Economic Dynamics

This forum contains all archives from the SD Mailing list (go to http://www.systemdynamics.org/forum/ for more information). This is here as a read-only resource, please post any SD related questions to the SD Discussion forum.
Locked
Mabel Fong
Junior Member
Posts: 2
Joined: Fri Mar 29, 2002 3:39 am

Economic Dynamics

Post by Mabel Fong »

Hello again:

What a nice welcome! I am feeling so appreciated that
I shall indeed continue to solicit your expertise.

First, to answer your questions: My official
introduction to SD began at this URL:
http://www.albany.edu/cpr/sds/DL-IntroSysDyn/index.html;
and I found nothing here at odds with what I learned
from the Electrical Engineers. I would rather not be
traced to my alma mater because of a recent political
dust-up: I have been able to convince a few younger
professors of Economics hereabouts (apparently to
their peril) that Economic Science has yet to consider
its proper paradigm; and public advertisements of our
debates might damage the career prospects of some
people I like.

Based on all my authority as a Bachelorette of
Science, I am convinced that nothing short full-on,
Newtonian dynamics can possibly support anything that
might reasonably be called a science of Economics. I
say this on behalf of a perfectly serviceable, control
theory approach that accomplishes things only
indicated by general equilibrium theory, supply/demand
analysis, game theory, econometrics, Keynesian
Macro-analysis, et al.

The model I am working with computes the quantity Q of
Good J held by Sector I in Economy K at time T for all
Qijk|@t. It seems to me that this model has analogies
to SD’s logistical models regarding what happens when
a market stock becomes completely depleted. The
economic model has no need to maintain a backlog
because (being an economic model) shortages or
excesses of goods are adequately expressed in a
commodity’s value. But the model does need to
conserve mass: it cannot use anything to produce
another good until the input itself has been produced.
This means that an allocation regime must kick-in
whenever output exceeds demand while nothing is on the
market.

I believe you all had a lengthy discussion a few weeks
back about how such a programming change might be
triggered, while avoiding formal references to
rates-of-change as if they were state variables. I
would be grateful for your critique of the economic
model’s method for addressing this problem. They
model goods-on-the-market as the integrated difference
between output rates and the rates at which expressed
demands are fulfilled. Their trigger for the
allocation regime is simply a negative value in the
market state variable itself. Markets are allowed to
go a tad negative, where “a tad” means “as far
negative as you can get, beginning from a positive
value, during one dt”.

This simple strategy is okay with me because, if the
dt is vanishing small from the standpoint of our
analysis, then so is “the tad”. I am not disturbed by
the possibility of small, one-sided violations of
mass-conservation adding up to instability because
“the tad” has to be made up by an excess of supply
over demand before the market state can achieve
positive values that turn off the allocation regime.
Seeing no violation of physical principles or modeling
principles, and noting that the overall model performs
smoothly and with stability, I am satisfied. Should I
be?

For a second question, please consider the common
Economic premise of a perfect market wherein prices
can be anticipated such that markets remain perfectly
clear, and disequilibria are expressed only in that
different economic agents value the same good
differently. Is such a formulation realizable within
the context of a truly dynamic analysis?

Vty,

Mabel Fong
may_belle_66@yahoo.com
zenabraham@aol.com
Junior Member
Posts: 18
Joined: Fri Mar 29, 2002 3:39 am

Economic Dynamics

Post by zenabraham@aol.com »

Hi May,

What I have most enjoyed about SD may be summed up in this joke I was sent a
while ago:

An economist was drawing a model describing forecasts of growth in milk production.
The person referred to rates of change, and population of consumers, etc.

But the SD person in the back of the room raised her hand and asked this
question:

"Where are the cows?"

The question I have for you is this: can you provide real world examples that
match your model? Also, in this new world of business collapse, you seem to
be assuming that the knowledge about the market is not only perfect, but
true. In other words, where do you consider that someone is simply lying
about whats going on for personal gain?

If you are going to use SD, I really do hope it causes you to become an
Institutional Economist. After all you cant have economics without people
and people build institutions.

Zennie Abraham
From: zenabraham@aol.com
=?iso-8859-1?Q?Andr=E9_Reichel?=
Junior Member
Posts: 14
Joined: Fri Mar 29, 2002 3:39 am

Economic Dynamics

Post by =?iso-8859-1?Q?Andr=E9_Reichel?= »

Mabel wrote:

| a market stock becomes completely depleted. The
| economic model has no need to maintain a backlog
| because (being an economic model) shortages or
| excesses of goods are adequately expressed in a
| commoditys value. But the model does need to

If there is no backlog of "waiting" or "unserved" demand, this demand will
just disappear and have no effect on the models reactions. Thats why...

| This means that an allocation regime must kick-in

If there is a backlog of "waiting" demand, this backlog is going to be the
"allocation regime". No need for interventions of any kind. This would also
relieve you from letting a state variable become negative, which always is
somewhat awkward: negative height (impossible), negative bank accounts (=
positive debts), negative welfare ("badfare"?) etc.

| This simple strategy is okay with me because, if the
| dt is vanishing small from the standpoint of our
| analysis, then so is "the tad". I am not disturbed by
| the possibility of small, one-sided violations of
| mass-conservation adding up to instability because
| "the tad" has to be made up by an excess of supply
| over demand before the market state can achieve
| positive values that turn off the allocation regime.

Your model, being a SD-model, will react on a negative state variable and
therefore produce a different outcome than without it (maybe less "real").
Of what kind is the decision policy in the attached rate variables? Is it in
the form of "if-then-else"? That seems to be the case, given your quote. I
would highly recommend to avoid using descrete decision rules in a
continuous simulation model. With a positive stock, e.g. a backlog thats
filling up with "waiting" demand, this problem would solve itself.

| Seeing no violation of physical principles or modeling
| principles, and noting that the overall model performs
| smoothly and with stability, I am satisfied. Should I
| be?

First principle of modeling is "clear purpose". The purpose is always
related to a real-world problem of interest -- a problem NOT a system.
Another principle is "no black boxes". No hidden variables, no sneeky little
tricks to produce "good" or "smooth" outcome.
So, what is the purpose of your economic model? What for is the Quantity Q
computed, what is believed to influence this, what is the time horizon,
believed or historic refernce mode, and so on.

| For a second question, please consider the common
| Economic premise of a perfect market wherein prices
| can be anticipated such that markets remain perfectly
| clear, and disequilibria are expressed only in that
| different economic agents value the same good
| differently. Is such a formulation realizable within
| the context of a truly dynamic analysis?

Yes, but only for the sake of realism. The neoclassical hypothesis of
perfect markets was just established to build a mathematically consistent
(and elegant) framework of the classical theory of the "invisible hand" --
and to bash the marxists as one said. In a dynamic economical model you
would have a different purpose: to model real problems and get real (and
hopefully implementable) answers. Like Zen Abraham said: look for the cows
if youre interested in milk!

Viele Grüße / Many regards
André Reichel
Universität Stuttgart, Germany
E-Mail: reichel@sofo.uni-stuttgart.de
Locked