PCT, Economics & Ludvig von Mises

CSGnet Archive CSG2000
1

I've just started reading Human Action by Ludvig von Mises. Subtitled "A
Treatise on Economics" it is touted as a monumental and magnificent work
and the only full-blown theory of economics that proceeds from a basis in
individual human action.

I've already come across some really interesting lines. Take this one for
example:

        "We call contentment or satisfaction that state of a human being which
does not and cannot result in any action." (p.13)

Sounds like a good name for the zero-error situation to me. Later on:

        "The ultimate goal of human action is always the satisfaction of the
acting man's desire." (p.14)

Substituting wants for desires (a not uncommon substitution) and that
sounds to me like actions controlling perceptions in relation to reference
conditions.

Before everyone goes off, I'm not suggesting that von Mises anticipated or
predates Bill Powers' work in any way. I picked up von Mises' book because
of its attempt to anchor economics in individual action instead of
mysterious "invisible hands" or some inferred collective consciousness. I
also think it's important that PCT can establish links to other great
thinkers and writings beyond Bill's often cited links to engineering.

As I make my way through what is a very BIG book (906 pages), I'll let you
know if von Mises made any more PCT-like observations.

···

--

Fred Nickols
The Distance Consulting Company
"Assistance at A Distance"
http://home.att.net/~nickols/distance.htm
nickols@worldnet.att.net
(609) 490-0095

2

[Mike Acree (2000.03.06 1017 PST)]

Fred Nickols (2000.03.05)--

It has always seemed to me that the methodological individualism of Mises'
microeconomic Austrian school made a more natural and promising fit to PCT
than the macroeconomic approach of both traditional theory and TCP's leakage
theory. The restriction to the analysis of aggregates would seem to present
many of the same problems in economics as in psychology. But I say that
without having read Mises. I did buy a copy of _Human Action_ a few years
ago, and the prospect of discussing it with someone friendly to PCT will
help push it to the top of my stack (a hefty push, as you say). I note that
Mises has at least this very obliquely PCT-relevant endorsement: Robert
Heilbroner, who contributed an encouraging blurb for _Leakage_, wrote in a
famous _New Yorker_ article a few years ago that Mises, ridiculed during his
lifetime for his claim 80 years ago that economic calculation was impossible
under socialism, is now generally acknowledged to have been correct on this
point.

Mike

3

[From Rick Marken (2000.03.06.1340 PST)]

Mike Acree (2000.03.06 1017 PST)

It has always seemed to me that the methodological individualism of Mises'
microeconomic Austrian school made a more natural and promising fit to PCT
than the macroeconomic approach of both traditional theory and TCP's leakage
theory. The restriction to the analysis of aggregates would seem to present
many of the same problems in economics as in psychology.

The problem with the analysis of aggregates in psychology is that
the aggregate data ia being used to test models of individuals. This
is not a problem for TCP's model, where aggregate data is used to
test a model of the aggregate.

TCP's leakage theory of the micro economy has a natural fit with PCT
because it is a closed loop control model of the macroeconomy (and
also, of course, because TCP is an anagram of PCT;-)). As in WTP's
PCT model, TCP's model recognizes that all variables in a closed
loop are changing _simultaneously_. This means that the aggregate
economy is simultaneously producing (acting on) and consuming
(perceiving) stuff (GNP), which is the controlled variable in the
loop (this is the way it works in my implementation of the TCP model).

TCP implicitly assumed that the reference for GNP is constantly
increasing, and he assumed that leakage reduces the rate of change
of that reference (and, hence, the rate of change of GNP). It's
this assumption that I questioned (it puts the effect of leakage
into the model as a premise rather than as a conclusion). But I
have now discovered (NB. Bill and Alice!!), thanks to actually
implementing the model as a computer program (in Excel), that the
effect of leakage on economic growth (dGNP/dt) _does_ fall out
of the TCP model _without_ putting it in as an assumption! But
it's not leakage (l) that affects economic growth; it the rate of
_change_ in leakage over time (dl/dt) that affects economic
growth!!!

The effect of dl/dt on economic growth is not _assumed to occur_; it
is a _side effect_ of the operation of TCP's GNP control model. This
is a very cool discovery (I think). It makes the effect if leakage
on economic growth a _prediction_ rather than an _assumption_ of
TCP's model of the economy. My preliminary look at the historical
data shows that the observed relationship between dl/dt and dGNP/dt
is just as strong as that between l and dGNP/dt. So the theory
may even be _right_!!

Anyway, it seems to me that the TCP model of the aggregate
economy is consistent with PCT because it _is_ a control model.
TCP's model is based on the assumption that an aggregate of
input controllers can be modeled as an input control system. The
fact that TCP never, apparently, saw the correspondence between
his model of the aggregate economy and his son's model of the
individuals who make up that economy has to be one of the great
scientific human interest stories of all time!

Best

Rick

···

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

4

QUERY:
[From Erling Jorgensen (2000.03.08.0800 CST)]

Rick Marken (2000.03.06.1340 PST)

Rick, please think of your reply as "TCP for Dummies," and
fashion accordingly. I'd appreciate it.

Is there a one-sentence reminder of what "leakage" entails? I seem
to recall it's stuff like savings, acquisitions, and inheritance money,
etc. that are (somehow) kept out of the recycling flow of the economy.
What determines whether they are classed as "leakage" or as "savings"
or as "speculations" or as something else?

TCP's model recognizes that all variables in a closed
loop are changing _simultaneously_. This means that the aggregate
economy is simultaneously producing (acting on) and consuming
(perceiving) stuff (GNP), which is the controlled variable in the
loop (this is the way it works in my implementation of the TCP model).

Second question: Where does leakage appear in your (implied)
diagram? Three possibilities occur to me -- can you confirm/
disconfirm each one [and then give me the real answer 8-}]?

1) It could be implemented as a negative disturbance ("a drain on
GNP stuff"). This would seem to make it an assumption of the model
rather than a prediciton, (as you rightly note, although at a different
place in the diagram, I think):

he assumed that leakage reduces the rate of change
of that reference (and, hence, the rate of change of GNP). It's
this assumption that I questioned (it puts the effect of leakage
into the model as a premise rather than as a conclusion).

2) It could be implemented as a siphoning reduction on aggregate
consumption, before it is compared to its reference -- the input side
of the diagram (i.e., "stuff that is consumed and _kept_").

3) It could be implemented as a reduction of aggregate production
-- the output side of the diagram (i.e., "a loss of productive capacity").
I think you are hinting at this possibility, with the following:

The effect of dl/dt on economic growth is not _assumed to occur_; it
is a _side effect_ of the operation of TCP's GNP control model.

Which if any of these is correct, as you are implementing the model?
Thanks for the clarifications.

All the best from dummie me,
        Erling

5

[From Rick Marken (2000.03.09.0900)]

Erling Jorgensen (2000.03.08.0800 CST)--

Is there a one-sentence reminder of what "leakage" entails?

Leakage is just the fraction of GNP that is not returned to the
composite producer as payment for goods and services. According
to TCP, the quantitative amount of leakage each year is reported
in the Statistical Index as "Personal Saving" and "Undistributed
Corporate Profits".

What determines whether they are classed as "leakage" or as
"savings" or as "speculations" or as something else?

Leakage is, by definition, dollars received by consumers (as
wages or profit) that are _not_ spent on goods and services.
TCP argues that "Personal Saving" and "Undistributed Corporate
Profits" are dollars distributed to consumers (as wages and
corporate profits, respectively) that are unspent. That is,
when wage and profit income is added up, "Personal Saving"
and "Undistributed Corporate Profits" are the amounts by
which wage and profit income exceed GNP (which is also a
measure of the dollar value of goods and services purchased).
This is how these two values are computed for the Statistical
Index; "Personal Savings", for example, is not a measure of
how much money people put into their savings accounts in a
particular year; it's just the excess of wages received over
goods and services purchased from wage income.

Here's an example of how "Personal Saving" and "Undistributed
Corporate Profits" are computed. If wage income for the year
is 100 billion and profit income is 200 billion and GNP is
250 billion, then "Personal Saving" is 1/3 of the difference
(50 billion) between income received and goods/services
purchased, or about 17 billion; "Undistributed Corporate
Profits" are 2/3 of this difference or about 33 billion.

Second question: Where does leakage appear in your (implied)
diagram?

Leakage appears in my model as a disturbance to producer income,
which I call P'Q', where P' is the selling price of the goods
and services, Q', that are being produced. It costs the composite
producer (in wages and profits) P to produce Q'; so PQ' is the
cost of production. The model is trying to keep P'Q' = PQ'; that
is, the composite producer must get paid back for the costs of
production. If there were no leakage, P would equal P' and
everything would be fine. Due to leakage, however, the composite
consumer doesn't have enough dollars to pay itself back (as the
composite producer). That is, leakage (when it's negative) tends
to make producer income (P'Q') less than producer expenses (PQ').

To make up for this leakage, the composite producer has to mark
up the price of goods; P'>P. This is called autoinflation. This
adjustment in P' (selling price markup) is the composite producer's
means of keeping it's input (income:P'Q') matching it's expenses
(reference for income:PQ'). In the process, Q' (with leakage)
ends up being less than Q'sans leakage; the composite producer
is producing less than it is capable of producing, by a factor
proportional to the current total level of leakage. Growth rate,
which is dQ'/dt (without the P so it is "inflation corrected", as
is the case when real GNP growth is reported) happens to be
inversely proportional the rate of change in leakage (dl/dt).

Does this help at all? It may be helpful to see the model in
operation. I am happy to distribute the model (it's an Excel
spreadsheet in fairly rough form) to anyone who wants it. But
I think the best thing to do first in order to understand this
stuff is to buy and read "Leakage" (order from Benchmark:
http://www.benchpress.com/Leakage1.htm ). Chapter 1 gives some
of the basic definitions and observations and Chapter 2 describes
TCP's model of the economy.

Best

Rick

···

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