U.S. CitiesCity-Data Forum Index
Go Back   City-Data Forum > General Forums > Economics
 [Register]
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
View detailed profile (Advanced) or search
site with Google Custom Search

Search Forums  (Advanced)
Reply Start New Thread
 
Old 08-17-2018, 06:57 PM
 
2,027 posts, read 816,055 times
Reputation: 1788

Advertisements

In a market economy, at the most micro level, you have consumers maximizing utility based on their income and the price of goods in the market. Similarly, you have producers maximizing profit by some combination of capital and labor. In aggregate, decisions by individual consumers form market demand, and decisions by firms form market supply. Classical economics teaches that the intersection of market demand and supply is Pareto Optimal. Soviet economists disagreed.

They basically said there was some inefficiency at the micro level. Consumers are not really maximizing utility because they don't have the information to do so. In other words, a consumer may compare prices of a desired good at a couple of stores; but in general, a consumer doesn't have the ability to choose the cheapest price in the whole economy. Soviet economists and mathematicians thought they had solved this "trial and error" inefficiency through linear programming.

They said they could use computers to solve these optimization problems better than the market economy. Basically the state would make sure citizens got the goods they needed the most efficient way possible with math and science. And to their credit, using linear programming to solve these questions was a fairly genius idea at the time. The problem was computational complexity. If you take an economy with millions of goods, thousands of industrial inputs, etc. it becomes insanely difficult for a computer to process. So instead, they made simpler models which were highly flawed; that coupled with politics is why so many grocery stores were empty and so many factories sat idle with no raw materials.

My question is this: given today's advances in supercomputing - would it be possible to achieve what Soviet scientists set out to do? Can you compute maximum efficiency with todays technology? I know many quantitative finance firms like to hire Russians because they are extremely adept at making algorithms to understand markets and make sh*t loads of money on the stock market. Models used by these firms are surely more complex than what they thought of in the 60's and 70's. So - its it possible?

*This isn't a communism vs. capitalism debate. It's about whether these ideas are even possible.*

Last edited by Drewjdeg; 08-17-2018 at 08:25 PM..
Reply With Quote Quick reply to this message

 
Old 08-17-2018, 08:10 PM
 
Location: Honolulu, HI
4,589 posts, read 1,145,353 times
Reputation: 6568
I think we can all agree that Communism failed so badly it's not even worth looking into whether it would be possible or practical again. No one would want it.
Reply With Quote Quick reply to this message
 
Old 08-17-2018, 08:23 PM
 
2,027 posts, read 816,055 times
Reputation: 1788
Quote:
Originally Posted by Rocko20 View Post
I think we can all agree that Communism failed so badly it's not even worth looking into whether it would be possible or practical again. No one would want it.
Yes, capitalism is objectively better than communism in many ways. That's not the point of this thread. I like to think about these things, and it's interesting to me to explore these ideas. That's why I'm asking the forum.
Reply With Quote Quick reply to this message
 
Old 08-17-2018, 08:33 PM
 
Location: Kansas City, MISSOURI
5,527 posts, read 1,659,914 times
Reputation: 4741
Hmm. Interesting question. However, it's not entirely clear to me exactly what they would need to calculate? How many acres of corn to plant? How many bars of rolled steel to produce? How many cars to produce, and what kind? You would also have to calculate some sort of demand for each of those things, which begs the question, what would be your inputs for calculating some sort of demand?

Funny thing is, I'm sure many big businesses probably do that sort of thing already these days.
Reply With Quote Quick reply to this message
 
Old 08-17-2018, 09:24 PM
 
2,027 posts, read 816,055 times
Reputation: 1788
Quote:
Originally Posted by James Bond 007 View Post
Hmm. Interesting question. However, it's not entirely clear to me exactly what they would need to calculate? How many acres of corn to plant? How many bars of rolled steel to produce? How many cars to produce, and what kind? You would also have to calculate some sort of demand for each of those things, which begs the question, what would be your inputs for calculating some sort of demand?
In the West we're taught that these problems actually have non-linear solutions. Which makes sense because of things like diminishing returns; also, it makes it way more complicated. So I'm not exactly sure how relevant this is in 2018, but this how I understand it.

The idea was that you could formulate the most efficient way to produce a good and deliver it to consumers. So if you wanted to plant corn, you could use data on climate, soil quality, etc. and pinpoint the best farms to plant corn - given that there are constraints on how much a farm can grow. The idea being that math was more accurate than the whatever a farmer decides to produce. If corn was an input, they could find the factories most efficient at processing corn. The reality was that they didn't have accurate data back then so it was hard to make an legit model. Nonetheless, I do believe the Soviets were successful using these methods in many industries.

Quote:
Originally Posted by James Bond 007 View Post
Funny thing is, I'm sure many big businesses probably do that sort of thing already these days.
They definitely do, and many of these firms are actually more wealthy than entire countries.

I recently heard that some firms now buy satellite data and use AI to scan the earth for the height of coal piles and other commodities. Basically they use pictures - to make an algorithm - to price commodities lol. Nowadays there are quants using game theory, and probably lots of theories I've never even heard of to make impossibly complex models. I just wonder what would happen if these people decided to run a command economy.
Reply With Quote Quick reply to this message
 
Old 08-17-2018, 09:32 PM
 
Location: The Berk in Denver, CO USA
13,115 posts, read 18,723,211 times
Reputation: 20425
No.
Reply With Quote Quick reply to this message
 
Old 08-17-2018, 10:33 PM
 
Location: Kansas City, MISSOURI
5,527 posts, read 1,659,914 times
Reputation: 4741
Quote:
Originally Posted by Drewjdeg View Post
In the West we're taught that these problems actually have non-linear solutions. Which makes sense because of things like diminishing returns; also, it makes it way more complicated. So I'm not exactly sure how relevant this is in 2018, but this how I understand it.

The idea was that you could formulate the most efficient way to produce a good and deliver it to consumers. So if you wanted to plant corn, you could use data on climate, soil quality, etc. and pinpoint the best farms to plant corn - given that there are constraints on how much a farm can grow. The idea being that math was more accurate than the whatever a farmer decides to produce. If corn was an input, they could find the factories most efficient at processing corn. The reality was that they didn't have accurate data back then so it was hard to make an legit model. Nonetheless, I do believe the Soviets were successful using these methods in many industries.
I'm not even talking about the most efficient way to produce things (which are surely computerized all the time now), I'm talking about deciding how much to produce, and when to produce it. It seems to me if you're going to devise a system to replace natural supply and demand, you'd need to calculate "artificial" supply and demand. How would you do that? You'd need some sort of input to your computer system that would calculate when X number of cars would be needed, for example. How would you determine when those cars would be needed, and how many X is? The processes for production are relatively trivial compared to figuring out how much, when and what kind. Not sure that could really be "calculated" even though it's the central issue in any command economy.
Reply With Quote Quick reply to this message
 
Old 08-17-2018, 11:16 PM
 
Location: Reno, NV
459 posts, read 142,885 times
Reputation: 1561
Whatever computers the government uses to calculate what to produce and how much and of it, consumers and business can use computers with the same power. Whether you use a super-fast computer or pen & paper, the difference between the two always boils down to the fact that only consumers know what they want, sometimes with only a 5 second decision making time. No government can possibly predict that.


So the answer is no, communism will not work with supercomputers. It was a dumb idea then and it's a dumb idea now.
Reply With Quote Quick reply to this message
 
Old 08-18-2018, 12:09 AM
 
2,027 posts, read 816,055 times
Reputation: 1788
Quote:
Originally Posted by James Bond 007 View Post
I'm not even talking about the most efficient way to produce things (which are surely computerized all the time now), I'm talking about deciding how much to produce, and when to produce it. It seems to me if you're going to devise a system to replace natural supply and demand, you'd need to calculate "artificial" supply and demand. How would you do that? You'd need some sort of input to your computer system that would calculate when X number of cars would be needed, for example. How would you determine when those cars would be needed, and how many X is? The processes for production are relatively trivial compared to figuring out how much, when and what kind. Not sure that could really be "calculated" even though it's the central issue in any command economy.
Yeah, I guess you would have to tweak the Soviet model a bit. Viktor Glushkov's utopian vision was to democratize the planning apparatus of the Soviet union. Instead of 100 bureaucrats deciding what consumer goods to produce, local factories would produce goods by responding to the needs of the local population. The factories would then request X inputs to make X number of unique goods. All of it would then be optimized.

In a way it does start to resemble a capitalism, which is why I suppose Glushkov's system was never implemented.
Reply With Quote Quick reply to this message
 
Old 08-18-2018, 12:21 AM
 
2,027 posts, read 816,055 times
Reputation: 1788
Quote:
Originally Posted by Alonso Gil View Post
Whatever computers the government uses to calculate what to produce and how much and of it, consumers and business can use computers with the same power. Whether you use a super-fast computer or pen & paper, the difference between the two always boils down to the fact that only consumers know what they want, sometimes with only a 5 second decision making time. No government can possibly predict that.


So the answer is no, communism will not work with supercomputers. It was a dumb idea then and it's a dumb idea now.
You're right, but it was fairly interesting for me to think about for an evening.
Reply With Quote Quick reply to this message
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.

Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.


Reply
Please update this thread with any new information or opinions. This open thread is still read by thousands of people, so we encourage all additional points of view.

Quick Reply
Message:

Over $104,000 in prizes was already given out to active posters on our forum and additional giveaways are planned!

Go Back   City-Data Forum > General Forums > Economics
Follow City-Data.com founder on our Forum or

All times are GMT -6.

2005-2018, Advameg, Inc.

City-Data.com - Archive 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35 - Top