| |
The METAMODEL contains two initial designations that must be made prior
to execution of the discovery system in the
computer. Firstly the user must designate which descriptive variables
among the current-valued variables are the
problematic variables, i.e. those that identify the
problem the theory is to solve and also the
cognizant profession.
In the application to the trade cycle
problem, the problematic variables are aggregate
employment and aggregate real income for the
national economy. Every macroeconometric model printed in the output state
description generated by the system will contain
these problematic variables and the equations
determining their numeric values. Secondly the user must designate which among the
current-valued variables are exogenous variables.
These variables have their values determined
for a generated model and not by it; the values are
determined independently by economic policy
decisions. The
exogenous variables designated in the trade cycle
application are Federal real aggregate fiscal
expenditures, Federal real aggregate fiscal tax
revenues, and the Federal Reserve’s measure of the
aggregate nominal money stock.
These two types of designations together with
other information such as the number of observations
in the time series data are entered into a control
record that is immediately read when the system is
executed. The
control record is followed by records containing the
character names of the input variables with separate
identifiers for current values and lagged-valued
variables, and then the time series data records
follow.
The METAMODEL discovery system is a FORTRAN computer program having an architecture consisting of a main
program, SLECTR,
and two called subroutines, REGRES
and SOLVER. SLECTR is the
combinatorial procedure that selects nonredundant
combinations of language elements.
It contains a switch, which is initially set
to the open state.
When the switch is open, SLECTR
selects combinations of time series from the
input file initially read by the system, and for
each selection it calls the REGRES
subroutine. REGRES is an ordinary-least-squares-regression procedure
that statistically estimates an intercept and
coefficients thereby constructing an equation for
the selection of variables passed to it by SLECTR. If the
estimated equation does not have a satisfactory
R-squared coefficient-of-multiple-determination
statistic associated with it, according to a minimum
value given to the system and stored in its control
record, then control is returned to SLECTR.
But if the R-squared statistic is
satisfactory, the equation is stored as a record in
an accumulation file before control is returned to SLECTR.
After SELCTR
has made all its selections for nonredundant
combinations of as many as six variables, the switch
is closed, and SELCTR
repeats its combinatorial procedure.
With the switch closed SELECTR
makes selections of estimated equations from the
accumulation file previously generated by REGRES,
and for each selection it calls subroutine SOLVER.
SOLVER
solves the multi-equation model as a
simultaneous-equation model, and then executes the
model to generate a reconstruction of the historical
data. In
order to accomplish this, there are certain criteria
that any selection of equations must satisfy, and SOLVER
checks for these conditions.
Firstly the combination of equations
constituting the model must contain equations that
determine the two designated problematic variables.
Secondly the model must be uniquely
determined, such that there are as many
current-valued endogenous variables as there are
equations. Thirdly
the model must be recursive, such that there is at
least one current-valued variable for each
lagged-valued variable describing the same
phenomenon. Fourthly
the model must be a minimal statement, such that it
contains no current-valued variables except the
problematic variables, that occurs but once in the
model and is not needed to evaluate a lagged-valued
variable describing the same phenomenon.
When SOLVER
finds a combination that does not satisfy all these
criteria, it returns control to SLECTR for another combination of equations.
Models that do satisfy all these criteria are
capable of being solved, and SOLVER
solves and iterates the model both to recreate the
history with synthetic data for the years 1921
through 1933, and to make a one-period out-of-sample
postdictive forecast for the year 1934.
The control record for the system also
contains a minimum error for the forecasts of the
problematic variables, and the final test for the
model is for its forecast accuracy.
Each model that also satisfies this criterion
is outputted to the printer and printed in
conventional mathematical form with each equation
listed together with its associated R-squared
statistic. The
output also lists the synthetic data generated by
the iteration of the model together with the
forecast values for all its endogenous variables.
The computing equipment available at the time
the METAMODEL
discovery system was created did not permit a
complete operation of the system, but partial runs
demonstrated that the system would generate a
satisfactory Keynesian model.
There are many and various
artificial-intelligence discovery system designs,
but Hickey’s design was motivated by his objective
of using the techniques and formalisms that are
actually used by econometricians.
Unlike the Logical Positivists, who relied on
symbolic logic to represent the language of science,
Hickey wanted to use the ordinary language
prevailing in the science for which he developed his
discovery system. Thus his system uses the
ordinary-least-squares regression statistical
estimation technique for estimating the parameters
of equations that are assembled into first-degree,
higher-order difference equation systems.
Two-stage-least-squares can be applied to the
outputted models if they are not just identified.
The truly noteworthy difference between
Hickey and the conventional neoclassical economists
using Haavelmo’s agenda is their respective
philosophies of science.
The neoclassicals practice the Romantic
philosophy of science, while Hickey is a Pragmatist.
Hickey’s combination of conventional
econometric modeling techniques with the
contemporary Pragmatist philosophy of science has
proved to be very fruitful for Hickey’s
professional research work; he has made his living
as a research economist for thirty years with his METAMODEL
system.
Four years after designing and testing his METAMODEL
discovery system with Keynes’ theory in
economics, Hickey had occasion and opportunity to
use the system to address a contemporary problem in
social science.
At that time he was a profit analyst in the
analysis and statistics bureau of the finance
department of United States Steel Corporation, the
largest domestic steel manufacturer. He had completed a conventionally Keynesian quarterly
macroeconometric forecasting model, but found that
the model was not performing satisfactorily.
This occurred during the years following the
large increase in crude oil prices imposed in 1973
by the Organization of Petroleum Exporting Countries
(OPEC), and no macroeconometric models available at
the time had the consequences of this shock in the
sample data available to estimate the models
statistically.
Many economists reacted to the structural
breakdown of their models with patience, and waited
the generation of new data.
Others, however, believed that more than oil
prices were at fault, and that there are more basic
reasons for dissatisfaction with their models.
One such group as mentioned above was the
rational expectations economists, and they had their
distinctive agenda.
Hickey also believed that more was involved
than inadequate sample data.
But unlike the rational expectations
advocates he views the phenomenon of structural
breakdown in the same manner as did Haavelmo, who
maintained that the problem is remedied by
introducing into the model new variables
representing missing factors, the absence of which
had caused the breakdown.
But unlike Haavelmo, Hickey agrees with the
Institutionalist economists that neoclassical
economics limits economic explanation to an
excessively small number of factors, and that it
assumes incorrectly that all the other complexities
in the real world are irrelevant.
Furthermore Hickey is not philosophically
sympathetic to the Romanticism in neoclassical
economics, and he prefers the explicitly Pragmatic
orientation of the Institutionalist economists.
However, Institutionalists did not make
econometric models; they were usually more
interested in the historical evolution of economic
institutions. Hickey ventured beyond conventional
Institutionalists and decided to integrate
functionalist sociology into his econometric model,
even though functionalists do not make econometric
models either.
Functionalism in sociology is the thesis that
all institutions of a national society are
interrelated. Therefore he used his METAMODEL
discovery system to investigate how variables
representing each of the five basic institutions of
the American society can be related by statistically
estimated equations of the type used in econometric
models. Both
the sociology in the model generated with the
discovery system and the truculent philosophical
rejection by the Romantic sociologists to his use of
the discovery system, are discussed in the sections
below. An
important complicating fact that is operative in the
sociologists’ rejection of Hickey’s work, is
that his system does not use the statistical and
mathematical techniques that might be called the
ordinary language of the sociologists; instead his
system generated models for which the
sociologists’ education leaves them professionally
incompetent and technically unprepared.
One noteworthy consideration in the present
context is the modifications he made to his METAMODEL system, which enabled him to run it as an integrated
system. In
later years he had access to much better computer
equipment than the machine he used to develop the METAMODEL,
but some modifications to the design of the system
were nevertheless needed. One modification made the system store in the accumulation
file only one equation of all those constructed for
the same dependent variable and having the same
number of independent variables. The one saved for further processing is that having the
highest R-squared statistic.
The effect of this modification is to reduce
significantly the number of equations available for
selection by SOLVER,
and therefore to reduce the number of models
generated for testing for output.
However, this modification has been
eliminated from the commercial version of the METAMODEL, which now runs without subroutine SOLVER. Nonetheless,
the fact that in any case the system generates many
alternative equations and models that are
empirically acceptable is an example of the
contemporary Pragmatist’s thesis of empirical
underdetermination of language and of scientific
pluralism. For
Romanticist and Positivist philosophers of science,
this is an argument against development of
hypotheses by data analysis, and an argument for
invoking some prior ontology with its concept of
causality. But
for the contemporary Pragmatist, pluralism simply a
routine fact of life in basic scientific research,
just as it was for Einstein who called such
pluralism an “embarrassment of riches.”
A second modification is the introduction of
a new test criterion in subroutine SOLVER that tests for the simulation of an inflection point in the
synthetic data.
A model assembled by SOLVER
that cannot simulate in the synthetic data the
occurrence in the actual data of an inflection
point, is rejected by SOLVER
and is not sent to the printer for display.
The modified version of the METAMODEL
was executed to make macrosociological models with
eleven current-valued input variables, and each was
allowed two lagged-valued variables.
The total number of equations estimated and
stored by REGRES
for further processing by SOLVER
was thirteen, and the total number of
macrosociological models generated and critically
accepted by SOLVER
for output was three models.
As it happens, two of the three models were
actually the same model for reasons that SOLVER cannot detect, and so the total number of models actually
outputted was only two.
In response to inquiries during the first
year of publication of Introduction to Metascience Hickey released a source-code listing of
the FORTRAN
statements of the METAMODEL,
and issued it as a supplement to the monograph.
This supplement also contained a listing of
the input data for the sociological application and
a list of the printed output models generated with
the modified version of the discovery system.
These functionalist macrosociometric models
generated by the METAMODEL discovery system were intended to be used as a guide for
integrating sociological, demographic, and human
ecological factors into an integrated
macrosociodemographic-econometric model of the U.S.
national society.
However, circumstances precluded Hickey’s
accomplishing this more ambitious objective until
the 1980’s, when he was the Deputy Director of
Economic Analysis and Senior Economist for the
Division of Economic Analysis of the Indiana
Department of Commerce.
The METAMODEL
discovery system as originally designed was
inadequate to such a project, and it was necessary
to revise the design.
Subroutine SOLVER
was the principal limitation; it could only make
models with as many as twelve equations, while the
integrated model required a number in the range of
one hundred equations.
Consequently Hickey designed and wrote a new METAMODEL
discovery system that performed only the functions
of SLECTR
and REGRES
in the old system.
The new system can accept as many input
variables as the installation's computer and its FORTRAN
compiler can handle.
A description of the resulting integrated
macromodel was published in the state agency's Perspectives on the Indiana Economy (March, 1985).
Later in the September 1985 issue Hickey
published "The Pragmatic Turn in the Economics
Profession and in the Division of Economic Analysis
of the Indiana Department of Commerce", in
which he described the new METAMODEL
and compared it with some VAR
models and the BVAR
system constructed by the rational expectations
advocates. The United States Department of Commerce
has issued him a registered copyright for both the
original and the commercial versions of his METAMODEL
discovery system.
Hickey has used the commercial version of the
METAMODEL
system for many other econometric and
sociodemographic modeling projects for various
employers and clients including USX/United States
Steel Corporation, State of Indiana/Department of
Commerce, BAT(UK)/Brown and Williamson Company,
Pepsi/Quaker Oats Company, Altria/Kraft Foods
Company, Allstate Insurance Company, and TransUnion
LLC. Monthly,
quarterly, and annual versions of the system exist,
and are used for both quantitative market analysis
and for quantitative risk analysis.
The METAMODEL
system has been licensed perpetually to TransUnion
for their consumer credit risk analyses using their
proprietary TrenData aggregated quarterly time
series extracted from their huge national database
of consumer credit files.
They use the models generated by the
discovery system to forecast payment delinquency
rates, bankruptcy filings, average balances and
other consumer borrower characteristics that
constitute risk exposure for lenders, especially
during the contractionary phase of the business
cycle. Hickey has also used the system to discover
the underlying sociological and demographic factors
responsible for the secular long-term market
dynamics of food products and other nondurable
consumer goods.
It might also be noted about these market
analyses that much of the success of the METAMODEL
system is due to Hickey’s Institutionalist
approach in economics.
A review of the membership roster of the
National Association of Business Economists (NABE)
reveals that economists in private industry are
almost never employed in the consumer nonfinancial
services and consumer nondurable goods sectors of
the economy that lie outside the financial,
commodity, or cyclical industrial sectors.
This is due to the education offered by the
graduate schools that is restricted to neoclassical
economics, which has become a kind of a Romanticist
ideology having the status of an orthodox theology.
Employers in the consumer nondurable goods
and nonfinancial services sectors, whose output
accounts for approximately half of the U.S. national
Gross Domestic Product, have no need for
neoclassical orthodoxy.
They have no need for macroeconomic aggregate
income theory of the business cycle, and very
limited need for microeconomic relative price theory
of commodities.
Microeconomic theory treats all industries as
commodities in which there is only price competition
to the exclusion of all franchise or branded
products where advertising and other forms of
nonprice competition prevail.
And it treats aggregate income as the only
aggregate factor to the exclusion of the many
underlying sociodemographic factors considered by
the Institutionalist economist.
The doctrinairism of the neoclassical
academic economists is costing their graduates a
very high opportunity cost in lost employment
opportunities. And it has also created an occupational vacuum which
Institutionalist economists like Hickey have not
hesitated to exploit financially.
From 1978 to 1982 Hickey submitted a paper
describing his macrosociometric model developed with
his METAMODEL
system to several sociological journals.
The paper was acceptable on empirical
grounds. But
the prevailing philosophy of science in academic
sociology is still Romanticism, and since Hickey is
a Pragmatist, the editors of all the journals
rejected the paper for publication.
Romanticism, an early philosophy of science,
is still alive and well in both American academic
sociology, which is still neo-Parsonian, and in
neoclassical economics.
Therefore before turning to Hickey's macro-sociometric
model, consider firstly the Romantic philosophy of
science prevailing in social science today.
Parsons’
Romantic Sociology
Twentieth-century sociology and
twentieth-century physics offer the philosopher of
science a striking contrast.
Physics saw revolutionary developments with
the relativity theory and quantum theory, and these
in turn occasioned the repudiation of Positivism,
the nineteenth-century philosophy of science, by
both the physicists and the philosophers of science.
Sociology on the other hand saw no
advancements like the developments in physics, and
attempted to rework both Positivism and Romanticism,
which contemporary philosophers of science view as
anachronistic.
The result has been the intellectual
stagnation of sociology and the decline of its
academic profession.
This section examines the reworking of the
nineteenth-century philosophies of Romanticism and
Positivism by two sociologists, whose names are
associated with these efforts in twentieth-century
American sociology.
The first and most influential of these is
the Harvard University Romantic sociologist, Talcott
Parsons. Parson’s
Romantic philosophy of science is very uncongenial
to such modern ideas as computerized discovery
systems, but his philosophy is still widely
practiced and is enforced by the editors of the
periodical literature of academic sociology.
This overview of Parsonian Romanticism is
also included here to explain its hostility to
artificial intelligence.
Talcott Parsons (1902-1979) was a professor
at Harvard University from 1927 until his retirement
in 1973. He
wrote an intellectual autobiography, "On
Building Social System Theory", in The Twentieth-Century Sciences (1970). He had majored in philosophy at Amherst University, where he
was also influenced by the Institutionalist
economist, Walton Hamilton, and he studied under the
anthropologist, Bronislaw Malinowski, at the London
School of Economics.
Parsons received his doctorate from the
University of Heidelberg, where he was influenced by
the views of Max Weber of Heidelberg, even though
Parsons attended Heidelberg after Weber's death.
Parsons' principal work is his Structure
of Social Action: A Study in Social Theory with
Special Reference to a Group of Recent European
Writers (1937), an eight-hundred page that
examines the social theories of four writers: Alfred
Marshall, Vilfredo Pareto, Emile Durkheim, and Max
Weber. This
magnum opus is a historical study in philosophy of social science.
Its thesis is that social theory has evolved
beyond Positivism by an "immanent" process
of development within the body of social theory, and
that the outcome has been a “convergence” to a
type of theory that Parsons calls the "voluntaristic
theory of action.”
The voluntaristic theory of social action
encompasses its own philosophy of science which has
evolved with it, and which in turn describes the
evolution of the voluntaristic theory of action set
forth in the book.
|
|
