money and PCT

[Frank_Lenk]

[From Frank Lenk (2009.09.21.16:57)]

I apologize for forgetting to time stamp my previous post, as
per convention on this list.

Frank

···

From: Control Systems
Group Network (CSGnet) [mailto:CSGNET@LISTSERV.ILLINOIS.EDU] On Behalf Of Richard
Marken
Sent: Monday, September 21, 2009 10:59 AM
To: CSGNET@LISTSERV.ILLINOIS.EDU
Subject: Re: [CSGNET] money and PCT

[From Rick Marken
(2009.09.21.0900)]

Frank Lenk (2009.09.20.22:55)

Until I can break free to respond with my own thoughts, I wanted to pass
along an old article (1913, I think) that according to my macroeconomics
professor is one of the best articles ever written on money.

I like it! It seems completely consistent with my (naively derived) view of
money as a symbol representing the exchange value of goods and services.
It certainly is consistent with my idea that money need have no intrinsic
“value” at all; paper and computer bits work as well as (or better
than) gold. I was focusing mainly on the “price” aspect of money; the
article makes the important point (which will be very important for modeling, I
think) that the money itself is a claim on goods; an IOU or “credit”.
So $100 is a claim on $100 worth of the goods and services produced by the
economy. The market has put a price tag on all goods and services (that are
available for exchange) so my $100 means that I can scoop up goods and services
with a total price of no more than $100. And when I hand my money to the people
from whom I purchase the goods and services, they are now getting IOUs that
they can use to exchange for the goods and services. So money is a symbol for
goods that have not yet been consumed. But that creates a problem; when I spend
my money (consume products) the money doesn’t get consumed, too; it goes to the
person who provided me with the goods/services. That person now has money representing
goods/services that no longer exist; this money can be used to get new goods
and services; but how do we know they are there? This is a problem I never
considered. Yikes. I must be making some kind of elementary mistake here. What
gives?

Best

Rick

–
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

[Bill_Powers1]

[From Bill Powers (2009.09.22.0943 MDTR)]

Innes says that to get the
credit, we first have sell something – our labor, or our property,
something. So, in general, we have to produce something first to get
money. So insufficient production is not typically the problem. The
problem arises when you decide not to spend your money, but hang on to
it, say because of uncertainty about the future. Then there is
production “out there” but insufficient money to buy it all back, which
can then cause difficulties for you the next time you want to sell your
production – no one may want it if what is already produced is not being
bought.

Right on. There is also a problem my dad called “leakage,”
which arises partly from people just hanging onto their money and partly
from other causes, and is represented by the fact that total spending on
goods and services has less by about 7% than total gross income for quite
a long time. And this is no statistal average over scatter plots: here is
the graph for the 30 years between 1951 and 1981:

The little smudges very near the lower of the two slanted lines are the
data points. (Powers, T.C. Leakage: The Bleeding of the American
Economy. Benchmark Publications Inc., 1996)

Getting the money started is something of a problem; how do you get paid
for work or sales of possessions if nobody else has any credit to start
with? I suppose that just has to be done on trust. It’s easier just to
assume the pipelines are full at the beginning.

I’ve started thinking of credit/money as part of the economic machinery;
it’s the gasoline without which car motors can’t run. You have to have as
much gasoline available as the motors need at their current speed. If you
have less, the motors slow down; if you have more, the excess is unused,
or causes strange changes in the system like a reduction in the energy
content of each gallon.

This puts a different light on the purely financial aspects of the
market, from interest rates to short-sale gambling. People think they can
simply siphon off some of the gas in other people’s tanks for their own
use, or drill a hole in the hose to collect some of the gas on its way to
the tanks. That’s easier than working (though it probably ends up using
just as much time and attention to manage it).

Electricity might be a better metaphor, since the electrons can be used
over and over.

Best,

Bill P.

···

At 04:40 PM 9/21/2009 -0500, Frank Lenk wrote:

[Richard_Kennaway]

[From Richard Kennaway (2009.09.22.1752 BST)]

[From Bill Powers (2009.09.22.0943 MDTR)]
Getting the money started is something of a problem; how do you get paid for work or sales of possessions if nobody else has any credit to start with? I suppose that just has to be done on trust. It's easier just to assume the pipelines are full at the beginning.

By Innes' account, trust is how it got started. Your initial credit is your personal character, abilities, and physical resources that persuade the people you want things from that they will be paid. According to Innes, trust is how it went on, with the greater part of trade in mediaeval times being carried on by means of records of credit rather than coins.

A fixed currency for denominating all credit and debt nationwide simplifies the problems of trust and coordination -- anyone's dollar is exactly as good as anyone else's, and everyone will accept them in payment for anything -- but according to Innes, the networks of credit transactions and clearing fairs worked in their time.

···

--
Richard Kennaway, jrk@cmp.uea.ac.uk, Richard Kennaway
School of Computing Sciences,
University of East Anglia, Norwich NR4 7TJ, U.K.

[rsmarken]

[From Rick Marken (2009.09.22.1140)]

Innes says that to get the credit, we first have sell something � our labor,
or our property, something. So, in general, we have to produce something
first to get money. So insufficient production is not typically the problem.
The problem arises when you decide not to spend your money, but hang on to
it, say because of uncertainty about the future.

In order to understand what's going on here with money I decided to
write a little simulation of an economy. It's coming along pretty
well. Basically it consists of a set of 10 controllers; each
controller specializes in producing one kind of product and has a
reference for consuming some amount of the 10 products produced by the
controllers (including themselves). The simulation starts of with
each controller having a stock of money and product to sell. Each
product has a preselected unit price. On each iteration of the
simulation a controller is selected; this controller buys his
reference amount of each product from all producers (controllers)
including himself, if he can afford it (has enough money in stock).
I've been starting the controllers off with enough money and product
so that all controllers have been able to control successfully (meet
their reference for all products without running out of money). Of
course, the controllers also make money when they sell their product
to another controller.

There is still quite a ways to go with this simulation but it looks to
me like a promising start and I've already learned some interesting
stuff from it -- stuff that you guys apparently already knew but I'm
just not smart enough to do this in my head, I guess. The main thing I
learned is that the total amount of money in the system always stays
the same, it is just redistributed across the controllers. This is
true even if each controllers sells all their stuff (new production=0)
or if production increases. So you are right; production has nothing
to do with money supply.

Right now my controllers (economic agents) are pretty simple; all they
do on each iteration is buy what they need (have references for) if
they can afford it; if they can't, then they don't buy it and they
have an error for that product. So the controllers control their
budget in a very simple way, all or none. Also, there is no control of
production; the controllers should vary production to maintain
inventory at some small positive value. So I will develop more
interesting (and plausible) strategies for the controllers. I'll
distribute the model as soon as it's presentable.

Best

Rick

···

On Mon, Sep 21, 2009 at 2:40 PM, Frank Lenk <FLENK@marc.org> wrote:
--
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

[Bill_Powers1]

[From Bill Powers (2009.09.23.0838 MDT)]

Richard Kennaway (2009.09.22.1752 BST) --

According to Innes, trust is how it went on, with the greater part of trade in mediaeval times being carried on by means of records of credit rather than coins.

Yes, I loved that part. He was debunking (with real data) the old imaginary picture of how it all went from barter, to some "medium of exchange", to money tokens, and finally to abstract credit. He made a pretty good case for the initial transactions being based on abstract credit! All the rest was just trying to keep track of who had credit for what and who owed whom how much.

A fixed currency for denominating all credit and debt nationwide simplifies the problems of trust and coordination -- anyone's dollar is exactly as good as anyone else's, and everyone will accept them in payment for anything -- but according to Innes, the networks of credit transactions and clearing fairs worked in their time.

His explanations go a long way toward making a real model of the economy feasible. I'm glad Kenny Kitzke gave us a prod on this. I don't think it's generally understood how modeling is used in engineering and physical sciences. What we can observe about systems like economies is only some set of relationships seen through a keyhole, a few variables at a time. We can say that if I pay you money, my bank account decreases and yours increases by the same amount, and if you simultaneously buy things from others, or from me, your bank account shrinks. If a bank loans money, the borrower has money and the bank has a debit as an asset, and the bank also collects interest and gets back part of which it lent, and where did that interest money come from after the loan is all paid off? And if people prefer more of one good than another, how does one person's debt get translated into credit usable for other goods, and so on and on. Each little relationship can be described as part of a model, but nobody can hold enough of this web of relationships in mind to make a prediction of what the whole system will actually do. The only possible way to see the consequences of all these little local relationships is to set up a huge array of simultaneous equations and solve them, or if that's not possible (the solving isn't), set up an analog computation in the form of a simulation, a model, which will act out the consequences of all these interlocking relationships. Then you see what you simply can't see in any other way.

This is certainly true of the PCT model. See chapter 3 in LCS3, the live block diagram. Just try to reason out what will happen if you double the gain in the system's output function. In fact you can't even predict the simple control behavior of the system correctly just by following cause and effect step by step around the loop. Yet that's how economists talk about the way economies work. They try to do it with sequential logic when it's really a system of interactions that are all occurring simultaneously. You can't predict the effect of a price change unless you can predict the amount and direction of the change, but to predict that you have to know what the effect of the price change is -- there are closed loops all over the place. People not used to dealing with circular causation simply don't know how to handle closed loops.

That's why we need an economic model, and why Innes' article is required reading that apparently never got done. Or more likely, the economists of 1903 and later were already too fond of their tale of the saga running from barter to credit to listen to anything that contradicted the story. PCTers are quite familar with that syndrome.

Best,

Bill P.

···

--
Richard Kennaway, jrk@cmp.uea.ac.uk, Richard Kennaway
School of Computing Sciences,
University of East Anglia, Norwich NR4 7TJ, U.K.

[Bill_Powers1]

[From Bill Powers (2009.09.23.09078 MDT)]

Rick Marken (2009.09.22.1140) --

In order to understand what's going on here with money I decided to

write a little simulation of an economy. It's coming along pretty
well. Basically it consists of a set of 10 controllers; each
controller specializes in producing one kind of product and has a
reference for consuming some amount of the 10 products produced by the
controllers (including themselves). The simulation starts of with
each controller having a stock of money and product to sell. Each
product has a preselected unit price. On each iteration of the
simulation a controller is selected; this controller buys his
reference amount of each product from all producers (controllers)
including himself, if he can afford it (has enough money in stock).
I've been starting the controllers off with enough money and product
so that all controllers have been able to control successfully (meet
their reference for all products without running out of money). Of
course, the controllers also make money when they sell their product
to another controller.

Way to go. This is the right way to start. Are you distinguishing between flow variables and rate variables? The integrations represented by inventories and cash reserves allow you to make all the rest of the control loops into simple proportional systems that are automatically stable (by themselves).

There is still quite a ways to go with this simulation but it looks to
me like a promising start and I've already learned some interesting
stuff from it -- stuff that you guys apparently already knew but I'm
just not smart enough to do this in my head, I guess. The main thing I
learned is that the total amount of money in the system always stays
the same, it is just redistributed across the controllers. This is
true even if each controllers sells all their stuff (new production=0)
or if production increases. So you are right; production has nothing
to do with money supply.

Right, but obviously nothing can happen unless there's at least the minimum required amount of money in the system. You can let the cash reserves go negative to simulate borrowing, so actually you could start with zero money, I guess. Interesting to try, when you get that far. If you included work and wages, it would be necessary to have enough money to pay enough for work on the production line so the product could be purchased -- an arrangement you're familiar with. Then you start adding capital income -- profits, dividends, pensions -- that some people live on and see what kinds of prices have to be set to make the circular flow come out right....

Kenny K., neither Rick or I has a model yet, but it's slowly taking shape.

Best,

Bill P.

···

Right now my controllers (economic agents) are pretty simple; all they
do on each iteration is buy what they need (have references for) if
they can afford it; if they can't, then they don't buy it and they
have an error for that product. So the controllers control their
budget in a very simple way, all or none. Also, there is no control of
production; the controllers should vary production to maintain
inventory at some small positive value. So I will develop more
interesting (and plausible) strategies for the controllers. I'll
distribute the model as soon as it's presentable.

Best

Rick
--
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

[rsmarken]

[From Rick Marken (2009.09.23.1215)]

Bill Powers (2009.09.23.09078 MDT)

Rick Marken (2009.09.22.1140) --

In order to understand what's going on here with money I decided to
write a little simulation of an economy.

Way to go. This is the right way to start. Are you distinguishing between
flow variables and rate variables? The integrations represented by
inventories and cash reserves allow you to make all the rest of the control
loops into simple proportional systems that are automatically stable (by
themselves).

Yes, got my stocks and flows in place. The control systems are very
simple at this point.

There is still quite a ways to go with this simulation but it looks to
me like a promising start and I've already learned some interesting
stuff from it -

Right, but obviously nothing can happen unless there's at least the minimum
required amount of money in the system. You can let the cash reserves go
negative to simulate borrowing, so actually you could start with zero money,
I guess.

TBD. Right now if you don't have the money you don't get the stuff you want.

Interesting to try, when you get that far. If you included work and
wages, it would be necessary to have enough money to pay enough for work on
the production line so the product could be purchased -- an arrangement
you're familiar with.

Right now I don't separate management from labor. I suppose I will
have to eventually. But right now I am just looking at how specialized
control systems can control for what they want using money.

Then you start adding capital income -- profits,
dividends, pensions -- that some people live on and see what kinds of prices
have to be set to make the circular flow come out right....

It's a long way to that point (maybe). Right now I've just got an
architecture that I like. I will soon distribute it see if people
think it's on the right track.

Best

Rick

···

--
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com