[From Bill Williams 31 January 2004 12:40 PM CST]
[Kenny Kitzke (2004.01.31)] said,
Rickster, I did not call you a fool.
I admire your restraint Kenny.
Bill Williams
[From Bill Williams 31 January 2004 12:40 PM CST]
[Kenny Kitzke (2004.01.31)] said,
Rickster, I did not call you a fool.
I admire your restraint Kenny.
Bill Williams
[From Bill Williams 31 January 2004 12:40 PM CST]
The problems that are becoming evident in the on-going discussion on the economics thread repeat in some respects the difficulties that have been experienced in economics. I've cited some representative publications that discuss these problems. Leamer's paper "Let's Take the Con out of Econometrics," provides a famous often cited discussion of the sources of difficulty.
Leamer, Edward E. 1983 "Let's Take the Con out of Econometrics,"
American Economic Review vol 73 # 1 March p. 31-43.
Caudill provides a discription of how bias on the part of the researcher may taint the result of econometric work. The methodological measures currently insisted upon in economics to minimize the problem of bias sometimes seem to me to be excessive. As much as, if not more time and effort is spent attempting to avoid a biased approach to inference, as is spent in conducting the inference itself.
Problems of Inference and Data Mining:
Caudill, Steven. 1990 "Econometrics in Theory and Practice."
Eastern Economic Journal vol 16 # 3 July-September p 249-56.
I am not recommending Boland's approach, but I mention his work
as a comparision to Powers' conception of modeling in control
theory.
Boland, Lawrence. 1989 _The Methodology of Economic Model
Building: Methodology After Samuelson." New York:
Routledge.
Sometimes a juicy quote comes in handy, and Bunge provides such a quote.
Frey, Bruno, et al 1984 "Consensus and Dissention Among Economists:
An Empirical Inquiry," American Economic Review vol 74 # 5
December p. 986-94.
Singham, Mano. 2OOO "The Science and Religion Wars."
Feburary Phi Delta Kappan p. 425-32.
p. 432.
Mario Bunge said,
"Walk a few steps away from the faculties of science, engineering
and medicine. Walk toward the faculty of arts. Here, you will meet
another world, one where falsities are manufactured in industrial
quantities ... We should expell these charlatans from the university."
Snow, C. P. 1969 _Two Cultures and a Second Look_ Cambridge
University Press
Cornelius, David K. and Vincent, Edwin St. Vincent 1964
_Cultures in Confict: Perspectives on the Snow-Leavis
Controversy_ Chicgo: Scott, Foresman and Company
Hollander, Samuel. 1973 _The Economics of Adam Smith_
University of Toronto Press
p. 3. Frank Knight
"Economic theory is nothing, or is signficant only as an intelletual
curiousity, unless it shows how, in a price economy, the price system
works out .... Canadian Journal of Economics and Political Science
I 1935 p. 171-96. reprinted in
_On the History and Method of Economics_ Chicago 1956 p. 42
p. 5.
footnote
"Practically all the fundamentals of economic analysis are too
obvious to be really overlooked by any observer who thinks about
the problems at all.
I am not convinced that any thinking is required for an observer
to come to the conclusion that economic relationships are "obvious."
Hollander, Samuel. 1979 _The Economics of David Ricardo_
University of Toronto Press
Leavis, F. R. 1972 _Nor Shall My Sword: Discourses on Pluralism,
Compassion and Social Hope_ New York: Barnes and Noble Books
"Yet Snow is in fact portentiously ignorant. " p. 41.
"The compentary I have to make on him is necessrily drastic
and dismissive." p. 43.
Leavis has a way with words, nearly all of us have much to learn
from Leavis.
Bill Williams
[From Bill Williams 31 January 2004 1:30 PM CST]
In discussion between Rick Marken and Kenny Kitzke the following definitions were specified:
[From Kenny Kitzke (2004.01.31PM)]
<Rick Marken (2004.01.29.2150)>
I am summarizing what you stated concerning your data and its source (the Statistical Abstract of the US, Table #650) from
1951-2001:
- Your lp is defined as Gross Private Domestic Investment
- Your lg is defined as Gross Government Investment
- Your [lp + lg] is defined as Gross Investment.
Further, you use for GDP (Gross Domestic Product) from the US Commerce Department from 1951-1998.
I am a bit puzzled why when as I understand it TCP's argument involved a relationship between investment, as he defined it, and a growth in income
the discussion and data analysis is being carried out in terms not of investment or income but rather in terms of what ought to be obviously measures of a different set of concepts.
Ip is clearly _NOT_ despite its name a measure of investment. Rather Ip is a report of investment _PLUS_ the user cost of capital experienced in the production of current income. Why anyone would expect that a cost of production which contributes nothing at all by defintion to dK/dt would be expected to contribute to a growth of income.
And, in similiar fashion, why if TCP was concerned with the relationship between investment and income is the analysis going on in terms of GDP which it is clear is not a measure of income but instead a measure of income _PLUS_ the user cost experienced in the provision of that income.
What is going forward, as is in evidence above, is a discussion of the cost of orange juice in terms of changes in the factor cost experienced by an apple farmer. Not that the discussion is entirely without interest, but the data manipulations as Kenny has so aptly described them, has drifted a long ways from the initial topic. This isn't altogather unusual in such discussion, but I thought I'd point it out.
It has been suggested here that the St. Louis Fed Bank's tables are widely considered the best availible. I wouldn't know, but to avoid a possible charge of maliciously withholding information I pass this along.
My thanks again to Kenny for his contribution to the discussion.
Bill Williams
[From Bill Powers (2004.01.31.1242 MST)]
Bill Williams 31 January 2004 12:40 PM CST --
I'll look at as many of the sources you cite as possible in my position.
However, I'm not convinced of the value of amassing a lot of quotations by
people with whose opinions you agree. Are those all the published opinions
that there are? And even if all opinions were on the same side, I would
still want to know why. There is very little of "why" in these citations.
Best,
Bill P.
[From Bill Powers (2004.01.31.1247 MST)]
Bill Williams 31 January 2004 1:30 PM CST --
I am a bit puzzled why when as I understand it TCP's argument involved a
relationship between investment, as he defined it, and a growth in income
the discussion and data analysis is being carried out in terms not of
investment or income but rather in terms of what ought to be obviously
measures of a different set of concepts.
On what basis do you think that TCPs's argument involved that relationship?
Ip is clearly _NOT_ despite its name a measure of investment. Rather Ip is
a report of investment _PLUS_ the user cost of capital experienced in the
production of current income. Why anyone would expect that a cost of
production which contributes nothing at all by defintion to dK/dt would be
expected to contribute to a growth of income.
I must be missing something here. User cost, you agreed, was defined as in
Keynes Chapter 6, page 52 or 53 as I remember. This cost is defined as
User cost = (G' - B') - (G - A1)
where G' = the value that the capital equipment would have had at the end
of the accounting period if the equipment had not been used and
B' had been spend on maintaining it,
G = the value of the capital equipment, "which includes both his
stocks of unfinished goods or working capital and his stocks of finished goods"
A1 = "a certain sum" spent on "purchasing finished output from other
entrepreneurs,"
Would this be the same thing as an entry called "Capital consumption
allowances"? If so, this might offer a way to convert the tables for "Gross
private domestic investment" to "Net private domestic investment" as is
done in S.A, for 1986 (but not any other year I have seen), table #899,
page 528.
The entries are as follows, in billions of current dollars (not corrected
for inflation):
Year Gross Net (private investment)
1984 627.8 234.5
1983 471.6 94.6
1982 414.9 56.0
1981 484.2 153.9
1980 401.9 108.7
1979 423.0 167.0
1978 386.6 164.1
1977 324.1 128.9
1976 257.9 82.8
1975 206.1 46.8
1970 144.2 56.2
1965 113.5 57.5
Those are pretty hefty "allowances." Rick, can you correct these numbers to
constant dollars and see what the results are for these 12 entries? I don't
suppose that there would be a corresponding government allowance, but
perhaps Bill W. can advise if there would be. At least you can compute a
rough ratio of Ip to GNP and dGNP/dt. Beware; the last two entries are not
consecutive years (the above is for all the years that were covered in that
table. No years were omitted).
And, in similiar fashion, why if TCP was concerned with the relationship
between investment and income is the analysis going on in terms of GDP
which it is clear is not a measure of income but instead a measure of
income _PLUS_ the user cost experienced in the provision of that income.
Again, what is your basis for saying TCP was concerned with the
relationship between investment and income? His analysis was in terms of
investment and GNP (rather than GDP).
It has been suggested here that the St. Louis Fed Bank's tables are widely
considered the best availible. I wouldn't know, but to avoid a possible
charge of maliciously withholding information I pass this along.
What is it you are passing along about those tables? That they are not the
best available? Actually, they are simply the most complete we have found,
so accusations of ignoring data that did not fit the hypothesis can be put
to rest. Judging from your lack of comment on that subject -- well, I had
better not judge. A comment from you would seem to be in order.
Best,
Bill P.
[From Bill Williams 31 January 2004 2:38 PM CST]
[From Bill Powers (2004.01.31.1242 MST)]
Bill Williams 31 January 2004 12:40 PM CST --
I'll look at as many of the sources you cite as possible in my position.
However, I'm not convinced of the value of amassing a lot of quotations by
people with whose opinions you agree.
# I didn't say I agreed with them. My purpose in citing them was to point
# out that the problems that are emerging with testing TCP's proposition
# are not unique.
Are those all the published opinions that there are?
# Do you want more? What kind of opinions do you want?
And even if all opinions were on the same side, I would
still want to know why.
# I would rather find the really good one liners myself.
There is very little of "why" in these citations.
# You mean to say that you've already read leamers "How to Take the Con .."
# paper already?
#
# I think this is precisely the point that Leamer "Take the Con out of
# Econometics" attempts to get at. But, I'm not sure that we are
# talking abut the same issue.
# I have a student who took a math degree at 19. But he didn't think that
# his raw mathematically ablity was adaqute to support a career as a
# first rate mathematicican, so he switched to economics. He says that
# there is a simiariarity between economic statistical studies and those
# that are carried out by people doing work in agricultural fields. The
# data in both fields are frequently inadaquate. The difference between
# the two fields is that there is often much less ideological bias
# involved in determining which seed is more productive than in economic
# questions in ecoomic research. Suppose we consider the question "Which
# is better Capitalism or Socialism?" I can immagine that this might pose
# some difficulties. Such as how do we measure which is better.
# We'd have to establish some criteria and either take or collect some data.
# What I'm getting to is, how come this number crunching going on regarding
# the connection between investment is using Gross Investment, and Gross # Product when these two measures include costs in addition to the investment
# and income?
Bill Williams
PS If you are going to do some reading, why not down load Professor
Druun's disertation. I hope there is a possiblity that a control
theory type agent could be inserted in her program in replacement
of a rule based agent.
[From Bill Williams 31 January 2004 3:00 PM CST]
[From Bill Powers (2004.01.31.1247 MST)]
Bill Williams 31 January 2004 1:30 PM CST --
I am a bit puzzled why when as I understand it TCP's argument involved a
relationship between investment, as he defined it, and a growth in income
the discussion and data analysis is being carried out in terms not of
investment or income but rather in terms of what ought to be obviously
measures of a different set of concepts.
On what basis do you think that TCPs's argument involved that relationship?
# The way I was first introduce to TCP's work was a description, as I
# remember it, was that "Investment doesn't cause a growth in income."
# I think that I would have remebered if what had been said was that
# "Gross Investment" doesn't have any effect upon "GNP." This would have
# been an entirely different style of presentation.
Ip is clearly _NOT_ despite its name a measure of investment. Rather Ip is
a report of investment _PLUS_ the user cost of capital experienced in the
production of current income. Why anyone would expect that a cost of
production which contributes nothing at all by defintion to dK/dt would be
expected to contribute to a growth of income.
I must be missing something here. User cost, you agreed, was defined as in
Keynes Chapter 6, page 52 or 53 as I remember. This cost is defined as
User cost = (G' - B') - (G - A1)
where G' = the value that the capital equipment would have had at the end
of the accounting period if the equipment had not been used and
B' had been spend on maintaining it,
G = the value of the capital equipment, "which includes both his
stocks of unfinished goods or working capital and his stocks of finished goods"
A1 = "a certain sum" spent on "purchasing finished output from other
entrepreneurs,"
Would this be the same thing as an entry called "Capital consumption
allowances"? If so, this might offer a way to convert the tables for "Gross
private domestic investment" to "Net private domestic investment" as is
done in S.A, for 1986 (but not any other year I have seen), table #899,
page 528.
The entries are as follows, in billions of current dollars (not corrected
for inflation):
Year Gross Net (private investment)
1984 627.8 234.5
1983 471.6 94.6
1982 414.9 56.0
1981 484.2 153.9
1980 401.9 108.7
1979 423.0 167.0
1978 386.6 164.1
1977 324.1 128.9
1976 257.9 82.8
1975 206.1 46.8
1970 144.2 56.2
1965 113.5 57.5
Those are pretty hefty "allowances."
# Pleas don't blame me. I just work here.
# Once again, I'd like to point out that despite as you grumble
# about my having a Ph.D in economics I am not an expert in regard
# to numbergrunching or account. But, the table above does look to
# me as if it may represent the relationship between gross and net
# investment that I've been attempting to bring into the discussion.
#
# It is sort of a thrill that the imps in the basement may really
# have been doing their work.
Rick, can you correct these numbers to
constant dollars and see what the results are for these 12 entries? I don't
suppose that there would be a corresponding government allowance, but
perhaps Bill W. can advise if there would be.
# I thought as stuff went flying by that I saw a symbol for Net governemnt
# investment. I cold be mistaken.
#
# Do you suppose that we could put the fan someplace else-- just for a
# while you understand.
At least you can compute a
rough ratio of Ip to GNP and dGNP/dt. Beware; the last two entries are not
consecutive years (the above is for all the years that were covered in that
table. No years were omitted).
# " No years were omitted). " Sad experience has resulted in imps being # trained to ask the most offensive questions and say the most offensive
# things. They hired a former academian from far away called B.F. Watson
# to improve the training program. Mostly this consists of the imps
# gathering in a small room and shouting the most offensive things at
# each other that they can immagine. I took Snips with me to my first
# session and Snips bit the trainer before we even really got going.
And, in similiar fashion, why if TCP was concerned with the relationship
between investment and income is the analysis going on in terms of GDP
which it is clear is not a measure of income but instead a measure of
income _PLUS_ the user cost experienced in the provision of that income.
Again, what is your basis for saying TCP was concerned with the
relationship between investment and income?
# My basis is how I remember your description of what your dad had
# done. We don't have a record of what you said then, but we do have
# a record of your recently desribing a speculative purches of stocks and
# bonds as investment.
His analysis was in terms of
investment and GNP (rather than GDP).
# Sorry I must have been thinking about Professor Druun again.
It has been suggested here that the St. Louis Fed Bank's tables are widely
considered the best availible. I wouldn't know, but to avoid a possible
charge of maliciously withholding information I pass this along.
What is it you are passing along about those tables? That they are not the
best available? Actually, they are simply the most complete we have found,
so accusations of ignoring data that did not fit the hypothesis can be put
to rest. Judging from your lack of comment on that subject -- well, I had
better not judge. A comment from you would seem to be in order.
# All I saying is that I'm not a number cruncher, so what I am reporting
# is second hand information. I think it is probably good second hand
# information. But, I have a habit of trying to be careful about making
# statements about stuff that I haven't used myself. Again, all I is
# that Ta-Ho says thinks that it is "the best." For what is is worth
# the St Louis Fed seems to have more interest in economic research
# than other Federal Reserve Banks.
# I sometimes ask myself, aside from having so much fun, why are we doing # this?
# Is the point of this exercise one of providing a basis for some
# policy measure? If investment doesn't have the effect of increasing
# income, does that mean that tax breaks for investment ought to be
# reduced?
Whereever the thread is going, it _is_ sort of interesting.
Zdravka Todorova a student here at UMKC (she's passed her comps)
recently subscribed to the CSGnet. This I think may answer one of
David's questions. The choice language used on CSGnet doesn't
phase her. She is Bulgarian you understand. We call her the Russian.
And when we do she used to turn red. There is, if you were unaware of
it, some misunderstanding between Russians and Bulgarians. And, she
used to resent our calling her "the Russian." Its all part of the
imps training and she shows promise of eventually becoming "difficult
to deal with." Now she justs smiles and tells us "to stuff it."
Bill Williams
[From Bill Williams 31 January 2004 3:55 PM CST]
[From Bill Powers (2004.01.31.1247 MST)]
The entries are as follows, in billions of current dollars (not corrected
for inflation):
Year Gross Net (private investment)
1984 627.8 234.5
1983 471.6 94.6
1982 414.9 56.0
1981 484.2 153.9
1980 401.9 108.7
1979 423.0 167.0
1978 386.6 164.1
1977 324.1 128.9
1976 257.9 82.8
1975 206.1 46.8
1970 144.2 56.2
1965 113.5 57.5
Those are pretty hefty "allowances."
# And, what is the answer to the question "What does a tax attorney drive?"
#
# Would you believe, "Anything he damn well wants."
# This might also be the time to consider how changes in investment
# as you seem to have managed to have broken it out in the table
# above relate to changes in income.
# This is definitely becoming interesting.
Bill Williams
[From Bill Williams 8 February 2004 2:10 PM CST]
The AFEE webb site is a bit of a mess, so I picked out the reference to the
Peach and Webb paper.
Peach, James T
Randomly specified macroeconomic models: some implications for model
selection, James T Peach and James L Webb, September 1983, p.
697-720
Journal of Economic Issues.
The arguement of the Peach/Webb paper is simple, with random specification
of economic models it is possible to get any result you want. How
wonderful!
Have the implications of the Peach/Webb paper been fully absorbed by those
attempting to model the economy? All too evidently the obvious lesson hasn't
yet been learned. As Kenneth Arrow said, "Data mining is a sin."
Kenny's description of this sort of approach, as "masturbation with the
data" is an acute perception.
How long it will take to for Bill and Rick to generate a Test Bed
approaching the problem this way may not be easy to predict, or even
describe.
However, are you familiar with the use of exponates as a way to express
very large numbers, or are you familiar, by chance with the Pre-Cambrion,
or
Jurastic eras and do you have some idea how very long ago that was?
Bill Williiams