[From Bill Powers (2004.01.30.0900 MST)]
Bill Williams 30 January 2004 6:42 AM CST --
Do you know the difference between the measure of investment that is a
contribution to total capital, and the measure of investment that is is
inclusive of user costs? Whether or not you know the difference, you've
used the wrong measures.
Are you referring to the User Costs defined on p. 53 of the General Theory?
Or is there a different definition in use?
And, while you are obtaining the net data, why not extend the analysis to
the interwar period, the Great Depression and WWII.
That is proving rather difficult. Can you work out which tables, or which
lines from different tables, contain the necessary data? It's very hard to
trace information backward through time, because the people compiling the
SA kept changing the data presentations. That may improve the information
going forward, but it plays hell with historical searches: the most useful
tables disappear as you go backward.
If Bill's dad had an opinion that his ratio didn't extend to the years of
the Great Depression and World War II
He mentioned WW II, not the Great Depression. But if the data are there we
might as well find them and put them into the analysis. We're not trying to
prove that the formula is right; only to find out if it fits the data, or
how much of the data it fits.
Rick says
# So we do get a positive correlation between the absolute level of lp #
(private investment) and growth.
You've confirmed What I was suggesting so many years to Bill's dad-- that
ordinarily it is thought that there is a positive relationship between
investment and economic growth. I gather that even though you are using
the wrong data, even gross investment is coorelated with growth.
Rick said that Ip -- private investment -- shows the highest correlation
with growth. However, the correlation is so low (0.27) that it barely
reveals a relationship and would be useless for predicting anything. A
scatter plot showing a correlation of 0.27 is a cloud of dots that is just
barely detectable as being oval instead of circular. Anyway, a positive
relationship does not reveal cause and effect. It is still possible that
investment follows whatever it is that actually drives growth. You'd see
the same positive correlation either way.
Since gross investment includes user costs, the assocation between net
investment and growth must of mathematical neccesary be higher, perhaps
much higher when the distortion of user costs are removed. It is
ordinarily assumed that dK/dt is proportionate to DY/dt. Who would have
thought otherwise?
The point is not what would be assumed or thought, but what is true. And
this proportionality does not reveal the direction of causation.
And, as I pointed out yesterday there is no reason to assume that net
collective investment is going to generate growth in national income as
measured by the GNP. Because as I explained, (see Peterson ) collective
investment doesn't generate a change in output that would show up as sales.
That sounds interesting. Can you elucidate? Is this like "taking in each
others' laundry?"
If, discounting for government's supposed inefficiency, just to keep Kenny
happy, there will be an implicit, non-market, and hense non-market
measurable, contribution to income by additions to government capital.
Since this flow of income isn't captured by GNP your use of a gross
measure doesn't matter, but conceptually you ought to use government
investment as reported in net terms-- if such a measure is made.
A useful suggestion. I believe it is shown. Could you explain what
"collective investment" means? What would be the indicator in the SA
tables? Or is the point that it doesn't show up there?
I ought to talk to Wally about giving a more extensive treatment to some
of the issues that have come up as a result of the controversy as to
whether investment is associated with economic growth.
I don't think "association" would be controversial, since that would say
nothing about the direction of causation. What may be controversial is the
proposition that investment follows from rather than precedes growth. I
don't mean that the growth happens first, only that conditions which would
lead to growth, such as a population increase or an improvement in
productivity, happen first, creating a situation in which growth is both
possible and needed or wanted. The growth can't happen in a money economy
unless there is some source of money with which to increase the size of the
plant or convert it to the new technology. Money is sought in those cases,
and when obtained is used for investment, if I am using the term correctly
now. Not _all_ growth requires investment, of course, at the micro level.
This way of looking at it gets around the puzzle of how production can be
increased sucessfully before there is money in the hands of consumers to
pay for the product, and how more money can be put in the hands of
consumers without first paying them to increase production. I have never
liked the assertion that supply creates its own demand, simply because
nobody has ever given a plausible explanation of how that can happen (and
there are too many easy counterexamples, starting with buggy whips). It
makes much more sense if the demand is there first, but of course if you
measure demand in terms of how much people spend to buy things, demand
can't come first. The things must be there to be bought.
A control system model gives us a different way of thinking of demand:
unsatisfied reference conditions. This can happen because of shortages, or
as you say because of advertising and other ways of getting people to raise
the reference levels, or just because many people never have satisfied all
their reference conditions. In any case, the people then demand, in the
non-technical sense, certain goods or services, and the risk involved in
tooling up to satisfy the demand is minimal.
I congratulate you on confirming what has been routinely understood since
Adam Smith.
I would say "routinely believed" rather than "routinely understood." Many
things we have found through control theory, such as the idea that behavior
is purposeful, have been believed for a very long time, by most people, but
the phenomenon of purpose has not been understood until quite recently. On
the other side of that coin, the fact that many people have believed
something for a long time is no guarantee that it is true -- I think the
odds are distinctly in the other direction. Proofs -- and disproofs -- are
useful ways of separating correct beliefs from incorrect ones. I'm sure
that economists since Adam Smith have believed at least as many things that
are wrong as are right.
Economic growth is associated with net investment. You've generated a
"proof" that economic growth is associated with gross investment. This
isn't neccesarily the case, but it is among the things that while of no
use, may be nice to know.
I'm hoping that ideas like gross and net investment, and user cost (Keynes'
definition) will be unnecessary in building the basic economic model, the
Test Bed. These technical terms are strongly associated with abstract
theorizing, and involve things that exist primarily in the psyche of the
consumer and entrepreneur, or in attempts to minimize tax liabilities by
accountants. This is not to say they are unreal, but only that they are not
necessary as a way of modeling the basic operation of the economy. They
belong more to models of the agents than to a model of the environment.
Keynes spends a good deal of time talking about the conditions under which
an entrepreneur will choose one course of action over another -- for
example, whether to sell the machinery at market, or to use the machinery
to produce goods which can be sold at a greater net profit. These scenarios
take place in the imagination of the entrepreneur, not out in the world of
real interactions. Once the entrepreneur has come to a decision about what
to do, and starts doing it, there is an action (or a policy) that can be
plugged into the model to see what its consequences will be. But none of
the entrepreneur's calculations or imagined alternatives, or unrealized
possibilities, have any physical effects in the real world. This was my
objection to Keynes' way of defining user costs: they were nothing but a
calculation in the mind of the entrepreneur. As such, they might very well
have influenced his perceptions, and thus his choice of actions, but only
the perceptions and actions would have real consequences. The actions not
taken are simply not taken.
Almost a year ago, I posted this extract from Clarence Ayres' (1946)
_The Divine Right of Capital_" to the CSGnet discussion list. Ayres' is
making a point about the concept of capital. But the point he is making
is applicable to the entire "capitalist" market society. The duality of
money and capital is repeated with variations when we attempt to think
about every aspect, and every element of the marketplace relationships in
a society.
"Even in early modern times industry was growing fast.
That growth was a function of industrial equipment of
the community, as anybody could see even in early times.
To identify that function with the accumulation of money
was to attribute the whole thing to the men who made money.
That is what we accomplish by calling both things 'capital.'
The argument runs as follows: Economic progress results
from the growth of the material equipment of industry, that
is to say, capital, that is to say money, is created by
'saving': therefore, econmic progress is made possible by
'saving.' In this fashion the money power became functional.
This is the idea which was forming in the minds of the
merchants and their spokesmen during the sixteenth century.
The word 'capital' first begins to appear in economic
writtings just about the middle of that century. By the time
of Locke it was a commonplace. By the time of Adam Smith it
was a 'law of nature.' In our time it is so familiar that we
are shocked to discover that every time we use the word 'capital'
we are indulging in double-talk.
The double-talk of capital has become an engrained habit.
It runs through virtually every textbook. Textbook writers
solemnly warn their students of the dangers of confusion
which lurk behind this word. Resolutely they insist that it
must be used to refer to one thing or the other, and not both.
With exemplary clarity they declare, for example, that they
propose to use it to refer only to the physical equipment of
industry: plant, manchines, raw materials, and so forth. And
in the very next paragraph we find them saying that capital
is brought into existence by saving! But now, obviously
they are talking about money!
That is how it is done." p. 10-11 ?? check page #
Admirably clear writing. I find it encouraging that Ayres is trying to get
back to more concrete identifications of what people have come to treat as
abstractions. This is essential for the modeling/simulation process,
because simulations, as opposed to conceptual models, have to deal in
things that happen as a function of other things that happen. I fully plan
to equip the "plant" in the Test Bed with machinery that costs money to
buy, install,. maintain, and operate, but there will be no need to identify
it as "capital." It is just machinery which plays a part in setting the
coefficient of productivity. And it will never be confused with money just
because it might hypothetically be bought or sold at some price.
The same goes for investment and savings. These are just different uses to
which money can be put. Basically, in the Test Bed, ALL money is saved when
it is received (though I'm still taking advice on how to handle non-bank
credit). Even borrowed money goes immediately into savings. By saying it's
saved, I mean only to say that it is accumulated in an account somewhere.
Then this money is taken out, immediately or later, to be used for paying
wages to workers and managers and capital income (including profits) to
owners and investors, and for buying more machinery or maintaining what
exists (real depreciation costs, as opposed to mere calculations of
lessened value). So "savings" is simply whatever exists in the place of
accumulation at a given time; it is being added to and depleted constantly
by all operations that lead to acquiring money or spending it, which go on
simultaneously and continuously. At present no interest accrues to savings,
but that is easily remedied.
So in the Test Bed, or perhaps we could say the Basic Economics model,words
like interest and savings and investment and capital do not have all the
nuances of meaning that they have to theoreticians in economics, or to
entrepreneurs, or even to consumers. They are just names for places where
money is temporarily kept, or for processes involving transfers of money
(or credit) from one entity to another. Also, goods are handled as goods,
not as equivalent money value. They have to be handled as goods&services
because consumers have reference levels for them as goods&services. They
can't eat money.
I don't know why I have had such trouble with saying this, but I think this
is the root of many of my difficulties on CSGnet. It is not my intent to
say anything at an abstract level about economics. The way we classify
things has no impact on how a simulation works. Whether something is called
profit or earnings or wages is of no consequence in the simulation; we
simply keep track of who gets how much, and for doing what, and make sure
the books always balance every 0.001 day.
I think TCP was trying for something similar, but not knowing how to do
simulations or mathematical modeling, he confused the two worlds, the world
of imaginary economic classifications, and the world of actual processes
and transactions. Also, he didn't believe that individual characteristics
had anything to do with economics, which made it a little difficult for him
to accept the idea that people have reference levels and try to satisfy
them. He wouldn't have touched such a proposition with the proverbial
304-centimeter Pope.
Best,
Bill P.