[From Rick Marken (2004.01.20.2120)]
Bill Williams:
Rick Marken (2004.01.20.1130)-
[Bill Williams] is correct if the income he is talking about is the
income for the aggregate consumer (Ic) and the expenditure he is
talking about is the expenditure for the aggregate producer (Ep). It
is not true if the income he is talking about is the income for the
aggregate consumer (Ic) and the expenditure is the expenditure by the
aggregate consumer (Ec).Strange as it may seem, _in the aggregate, Ic is equal to Ec. Rick,
you have yet to comprehend the logic of the agregate context. In
the aggregate, "Income is expenditure." And, conversely, "Expenditure
is Income."
I agree that I don't understand why, in the aggregate, consumer income
equals consumer expenditure. Why is it necessarily true that the
aggregate consumer spends all that it gets as income? I believe it
could happen. I just don't see how it could be a law or tautology, as
you imply. If GDP is consumer income (and producer expenditure) then,
it seems to me, if all consumer income were also consumer expenditure
then there would be no inventory since all that was produced would have
been bought.
Bill Williams (20 January 2004 7:00 PM CST)--
I should have been more explicit. It would have been clearer if I had
said "Money Income is equal to money expenditure."
No. That would still not be clear to me. What I don't understand is why
the aggregate consumer must be assumed to spend all its income. What
the aggregate consumer gets as income is GDP dollars. I agree that some
consumers (the poorer ones) will spend all their portion of this
income. So for these elements of the aggregate consumer income will,
indeed, equal expenditure. But other consumers (the much richer ones)
will not spend the entire amount of their portion of this income. At
least, that income will not be used to purchase GDP goods and services
in proportion to that income. I think the fact that all the income of
the aggregate consumer is not spent on goods and services is
demonstrated by the fact that there is always an inventory of goods and
services on which consumer income has not been expended. But maybe I
have this wrong. It would be great if you could set me straight.
Now Rick says, that my assertion that "Income is expenditure" does
not apply to the following situation-- as Rick says,>It is not true [that income equals expenditure] if the income
[Williams]
> is talking about is the income for the aggregate consumer (Ic) and
the
> expenditure is the expenditure by the aggregate consumer (Ec).
However, as I pointed out in Boston a few years ago, in my attack on
Rick's presentation
Well, it's nice to hear such breathtaking honesty. But it would be
nice to get help instead of attacks.
Consider the example of the aggregate circular flow diagram in
economics. .________<____(2)_____
> >
C P
> >
-----(1)---------->---------------To start with say the rate of income is 100 units. In a simultaneous
analysis the income is 100 units everywhere in the closed, circular
flow loop.
I don't think this is quite right. In a simultaneous analysis, all
variables are changing simultaneously but they don't necessarily all
have the same value simultaneously. In a simultaneous analysis of a
closed loop system, like the one you show, the effects of one variable
on another propagate around the loop over time. The slowing factors and
transport lag determine how rapidly this propagation takes place. In
the loop you show, a simultaneous analysis takes into account the fact
that P is having an effect on C _while_ C is having an effect on P. But
the values of P and C at any particular instant are not necessarily
(and almost certainly are _not_) the same.
Suppose the consumer attempts to save by reducing expenditures
relative to income. What will be the effect? An individual consumer
can, of course, choose to reduce expenditures out of a, more or less,
constant income. But, this is not true in the situation under
consideration here. The analysis here is, as has been agreed by all
parties, a simultaneous aggregate analysis. Consistent with this
specification, there is in the circuit above no loop or time delay.
If an expenditure is inserted into the loop at point one it will
appear without any delay what-so -ever at two. It may be helpful to
think of the circular flow in terms of a perfect circle in which the
rate of income is equal to the diameter of the circle. When income
changes in the context of a simultaneous analysis the circle either
expands or contracts-- never-the-less where ever the expenditure is
inserted into the flow the circle is maintained without distortion.--
by expanding or contracting everywhere-- simultaneously.
I think I lost you. What will be the effect if the consumer attempts to
save by reducing expenditures relative to income? Won't the effect be
that the consumer will spend less and thus take in more income relative
to expenses?
Suppose the producer in the figure above attempts to make a profit.
An attempt to reduce expenses below income will have the effect, not
of creating a difference between income and expenditure, but rather
the effect of reducing _both_ expenditure and income. And, reducing
them in such a way that they are always equal to each other.
But this seems wrong because producers can increase profit by reducing
expenses. Your "simultaneous analysis" seems to be saying something
happens that clearly does not happen. A control loop can be easily
designed to control for profit by adjusting expenses.
Rick's H. Economicus use of a controller to bring two parts of the
wheel of circulation into equation with each other displays a
fundamental misunderstanding of the principles involved in a
simultaneous equation model of transactions in the aggregate. In the
aggregate sales equal purchases. That isn't all one needs to know.
I don't see what this part of the model (the part that controls the
difference between what is spent by the aggregate producer (PQ') and
what is returned as income by the aggregate consumer (P'Q') ) has to do
with sales and purchases. PQ' is not a measure of what is sold; it's a
measure of what is _produced_. Is it all the goods and services
produced in one unit time by the aggregate producer, measured in terms
of the average dollar value paid to produce those goods and services
(P). P'Q' is a measure of purchases; it is a measure of the amount
spent to purchase whatever portion of PQ' is consumed by the aggregate
consumer. When P'Q' = PQ' then all goods and services produced in a
unit time by the aggregate producer are consumed by the aggregate
consumer.
However, violation of this obvious equation is going to result in
mischief-- which is what I said in Boston, and also in St. Louis as
well (if my memory is correct).
It's certainly true that, in an economic exchange, at the individual or
aggregate level, the dollar value of what is sold is exactly equal to
the dollar value of what is purchased. But this is not quite what H.
Economicus is about. The model is really just two control loops, one
controlling for P'Q' (goods and services in dollar value P') and the
other controlling for PQ'-P'Q' (balanced books). The first loop
(Composite GNP Controller) is trying to purchase a reference amount of
goods and services, perceived as P'Q', which can be viewed as the spent
component of consumer income. The second controller (Composite Manager)
is trying to keep the difference between what is spent by the aggregate
producer to produce GNP (PQ') and what is taken in as income for the
portion of it that is consumer (P'Q') equal to zero.
Now that I go through this reanalysis I think I can see why Bill P.
might not have liked the model. I guess the best way to describe it is
that it doesn't keep the variables clearly separated. For example, Q'
shows up in both producer expenses (PQ') and consumer purchases (P'Q').
Since Q' is the amount of goods and services purchased and produced,
this value should be different in the two different equations, say as
Q' and Q. The amount of Q' purchased would be Q and it will only equal
the amount produced Q' when all of consumer income is spent (no
leakage). I take this into account properly in the spreadsheet because
inventory is greater than zero when there leakage. But, still, the
whole model could be built up more coherently. My only excuse is that,
as I said, I was building the model to match TCP's analysis, where Q
(as opposed to Q') meant output without leakage, not output consumed.
Table 1 shows that the behavior of my model matched TCP's analytic
results exactly, at least in terms of the behavior of Q'/Q and
inflation rate, both as a function of leakage. The prediction that
didn't work was rate of growth as a function of leakage. It's this
result that led to the discovery that the effect of leakage on growth
rate is an assumption rather than a derivation of the leakage model.
Best regards
Rick
···
---
Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400