| |
The data are drawn from the U.S. Commerce
Department's Historical
Statistics of the United States (1976) and
annual issues of Statistical
Abstract of the United States.
The historical time series are from 1920
through 1972.
Firstly the data are either aggregated into
four-year periods before per capita rates are calculated or are four-year averages of per
capita rates calculated by the source.
Then these four-year per
capita
rates are transformed into period-to-period change
rates to enhance sensitivity of the equations and
eliminate collinearity, and then the change ratios
are transformed into index numbers having the last
historical period, which is the first out-of-sample
forecast period, as the base period. The
quantitative theory is a functionalist theory not
only in the sense that it is expressed in
mathematical functions, but also in Merton's sense
of functionalism, because it describes the
interdependence of the types of institutional groups
and the consequences of their interaction for the
macrosociety as a whole as represented by the whole
system of equations.
The mathematical model is a recursive,
first-degree, higher-order difference equation
system.
This
mathematically expressed macrosociological theory is
used to examine social change in the U.S. national
macrosociety with both static and dynamic analyses.
Consider firstly the functionalist static
analysis.
The objective of a static analysis is to
determine whether or not there is a stable
equilibrium, that is, a solution in which the
numeric value of each variable is the same for an
indefinite number of successive time periods through
which the model may be iterated.
Since the numeric values are change rates of per
capita rates, the mathematical equilibrium
solution is one of constant change rates of the per
capita rates for all of the variables.
However these constant change rates may be
positive, zero, or negative, and this is not
necessarily a macrosocial equilibrium.
The classical-functionalist-consensus
equilibrium extended to the scope of the
macrosociety is represented by constant per
capita rates for all the institutional
variables, and these per
capita rates must be high if not actually near
their maximum values to represent a high degree of
consensus and macrosocial integration.
Since constancy of such high per
capita rates implies zero growth of the change
rates, nonexistence of an equilibrium solution
consisting of zero change rates implies the
nonexistence of a classical consensus stable
equilibrium.
As it happens, examination of the equilibrium
solution of the model reveals that a mathematically
consistent zero-growth solution for all the
institutional variables does not exist, and
therefore reveals that consensus equilibrium is not
possible.
In classical functionalist terms this means
that the interinstitutional system of cultural value
orientations of the American national society is
inconsistent or "malintegrated.”
Consider next the dynamic analyses, which
consist of iterating the model to examine its
response properties.
The findings from three simulation
experiments were described in the paper.
In the first experiment the model is iterated
with all its exogenous variables zero growth of
their per
capita rates.
The iteration propagates a time path, which
oscillates with increasing amplitude generates an
intergenerational twenty-eight-year cycle.
Examination of the structure of the model
reveals that the equations determining the change
rates of the per
capita birth and marriage rates are interacting
to create the cycle, and therefore that the cycle
reflects changes in the national demographic
profile, i.e. the age composition of the population.
The explosive instability, however, is due to
the constancy of the per
capita real aggregate income variable occurring
in the marriage rate equation, which represents
fewer marriages in the Great Depression.
The exogenous status of the economic sector
means that there can be neither a dampening feedback
on per capita
real income from the exploding birth rates, nor any
effect from productivity improvements resulting from
technology improvements represented by the patents
for inventions.
Hickey maintains that any macrosociological
model should be integrated into a model of economic
growth, just as the contemporary Institutionalist
economist will maintain that conventional
neoclassical econometric models of economic growth
should be integrated into a macrosociological model
of institutional change.
In the second experiment and in all
succeeding experiments this demographic cycle is
eliminated by removing the birth rate equation from
the model and by making the birth rate exogenously
determined with an assigned constant zero-growth
rate.
When the model thus modified is iterated, it
generates a damped oscillating path which is also
intergenerational, and which converges into a stable
equilibrium of constant growth rates for all its
endogenous variables including the institutional
variables.
The negative feed back producing the
dampening effect involves an interaction between the
equations determining the homicide rate and the high
school graduation rate.
A sustained decline in voluntary compliance
with the minimum conditions for social order as
codified into criminal law, i.e. an increase in the
homicide rate, occasions in due course a corrective
reaction that involves the socializing function of
the educational institution.
However, the resulting growth equilibrium
does not necessarily result in a movement toward
consensus for all the types of institutional groups.
Additional simulations reveal that only when
the growth of the exogenous real per
capita income is made to occur at a rate of no
less than four and one-half percent compounded
annually, does the equilibrium solution result in
positive growth rates for all of the institutional
variables excluding the divorce rate.
This is the minimum annual rate of per capita real economic growth required for the dampening negative
feed back to stabilize the national macrosociety in
an equilibrium growth toward macrosocial consensus
integration.
It is also the minimum annual growth rate for
a full-employment economy.
But the static analysis revealed that this
state of affairs can only be temporary, since the
cultural value system is malintegrated.
Furthermore, historically a four and one-half
percent annual growth rate for real per
capita income has not been sustainable by the
U.S. macroeconomy, because it typically results in
destabilizing inflation rates.
The third experiment consists of shock
simulations in which the model is given an
unrealistically large one-period increase after it
is iterated sufficiently to settle into its
equilibrium growth path.
The limits imposed in reality by the per
capita rates are disregarded in the simulations,
to exhibit the dynamic properties of the model.
In each simulation the shock consists of a
permanent doubling of the per
capita rate for one selected noninstitutional
variable.
In all but one case the shock propagates a
damped oscillating path that settles back into a
stable equilibrium solution.
The exceptional case is the per
capita urban residence rate, and the outcome of
the shock is a nonoscillating explosive
destabilization of the macrosociety.
In this latter case the equations for both
the birth rate and the urban residence rate have
been removed from the model, such that the
population growth cannot be accommodated by internal
migration, and the macrosociety is disturbed beyond
the stabilizing capacity of its interinstitutional
integrative mechanisms.
In summary there are four findings from both
the static and dynamic analyses.
Firstly the interinstitutional cultural value
system of the American national macrosociety is
malintegrated, such that Parsonian macrosocial
consensus equilibrium is not possible.
Secondly if demographic cycles are
exogenously determined, the national society
exhibits a nonsustainable tendency to macrosocial
consensus when per
capita real income grows at a minimum rate of
four and one-half percent compounded annually.
Thirdly this tendency to consensus
equilibrium at this economic growth rate stabilizes
in growth equilibrium, because the
interinstitutional cultural value system contains
relationships that create intergenerational negative
feed back integrative mechanisms, which involve a
corrective reaction to criminal social disorder and
operates through the socializing functions of the
educational institution.
Finally internal migration, an ecological
adjustment to population growth, is necessary for
the institutional integrative mechanisms to be
effective.
Such is a summary of Hickey's findings.
A few years later Hickey developed a larger
model based on the above described macrosociometric
model, which integrated the sociological and
macroeconomic sectors of the nation into one model.
Hickey’s description of the model and
findings from it were published with the title
"The Indiana Economic Growth Model" in the
periodical Perspectives
on the Indiana Economy (March, 1985)
published by the Indiana Department of Commerce.
The model contained over one hundred
equations and was instrumental to the economic
development planning by the government of the state
of Indiana during the eight years of the
administration of Governor Robert Orr.
A
Pragmatist Critique of Academic Sociology’s Weltanschauung
This section consists of Hickey's criticism
of the referee reviews and consequent decisions by
the editors of four sociological journals to reject
the paper. Contrary to these editors Hickey regards
his paper as worthy of publication.
His reports of the sociologists' attempts at
scientific criticism are based on the
correspondences from the editors, which Hickey has
retained.
This sample of seven referee criticisms from
three academic sociology journals is not a random
sample.
It is a selected sample made by the journal
editors, who presumably chose the critics they
deemed best for the topic of the paper.
And it is consistently representative of
academic sociology's institutionalized values and
German Romantic Weltanschauung.
In this respect it is noteworthy that
virtually none of these referee criticisms of
Hickey's paper are empirical, but rather are
attempted criticisms in philosophy of science.
Hickey's basic rejoinders set forth herein
are firstly that sociologists are technically
inadequate to the Hickey’s mathematical modeling,
secondly that they are ignorant of the contemporary
Pragmatist philosophy of science, and thirdly that
they reveal obstructionism.
Consider the sociologists’ technical
inadequacies. Before constructing his national
macrosociological theory with his METAMODEL
discovery system, Hickey undertook an extensive
search of the academic sociological literature to
determine what factors should govern his selection
of the manifestly sociologically relevant time
series data as inputs to the discovery system.
He also wanted to find some example of the
writing style used in sociology for reporting
findings from such modeling analyses.
In his literature search he could find no
precedent for his dynamic macrosociometric model.
Empirical work in sociology consists
exclusively of survey research using written
questionnaires and/or oral interviews.
And the purpose of the surveys is to examine
social-psychological hypotheses.
Furthermore the survey research findings are
summarized in tables, but are not analyzed with any
statistically estimated models.
One consequence of this condition is that any
sociologist selected by an editor to be a critic
could not reference any previously published
equation that is empirically superior to any
equations explaining the time series data used by
Hickey.
A second consequence of the unprecedented
character of Hickey's macrosociometric model is that
it reveals that academic sociologists are not
educationally prepared to work with higher-order
difference equation systems, such as those
constituting Hickey's model.
Hickey's professional education is in
economics, and since the publication in 1939 of
"Interactions between the Multiplier Analysis
and the Principles of Acceleration" in Review
of Economics and Statistics by Nobel laureate
economist Paul Samuelson, such difference equations
have become a staple technique in mathematical
economics and econometrics.
And Hickey’s use of the technique of shock
simulations was introduced into economics in 1933 by
the University of Oslo, Norway, economist Ragnar
Frisch in his
"Propagation Problems and Impulse
Problems in Dynamic Economics" in Economic
Essays in Honor of Gustav Cassel.
Ironically Hickey’s macrosociometric model
is not a simultaneous-equation model, and any
reasonably bright high school student could
replicate Hickey’s findings using Hickey’s model
with index numbers having any arbitrary but uniform
base year and by iterating the model in a computer
spreadsheet program.
And any undergraduate who was sufficiently
motivated to search back issues of the U.S.
Statistical Abstract in a public library or a
college library, could replicate the estimation of
Hickey’s model in a computer spread sheet’s
multiple regression routine.
But these techniques are not taught in the
curriculum of the Ph.D. sociologists. Thus the
referees were suspicious and dismayed by the
findings drawn from the simulation and shock
analyses in Hickey’s paper.
The outcome was that the sociologists deemed
by the editors to be sufficiently reputable as to be
worthy to function as referees for his journal,
showed themselves to be incompetent in the formal
techniques in Hickey's paper.
And it may be added that the editors who
rejected Hickey’s paper gave no evidence that they
are any less technically ignorant.
Hickey comments that it may be human to
reject what one does not understand, but it is not
professional.
Consider next the sociologists’
philosophical inadequacies.
When people do not know what to do, they do
what they know, whatever it may be; and what the
critics of Hickey's paper know is a reductionist
social-psychological Romanticism, which even today
still requires verstehen criticism.
The referees selected by the editors to whom
Hickey sent his paper did not announce explicitly
that they practice verstehen
verification.
But just as critics of papers in mature
sciences do not announce that they practice
empirical criticism, so too sociologists simply go
about practicing criticism unreflectively according
to the institutionalized value standards that they
had learned in their educational experience and that
are approved by their colleagues.
These editor-selected sociologists used
language that makes apparent their verstehen practice by the rhetoric and vocabulary in the stated
reasons they set forth as criticisms.
They criticized Hickey's equations because
they do not "make sense", because they are
"counterintuitive",
"meaningless", "bizarre",
"surprising", etc. thereby making apparent
their practice
of verstehen
criticism.
Sociologists may use the verstehen
for a priori criticism either before or after testing.
It operates before testing when it controls
discovery.
The sociologist empathetically formulates on
the basis of his own personal reality a hypothesis
about the mental states that motivate the social
actors' behavior that he may or may not intend to
investigate by survey research, with the result that
hypotheses that do not satisfy the a
priori verstehen
criterion are excluded from consideration for
empirical testing.
And verstehen
operates after testing when the criticizing
sociologist is confronted with a report of findings
from another sociologist's empirical work, and when
the report leads the criticizing sociologist to
reject out of hand an unexpected but empirically
valid finding with which he cannot empathize on the
basis of his personal or vicarious experience.
Consider finally the consequent sociological
obstructionism.
Sociologists like to view themselves as the
professional experts in all matters sociological.
As experts they earn their livelihoods as
academic professors of their subject in recognized
universities, award the universities’ credentials,
and condescendingly deem all others who might
discourse on the subject to be laymen and amateurs
who lack the academic instruction that professional
sociologists market. Therefore should the submitting
layman employ mathematical and statistical
techniques in which the academic sociologists are
incompetent, then the submission is viewed as an
embarrassing expose of the professionals’
inferiority. Consequently the submission of such a
paper by an outsider like Hickey, a philosopher and
econometrician, is not welcomed by the academic
sociological journals.
Were sociology a modern science with
institutionalized empiricist value standards, these
editors and their referees would have damaged their
professional reputations by dismissing Hickey’s
paper.
As it happens in the same year that Hickey
began submitting his paper to these sociological
journals, the editor of the Journal
of the American Society of Information Science (Jan.
1978, Vol. 29, No. 1, p. 3.) stated in his
"Editor's Notes" that referees sometimes
use the peer review process as a means of attacking
a point of view, and object to the content of a
submitted paper.
He said that often rather than rejecting a
paper so treated, it would be better to publish the
submitted paper with the reviewer's comments.
In his autobiography, Work
and Academic Politics (2002), William H. Form
who was the American
Sociological Review editor to whom Hickey had
submitted his paper, portrays academic sociology as
a mediaeval guild and himself as a journeyman in the
guild. A guild is a kind of trade association of
craftsmen or merchants that flourished in Europe
between the 11th and 16th centuries, and that was
formed for the mutual aid and advancement of its
members by monopolizing its trade.
Based on the attempted criticisms by
editor-selected sample of referees Hickey concludes
that American academic sociology is much worse than
a mediaeval guild operating in restraint of trade;
he believes that the philosophy of science enforced
in American academic sociology is so inbred that its
information pool is as degenerate as the gene pool
of an incestuous hereditary dynasty.
Sociological
Methods and Research
The first academic sociological journal to
which Hickey sent his paper was Sociological
Methods and Research published by Sage
Publications, Inc.
This journal did not acknowledge receipt of
the paper, but Hickey's retained U.S. Postal Service
receipt documents that the paper was received on 18
December 1978, the date that Hickey uses to document
his priority, although in fact his macrosociological
model was actually created in the latter half of
1976.
On 22 May 1979 Hickey received a letter
rejecting the paper for publication from the editor,
a Mr. George W. Bohrnstedt of Indiana University.
With the letter were enclosed criticisms by
two referees, both of whom offered a recitation of
their Romanticist philosophy of science.
The first referee described Hickey's model as
a reification of the worst type, and ridiculed
Hickey's artificial-intelligence discovery system as
a self-cooking program.
This rhetoric is symptomatic of Romanticism
and also suggests a Luddite mentality.
This critic also described Hickey's model as
value-based modeling, and said that it is inferior
to a demographic accounting framework advocated by a
sociologist named Kenneth C. Land.
Land had proposed a modeling approach in his
"A General Framework for Building Dynamic
Social Indicator Models: Including an Analysis of
Changes in Crime Rates and Police Expenditures"
in American
Journal of Sociology (1976).
Land’s ideas have their origin in a 1971
technical report, Demographic Accounting and Model Building, written by a Professor
Richard Stone, and published by the Organization for
Economic Cooperation and Development. Conceptually
the demographic accounting system is analogous to a
perpetual inventory accounting system as might be
found in a retailing business; it has beginning and
ending inventory stocks, and inflows and outflows
explaining the changes in stocks over an accounting
period.
In the demographic system the stock variables
represent population head counts with the inflows
due to births or immigration and the outflows due to
deaths and emigration. Stone also describes how the
accounting system may be expressed as a matrix with
the inflows and outflows treated analogously to the
economist's input-output models with the rows
representing inflows, the columns representing
outflows, and the cells representing transition
coefficients calculated by dividing the aggregates
in each row by its row total.
Since these transition coefficients will
change from period to period, there is an additional
problem of projecting these changes if the model is
to be used for any kind of policy analysis.
Land's paper proposed using the econometric
type of models statistically estimated over the time
series of transition coefficients, which he
furthermore says can be interpreted as measures of
opportunities for social benefits.
He therefore calls this the
opportunity-structure approach, which he says is
based on ideas originally proposed by the
sociologist William F. Ogburn.
Land's approach seems interesting and might
be fruitful, if and when it is ever carried out.
However the equation set forth in his paper
is not estimated over vectors of transition
coefficients from any demographic model.
In any event it is not clearly an alternative
to Hickey's, since his value-based approach might be
used to model the changes in the transition
coefficients.
But Hickey is not indebted to this approach,
and he was unwilling to be conscripted to the
support of this agenda as a condition for
publication.
In fact he referenced it in future versions
of his paper to distinguish his work. The second
referee selected by Bohrnstedt started his criticism
by saying that he can't quite figure out whether or
not the paper is a "put on”.
Hickey decided that Sociological
Methods and Research is not a suitable
publication for his paper, because he concluded that
the model is beyond the competence of the editor and
his selected referees, and he did not send the
editor any replies.
He did not know at the time that referees for
other sociological journals would offer even more
dogmatically Romantic criticisms.
Nor did he know at the time that Bohrnstedt
is philosophically opposed to the contemporary
Pragmatist philosophy of science until Bohrnstedt,
Knoke and (later) Mee later published an
undergraduate textbook titled Statistics
for Social Data Analysis, which virtually echoed
the philosophy of science expressed in
Bohrnstedt’s selected referees.
The authors would limit modeling to a testing
role, and advocate a version of Haavelmo’s
structural-equation agenda with its romantic
ontology. Like Haavelmo they distinguish unobserved
conceptual variables and observable indicators, a
gratuitously equivocating semantical dualism.
And they propose criteria for identifying
causality prior to statistical modeling and
empirical testing.
|
|
