Optimizing things in the USSR (2016)

The efforts to sabotage and disband central planning started as early as 1954.

In 1954 an executive order of 14 Oct 1954 reduced the amount of administrative personnel by 450 thousands.

The amount of metrics went down from 9 940 to 6 308 in 1954, to 3 081 in 1955, and to 1 780 in 1958.

Khrushchev moved most of the planning power from central planning institution to the regions and down to the factories and enterprises. What previously was strict targets from the center now became soft suggestions.

Article misses those basic things completely and instead goes into tangential theoretic topics and even goes to state things like that:

  central planners preferred to use the CIA’s analyses of certain Russian commodities rather than reports from local Party bosses

I wonder if one could make a fun and entertaining game that's inspired by Soviet central planning and Factorio alike - so much of Factorio is balancing raw inputs, intermediate products, and the need for "final" products (e.g., ammunition, rocket fuel, etc.).

I suspect that you can definitely use planned economy tools down to a certain scale and then let the markets sort the rest.

At some level, Indian and Chinese economies have their development centrally planned, not just financially regulated, like in the West.

Probably a lesson in all this for management-by-poorly-conceived-metrics dashboard enjoyers

https://bactra.org/weblog/918.html

In USSR, Optimization Solves You.

Markets are optimal but capitalism is not. If you let supply and demand set prices but don't allow extracting value you get an economy that is both fair and efficient. The rule is all profits must be reinvested or paid out to workers and the state can impose price controls if margins are too high.

Yugoslavia tried it and they had the best economy from the eastern bloc growing faster than a lot of the west. China and Vietnam went further in their reforms and allowed private investment but they still kept state owned corporations instead of privatizing everything, and they don't worship markets and property like America does.

I don't think it's an accident that all these mixed economies got ahead of pure socialist or capitalist systems.

"Red Plenty" the book being discussed in the article is well worth a read.

The computational complexity is a bit of a red herring. Finding the most optimal solution is hard. But if you settle for a close-to-optimal solution, then it can be solved in a much more reasonable time.

(2016) but still worth the read

It is worrying that in the first "describing history" he describes the main data quality problem being that factory managers and party bosses outright lied to the central planners and to the communist party in Moscow, and then totally disregards that problem in the data quality issues for the future. Elsewhere you find the same: the very people who needed to secure against lying factory managers simply took their cut.

I've heard the same problem described from Chinese people as occurring on an equally large scale in China, even now. That what people the control Beijing has is vastly overstated due to the very widespread lying to acquire resources. Followed by complaints that the lying goes the other way as well: local governments telling home owners that a rail line will be built through their house, "disown them" with documents nobody in government knows about, then a company owned by a party boss builds a new house there and sells it, or doesn't even build anything, just sells the house they just acquired through fraud. Same with businesses.

I'm not sure what the fix is for this, but I can imagine that this is how the game is played in any authoritarian system. I doubt it will be much of a challenge to lie to the "internet of things".

[dead]