Five Easy Pieces

CSGnet Archive CSG2000
1

[From Rick Marken (2000.03.02.1250)]

1. I think we have about 30 people committed to attending the
CSG 2000 in Boston. It would be nice to get a few more. I
should be receiving the contract from Boston U this month,
at which point I will ask people to _pay_ for the conference;
then we'll see who is really committed. But if you plan to
attend the conference and have not told me yet, please let me
know. If you (or people you know) will be attending the
conference but you (or those people ) are currently not on
speaking (or e-mailing) terms with me, please contact CSG
President Lloyd Klinedinst at lloydk@icon-stl.net and let
him know that you (or they) will be attending. Thanks.

2. Re: John Anderson (2000.03.03.1035 EST) --

Does anyone know how to decode such attachments?

I don't know how myself. I think if people have attachments
they should just paste them into the body of the e-mail as
text. The word processor document takes up a lot of space as
an attachment (more than the ASCII text version) and pasting
the text into e-mail would make these documents more readily
accessible to those who read only the digest.

3. Has anyone else (besides David G.) tried my Economics demo
(http://home.earthlink.net/~rmarken/ControlDemo/Economics.html)?
If so, any comments? (By the way, it works fine in Netscape 4.7
on a PC; the performance of this demo is, unfortunately, quite
Browser, and platform, dependent).

4. My much anticipated paper on "Controlled Variables" is not
likely to appear before Summer, 2001. This is a bit of a
disappointment because it means that behavioral scientists
are not likely to switch to doing the appropriate kind of
research for at least another year.

5. I have been working on implementing TCP's macro economic
model as a dynamic simulation. I have discovered (I think)
that TCP's model _does not_ explain _why_ leakage affects
economic growth _rate_. The model (as I've implemented it)
is a simple control model in which the aggregate producer
produces PQ to meet demand, compensating for leakage by
varying price (P) as necessary to collect from the consumer
the amount spent for production. This model nicely accounts
for the effects of leakage on relative output (Q/Q') and
on the index of automatic inflation (shown in Table 3-6,
p. 191 of _Leakage_). But the effect of leakage on growth
rate (r) does _not_ fall out of the model. It turn out that
the effect of leakage on growth rate is an empirical observation
that is incorporated into the model as an _assumption_. That
is, TCP just assumes (I think it's on p. 100 or so) that
dQ/dt= z - l, where dQ/dt is growth rate, z is intrinsic
growth rate and l is leakage. I think dQ/dt should fall out
of the model operating according to its own properties.
This can only happen if the model includes properties that
allow leakage to (somehow) influence dQ/dt. I'm currently
trying to fiddle with TCP's model so that dQ/dt does vary
with leakage (as it is observed to do) without putting this
effect into the model as an assumoption. If anyone has any
suggestions regarding _how_ leakage might affect growth rate
I would sure appreciate them.

Best

Rick

···

---
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

2

[From Bruce Nevin (2000.03.02.2002 EST)]

Rick Marken (2000.03.02.1250) --

···

At 01:16 PM 03/02/2000 -0800, Richard Marken wrote:

I think if people have attachments
they should just paste them into the body of the e-mail as
text. The word processor document takes up a lot of space as
an attachment (more than the ASCII text version) and pasting
the text into e-mail would make these documents more readily
accessible to those who read only the digest.

Another suggestion: just send the attachment to one person who volunteers
to post it for download on the CSG web site. Then send an abstract in ASCII
in the text of your message, with instructions where to go for the download.

        Bruce

3

[From Bill Powers (2000.03.03.0435 MST)]

Rick Marken (2000.03.02.1250)--

1. I think we have about 30 people committed to attending the
CSG 2000 in Boston. It would be nice to get a few more.

Kirk Sattley will be attending some of the sessions. His wife will not.
He lives in the area so will not be needing a room.

2. Re: John Anderson (2000.03.03.1035 EST) --

Does anyone know how to decode such attachments?

I don't know how myself.

There's a program called Mime64.exe that will do it for a PC. I sent it to
John and he said it works. It's obtainable free on the WWW.

I think if people have attachments
they should just paste them into the body of the e-mail as
text.

That's fine if what you're sending is text.

3. Has anyone else (besides David G.) tried my Economics demo
(http://home.earthlink.net/~rmarken/ControlDemo/Economics.html)?

Yes. It works well, and I like the way the bread and meat proportions are
adjusted along with the total consumption.

I have been working on implementing TCP's macro economic
model as a dynamic simulation. I have discovered (I think)
that TCP's model _does not_ explain _why_ leakage affects
economic growth _rate_. The model (as I've implemented it)
is a simple control model in which the aggregate producer
produces PQ to meet demand, compensating for leakage by
varying price (P) as necessary to collect from the consumer
the amount spent for production.

Does it also show that the composite consumer's income derives totally from
the composite producer?

This model nicely accounts
for the effects of leakage on relative output (Q/Q') and
on the index of automatic inflation (shown in Table 3-6,
p. 191 of _Leakage_). But the effect of leakage on growth
rate (r) does _not_ fall out of the model. It turn out that
the effect of leakage on growth rate is an empirical observation
that is incorporated into the model as an _assumption_.

...

I think dQ/dt should fall out
of the model operating according to its own properties.
This can only happen if the model includes properties that
allow leakage to (somehow) influence dQ/dt.

I totally agree. But I couldn't tell TCP that without risking apoplexy (his).

There's something about what happens when money is lost from the circular
flow, so there's not enough to buy the whole product. The only way to
compensate for that loss is to raise prices (the automatic inflation effect
of leakage), but that means the sum of wages and capital income is no
longer enough to buy the whole product, and inventories start to rise
unless production and consumption are reduced and people are laid off or
their salaries are reduced (below the level of increase that normally
occurs). Maybe what's needed in a model is a control system to keep
inventories exactly at zero (or some low amount). The effect on growth rate
might show up through this indirect link.

Something that needs to be made explicit in the model is savings. TCP
didn't really deal with this, using the justifiable approximation that the
net savings rate must be close to zero (the composite consumer withdraws as
much as is deposited, to a first approximation). But actual savings can
fluctuate over the short term, with a net flow into or out of savings. By
stretching the meaning of savings to include all ways of temporarily
removing money from circulation and then returning it, we can handle rho
leakage.

One of TCP's main theses was that growth is driven by increases in
productivity, and productivity is increased by human ingenuity, not by
money. The "intrinsic growth rate" is supposedly a measure of the rate at
which human ingenuity raises productivity when money is freely available
and the circular flow is free of leakage. The actual growth rate is thus
this intrinsic growth rate minus all the effects that tend to reduce it. If
leakage produces autoinflation and autoinflation leads to a reduction in
production and consumption, the economy can still be growing, but at a
slower rate. The continuing increase in productivity in the background
prevents the effect from running away, if it's not too large.

At least that's the arm-waving version. It would be nice to see it
happening in a model, coming out of the properties of the model instead of
verbal reasoning.

Best,

Bill P.

4

[From Rick Marken (2000.03.03.0900)]

Bill Powers (2000.03.03.0435 MST)--

Kirk Sattley will be attending some of the sessions. His
wife will not.

So we're down to 29 total.

Me:

3. Has anyone else (besides David G.) tried my Economics demo
(http://home.earthlink.net/~rmarken/ControlDemo/Economics.html)?

Bill:

Yes. It works well, and I like the way the bread and meat
proportions are adjusted along with the total consumption.

Great. I like it, too. Thanks.

Me:

I have been working on implementing TCP's macro economic
model as a dynamic simulation...

Bill:

Does it also show that the composite consumer's income derives
totally from the composite producer?

Yes. It basically solves TCP's simultaneous equations for PQ and
P'Q'. I know it's basically correct because it produces TCP's
analytically derived results in Table 3-6 (p.191 of _Leakage_).
that is, given total leakage (first column in Table 3-6) the
model produces the appropriate Q/Q' (col 2) and auto inflation
(col 4). It will also produce the right Rate of growth, but only
if leakage is put directly into the growth formula as an input
to the model. The model basically acts to make the money paid (to
consumers) for the constantly increasing production of goods
match the money received for these goods from the consumers
themselves.

Me:

I think dQ/dt should fall out of the model operating according
to its own properties...

Bill:

I totally agree. But I couldn't tell TCP that without risking
apoplexy (his).

So my modeling efforts have gotten me to the point where you got
years ago by just thinking. That's OK with me; I'm getting used
to it. At least I get there _eventually_.

Thanks for the suggestions about what to add to the model. Very
helpful. I think TCP made a great start at an aggregate model of
the economy but now we are in a position where we can add the
properties that will make the model work so that the effect of
leakage on dQ/dt (a brilliant observation on TCP's part) becomes
a prediction (rather than an assumption) of the model.

I think this model economy is a good example of an "aggregate
level" control model. It is also nice because there are copious
amounts of aggregate level data against which to test the model.
I think I've found something that will keep me off the street
for another couple of years.

By the way, does anyone out there know of any other working
model of the macro economy -- one that actually produces
variations in observable variables like GNP, inflation,
unemployment, etc? I would really like to see a reference to
such a model. When I say "model" I don't mean "regression
equation", by the way. I mean a model of the processes thought
to be involved in the economy that result in the observed
behavior of economic variables.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

5

[Martin Taylor 2000.03.04 11:54]

[From Rick Marken (2000.03.03.0900)]

By the way, does anyone out there know of any other working
model of the macro economy -- one that actually produces
variations in observable variables like GNP, inflation,
unemployment, etc? I would really like to see a reference to
such a model. When I say "model" I don't mean "regression
equation", by the way. I mean a model of the processes thought
to be involved in the economy that result in the observed
behavior of economic variables.

Not a working model in the sense of a simulation, but an analytic
description in terms of processes and mechanisms that could probably
be turned into a working model with current computers--Samuel Bagno,
IRE Conference record 1955 Part 4. I recently got permission from the
IEEE to reproduce this paper on my Web site, and have done so:
<http://www.mmtaylor.net/Economics/Bagno/index.html&gt;

Since I first read it in 1955, this work has seemed to me to have
much the place in economics as Mendel's work has in genetics--a first
start in the direction economics will eventually have to follow.
Unfortunately (to my mind) it has been ignored by economists whose
theories have caused so much destruction in the world. Bagno deduces
that there is a fundamental requirement for moderate inflation in a
stable economy. His analysis also suggests that to balance the budget
in the long term is a recipe for economic decline and devastation, at
least in a closed economy. And if I understand "leakage" correctly,
Bagno describes the process by which it occurs, using the perceptions
and experiences of the individual as the basis. I haven't read TCP,
and what I know about leakage comes entirely from reading CSGnet, so
I may be way off base here, but I suspect that TCP is describing a
phenomenon that occurs, whereas Bagno is predicting that it should
occur.

Martin

6

[Martin Taylor 2000.03.04 14:04]

[Martin Taylor 2000.03.04 11:54]

I said...

... Samuel Bagno,
IRE Conference record 1955 Part 4. I recently got permission from the
IEEE to reproduce this paper on my Web site, and have done so:
<http://www.mmtaylor.net/Economics/Bagno/index.html&gt;

It was pointed out to me that there was a case error in the links.
That's been fixed and you should be able to see the paper now, rather
than getting a 'missing image' icon.

Sorry about that.

Martin