone.

Soviets invented a lot of things, in terms of aviation, computers, celestial navigation in intercontinental missiles. There was also an article, which I read recently but failed to find it now, about how they revolutionized mechanical watches, and came out with their own quartz movements, without copying anyone.

Oh lastly, they did the "iPhone of sea going vehicles", the erkanoplan, and VLIW computing which led us to Itanium at some point [0].

[0]: https://www.abortretry.fail/p/the-itanic-saga

Soviet citizens were always at the mercy of central planners deciding if they got to enjoy any innovations (mostly aped from the west). Toasters only started production in 1967 and were mostly unobtanium for the next 25 years.

The watch article was on HN a few day ago.

https://news.ycombinator.com/item?id=40365956

While the title is provocatively titled, "How the Soviets revolutionized wristwatches", it seems to be more of a play on words around the Soviet revolution. The article is about how they bought watchmaking equipment from the US, looted machinery from Germany after WWII, and copied swiss designs.

Soviet watches were workhorses, but not revolutionary at all.

>You might be able to plan, but what about the advancements that nobody would plan for

You obviously adjust the plan? Planning was never supposed to be some static system (and the "5 year plans" are not about market planning, they were more like "5 year initiatives to improve this or that aspect", e.g. "develop more tank making capacity", or "take a man to the moon").

>I can’t ever see a communist system coming up with an iPhone.

Communists came up lots of inventions.

Adjust it to what?

“Let’s develop a product nobody asked for”

And I’m happy to take a look at this list of “lots of inventions communists came up with” if you want to point me to them.

Fundamentally centralized control has an information problem. Decisions about how to allocate resources require complex information that mostly exists at the edge, but communicating all of that back to the central controller would just overwhelm them. This is why markets work so well, the decisions about how to allocate the resources are pushed down to where the information is. But this is also why market are bad at solving systematic issues, in that case the information is above where the decisions are being made. You can see where there are tradeoffs and cases where one system works better than the other, so anybody who issues a blanket statement like "markets are always the best solution" is not operating with a complete picture of the world.

> but is almost never on the line

The water surface in the ocean is almost never at "sea level".

Markets are a chaotic system.

> the booms seem roughly balanced with the busts until the great depression

Yup. Like waves in the ocean!

> The prosperity of WWI seems to have set up the cycle of bust for the great depression.

The setup was to peg the dollar exchange rate with gold, and then inflate the money. Such "pegging" historically has always led to a massive correction.

> The water surface in the ocean is almost never at "sea level".

> Markets are a chaotic system.

Don't blame the boom-and-bust in the US on markets. Their heavy-handed policies make them especially crisis prone. Their northern neighbour of Canada had much smoother sailing, thanks largely to a more laissez-faire approach to banking.

How do you meaningfully compare a tiny economy like Canada’s to the 800 lb gorilla of the USA, the global military and economic superpower whose currency rules the world which is conservatively ten times larger than Canada and doesn’t need/act to police the global dollar trade?

Seems a non sequitur at best.

In the relevant time period, the US was not the global hegemon, and there was no global dollar trade.

Also a lot of the policy mistake in the US were made at the state level, and that is comparable in scale with Canada.

In any case, it's easy to compare different economies, as long as you are careful what you are doing. And, if anything, overseas demand for your currency should make it easier to run your monetary and financial system.

>The water surface in the ocean is almost never at "sea level"

I'd expect it to be "at sea level" though with very small averaging (smoothing) to account for the waves.

The economic plots however would amount to Everest-sized waves.

Historically, the US was a particularly crisis prone economy. Have a look at Canada for something much more stable. See eg https://www.cato.org/blog/there-was-no-place-canada

> I have to wonder if more balance with this would result in better outcomes, or worse.

After reading Taleb, these "boom or bust" cycles are a feature not a bug. My own simplified interpretation: You need booms to fund moonshot ideas and busts to get rid of ideas that don't work. Also a bit similar to neural network training. You first pump in all kinds of information and then aggressively prune to get rid of mostly bad information.

That doesn't explain why they all have to line up in time. If each individual boom or bust was spread over time, it would disappear in the noise.

They don't always all line up in time. Companies grow and die all the time.

Trends, Hypes and Bubbles are probably a consequence of human nature: most investors have a deep fear-of-missing out, so in general money loves to jump on the hype-train. Likewise, fear and uncertainty easily spread: markets moving together. When a couple of businesses fail, that insolvency or illiquidity spreads to related businesses.

Booms and busts have cyclic effects. Booms provide more capital and credit that can be reinvested and beget more booms; busts collapse demand and can contagiously bust other sectors.

The responses to your observation are hilarious. They'd have you think that the mass hiring by FAANG in 2021 and early 2022 followed by layoffs was a good thing. You accurately observe it as being screwy, which it is, but people are brainwashed into thinking it isn't - or is even a good thing.

I'm trying to figure what the black "boom or bust" line actually is mathematically speaking. It looks a bit like they've run a regression on GDP and show the deviation above or below trend? Maybe it's in the text? Trying to figure it.

It took me a few looks to figure this out, but out of all the various y axes on this chart, this appears to correspond to "business volume scale," which is exactly 0 on the flat line. The axis itself is found under the years 1845, 1895, and 1930.

What "business volume scale" means or how it is measured, however, I don't see specified anywhere, and after an admittedly cursory web search plus FRASER search of 20 minutes or so, I can't figure it out. The text blurb for "business activity" seems to mean it is probably some aggregation of industrial production and consumer spending, but what 0 is indexed to and what the deviations mean is less clear.

Stability will beget over-confidence which will bring exuberance. If the system feeds on its own information it might be impossible to stabilize.

And do you even want stability? The boom cycle produces exuberance that allows funding risky projects of which a small number become smashing success. The bust destroys the rest of risky project, which allows recycling of labor and resources. Arguably this is a better model for economic advancement, compared to a steady state where stagnating businesses continue to stagnate, hoarding resources in under-productive pursuits.

Reply on news.ycombinator.com