Citation
The configuration of interest groups influential in international population policy during UNCED, 1990-1992

Material Information

Title:
The configuration of interest groups influential in international population policy during UNCED, 1990-1992
Creator:
Campbell, Martha Madison
Publication Date:
Language:
English
Physical Description:
ix, 271 leaves : ; 29 cm

Subjects

Subjects / Keywords:
Population policy ( lcsh )
Pressure groups ( lcsh )
Population policy ( fast )
Pressure groups ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 249-271).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Doctor of Philosophy, Department of Public Administration
Statement of Responsibility:
by Ellen O'Connor Dickson.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
32077763 ( OCLC )
ocm32077763
Classification:
LD1190.P86 1994d .C36 ( lcc )

Full Text
THE CONFIGURATION OF INTEREST GROUPS
INFLUENTIAL IN INTERNATIONAL POPULATION POLICY
DURING UNCED, 1990-1992
by
Martha Madison Campbell
B.A., Wellesley College, 1963
M.A., University of Colorado 1989
A thesis submitted to the Graduate School of the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Administration
1994


Campbell, Martha Madison (Ph.D., Public Administration)
The Configuration of Interest Groups Influential in International Population
Policy during UNCED, 1990-1992
Thesis directed by Professor Robert W. Gage
ABSTRACT
In this thesis, epistemic community theory by Peter Haas is examined
as a framework for understanding the configuration of competing interest groups
during the United Nations Conference on Environment and Development
(UNCED), January 1990 to June 1992. An epistemic community (EC),
according to Haas, is a group possessing authoritative expertise in a policy issue-
area, to which policy-makers turn for advice in the presence of uncertainty. An
EC exhibits shared interests (policy enterprise), causal beliefs, principled
beliefs, and validity criteria (the latter internally validated within the group). Its
members are not bound by any organizational relationship. For Haas, interest
groups have shared interests and principled beliefs but not shared causal beliefs.
Five separate groups are identified as influential in population policy
during UNCED: population growth-concerned people (POP), those supporting
private enterprise and market systems (MKT), developing country voices
preferring to focus on poverty and inequitable distribution of resources rather
than on population growth (DST), organized women promoting a population
focus on reproductive rights and health rather than on reducing population
IV


growth (WIN), and the Vatican (VTC). These groups are widely diverse in
organizational as well as philosophical terms.
The primary analysis tests the groups for shared causal beliefs, using
non-quantified content analysis of publications and speeches. All show shared
causal beliefs. Only DST, with scant literature on population, is judged to lack
the required expertise and not qualify as an EC. A second analysis reveals
asymmetry among the groups, defined by the author. Four amendments to the
Haas theory are recomihended. 1) Complex, multi-cultural, value-laden policy
issue-areas can have multiple ECs, including some from outside the field who
redefine the issue-area according to their own perspectives, and thus must then
be counted as experts in the field. 2) Professional pedigrees, seen by Haas as
distinguishing ECs from other groups, are not necessary, as demonstrated by
WIN. 3) Interest groups may have shared causal beliefs but lack expertise of
ECs. 4) Faced with anomalies undermining their beliefs, members of ECs do
not necessarily withdraw from the debate as Haas has suggested. Rather, they
tend to reject dissonant information in ways observed by cognitive theorists cited
in the research.
This abstract accurately represents the content of the candidate's thesis. I
recommend its publication.
Signed
V


This thesis is dedicated to my father,
Kenneth Campbell,
and to the memory of my mother,
Margaret Macon Campbell.


ACKNOWLEDGMENTS
My grateful thanks are due to my thesis committee for their hard
work and support throughout this project. In particular, I want to thank Bob
Gage for his steady stream of solid advice and timely caveats in my design and
completion of this project; Peter deLeon for his creative ideas in theoretical
aspects of this work, especially in directing me to critically important literature
in several directions; Lloyd Burton for his insights as an environmental
specialist, and also for his suggestions on the use of the central theoretical
literature; political scientist Steve Thomas for advising me early in this process
to stay narrow and straight in my approach to a subject that had potential for
being dangerously broad, and for his continuing support through both this
project and my earlier political science master's thesis on Mexico's population
policy reversal; and John Sbarbaro, M.D., for recognizing the health
implications of this research, welcoming it into his Colorado Prevention Center,
and assisting in obtaining the foundation support for the project.
Special thanks go to Lisa Waugh for her insightful analysis of data
from UNCED country reports, and to both Lisa Waugh and Barbara Bailey for
their hard work in a test of intercoder reliability of the analysis.
I am grateful also to the officers of the Rockefeller Foundation and
the Compton Foundation for supporting the research for this thesis.
Vll


CONTENTS
CHAPTERS
I: Introduction..........................................................1
iff Literature Review....................................................14
III: The International Population Policy Arena: Laboratory for the Study.58
IV: Designing the Analysis...............................................84
V: Outcomes of Stage 1 of the Analysis...................................141
VI: A Second Look, Through a New Framework...............................198
VII: Conclusions.........................................................210
APPENDIX
A: Agenda 21 Chapter 5, the population chapter (11 pages)................229
B: Women's Alternate Agenda..............................................240
C: Agenda 21 Substantive Categories......................................242
D: Early Subjects List...................................................243
E1-E5: Early Lists of Causal Beliefs.....................................244
BIBLIOGRAPHY................................................................249
VUl


EXHIBITS
Exhibits:
1: Distinguishing Epistemic Communities from Other Groups...................21
2a: Coding Frame..............................................................104
2b: Subjects List.............................................................107
3: Star Model of Schools of Thought on Population............................113
4a: Matrix: POP's Causal Beliefs..............................................146
4b: Matrix: POP's Principled Beliefs..........................................147
4c: Matrix: POP's Data Sources................................................148
5a: Matrix: MKT's Causal Beliefs..............................................155
5b: Matrix: MKT's Principled Beliefs..........................................156
5c: Matrix: MKT's Data Sources................................................157
6a: Matrix: DSTs Causal Beliefs...............................................163
6b: Matrix: DST's Principled Beliefs..........................................164
6c: Matrix: DST's Data Sources................................................165
7a: Matrix: WIN's Causal Beliefs..............................................171
7b: Matrix: WIN's Principled Beliefs..........................................172
7c: Matrix: WIN's Data Sources............................................... 173
8a: Matrix: VTC's Causal Beliefs..............................................182
8b: Matrix: VTC's Principled Beliefs..........................................183
8c: Matrix: VTC's Data Sources................................................184
9: Stage 1, Causal Beliefs and Principled Beliefs............................190
10: Transition from Stage 1 to Stage 2 of the Analysis........................201
11: Stage 2, Positions of the Five Groups.....................................202
12: Model: Patterns of Acknowledged Information...............................209
13: An Epistemic Community Model..............................................216
IX


CHAPTER I
INTRODUCTION
This thesis examines a theory of epistemic communities that Peter M.
Haas has presented in the context of international cooperation (1992b). An
epistemic community is defined as a knowledge-based group recognized for its
expertise in a given domain or policy issue-area. Haas suggests that decision
makers turn to these communities in policy situations characterized by
uncertainty, and that such communities can have substantial influence on a
policy outcome. Members share their policy interests through their transnational
professional networks, rather than within strict organizational structures, and can
be effective in developing consensus on complex issues within their respective
governments.
The theory's emphasis on seeing the epistemic communities as
exhibiting far more in common than political interests suggests that it might be a
valuable analytical tool for gaining understanding of the complicated
international population policy arena. This policy issue-area is a chronically
contentious one in which there are several widely different schools of thought
about what would appropriately constitute an international population policy.
It is on the differences between the interest groups influential in the
international population arena that this exploratory research will focus, using as
a framework Haas' concept of epistemic communities. In particular, the
l


research will examine the competing schools of thought influential at the time of
the United Nations Conference on Environment and Development (UNCED),
the world conference that began in January 1990 and culminated in the Earth
Summit and the Global Forum in Rio de Janeiro in June 1992.
"Rush off to Rio de Janeiro, talk about the future of the world, no
less, and rush home again: that is the lot of many world leaders next week." So
began a description of the Earth Summit in the Economist's editorial as the Rio
conference began. Those editors were bemoaning the failure of UNCED to
highlight the importance of population growth in the global environmental
challenges it planned to face, and suggested that excitement over the new global
warming treaty that would be signed was the reason for that failure. "[GJlobal
warming is fashionable, and tantalising...." (The question Rio forgets 1992, 11).
This research in conjunction with Haas' theory will attempt to show
that, rather than the enthusiasm over the global warming treaty, the nature of the
differences among the competing interest groups involved in population policy
can reasonably be seen as responsible for the failure of UNCED to present
global population growth persuasively as a critical obstacle to avoiding
environmental damage.
In this author's observation of the UNCED policy process, it
appeared that the groups of people representing very different ways of thinking
about population^ were exhibiting not only different sets of interests, but also
1 The term "population" is being used narrowly in this thesis. Studies of population (broad sense) are
generally conducted in one of four areas: fertility, mortality, migration or urbanization. In this present
analysis, population, except where otherwise indicated, will refer to population growth and fertility issues, and
in particular the relatively rapid or persistent population growth occurring in many of the world's developing
2


different sets of beliefs in terms of what was accepted as fact. Further, then-
ways of thinking about the population subject appeared to be characterized by
not only contrasting approaches to the same information, but also, to a
noticeable degree, attention to different sets of information. It seemed
incongruous to see intelligent people exhibiting what appeared to be self-limiting
mechanisms on acceptance of information. In short, each interest group not
only exhibited a firm objective, but also seemed to be revealing a body of
knowledge not consistent with the other groups' expressed knowledge.
Haas views an epistemic community as having an impact on the
policy process when, through the influence of its network of experts, policy
makers come to adopt its information as consensual knowledge. He indicates
that an epistemic community's influence on international policy development
takes place insofar as it "consolidates bureaucratic power within national
administrations and international secretariats" (Haas 1992b, 4). He explains this
process in terms of the diffusion of the knowledge of the epistemic community's
experts to the policy makers, possibly resulting in "the creation of toe proper
construction of reality with respect to toe particular issue-area...." (1992b, 27;
emphasis by Haas).
This exploratory research will seek to determine which of toe
competing groups might reasonably be seen as epistemic communities, each with
its distinct form of authoritative expertise on toe subject of population. This
countries. The justification for using this narrowed definition during this research is that human fertility is the
factor that drives most international population policy formulation, and fertility issues tend to be the bones of
contention in the population policy arena. The word "population" has commonly been used in this narrowed
sense in the United Nations, including the UNCED process.
3


research is expected to support Haas' view of the conditions that allow for an
epistemic community's influence on international policy development, by
demonstrating that several active, influential epistemic communities were
attached to the population issue-area, each representing its own version of
expertise for that issue-area; and that the involvement of these competing groups
could be expected to make it difficult for any one of them to consolidate power
in the international secretariat, or in the case of UNCED, the delegates of the
178 countries represented.
The first of the competing interest groups or schools of thought in the
policy formulation process during UNCED was a set of people concerned about
population growth. The second was another large set of people who strived to
keep the population subject off the UNCED agenda in favor of attention to
international equity issues and attention to the impact of profligate Northern
consumption patterns on the environment. The third group was a newly visible
team of organized feminists decrying attention to population growth and birth
rates. The fourth group was the Vatican. In addition, with President Bush still
in office, a fifth group representing a preference for market systems was
influential in international population policy during the UNCED period, but not
in the conference process itself. The latter four of these five groups all resisted
attention to population growth, all for different reasons, and they represented
clearly visible, disparate interests. Later in the dissertation, they will be seen, in
James Sebenius' terms, as blocking coalitions, or as, unintentionally, a single
blocking coalition. Not only did these four groups have competing objectives,
4


but also they appeared to define the population subject in entirely different ways.
Given this divergence, along with an absence of pressure to compromise their
positions (for reasons to be explained in Chapter ID), it is understandable that in
UNCED no one epistemic community's knowledge could become the consensual
knowledge in the policy development process. Indeed, the resulting written
agreement on population that was produced in the UNCED process was a
relatively weak statement, and was distinctly additive in form in an attempt to
satisfy the requirements of all of the groups' positions, which remained
essentially uncompromised throughout the process.
The kind of expertise to which Haas refers throughout his description
of epistemic communities tends to be technical in nature. He states that his
theory applies to social policy as well as the natural sciences, but in spite of this,
his policy examples are relatively technical ones. Because of this, there are no
examples given in which the meaning of the policy issue-area itself has been
disputed, or the problems to be addressed by policy are defined in radically
different frameworks by groups exhibiting very different world views. Further,
Haas gives no indication of any awareness of this possibility. This dissertation
will argue that, while Haas' concept of epistemic community appears to be of
value in technical issue-areas, it may have, although possibly unrecognized by
Haas, even greater potential as an analytic tool for increasing understanding of
the complex population policy arena, and possibly also other social, complex,
multi-sided, international, multi-cultural policy debates as well.
5


Haas provides a way to distinguish between epistemic communities
and other identifiable collections of people such as interest groups, social
movements, legislators, bureaucratic agencies, bureaucratic coalitions,
disciplines and professions. He sees epistemic communities as the only type of
group contributing expertise for an issue-area and at the same time exhibiting
shared causal and principled beliefs, shared validity criteria for the acceptance
or rejection of information, and a common policy enterprise. Using Haas' own
defining criteria as a starting point, this research will explore the characteristics
of the groups active in the population policy arena to test their qualifications as
epistemic communities. The reason for doing this is that if these groups are
epistemic communities, the population policy arena is exhibiting something Haas
has not, evidently, anticipated: epistemic communities whose expertise comes
from outside the policy issue-area, who resist accepting the traditional meanings
of that issue-area, and who, in effect, redefine the issue-area in ways consistent
with their own interests. This research will provide a view of the configuration
of the five competing groups. In the process, it is expected that it will become
evident that Haas' attempt at distinguishing between interest groups and
epistemic communities is, at least in this policy issue-area, unwarranted.
Chapter n, the literature review, will be divided into two sections.
In the first, it will explore in depth Haas' concept of epistemic community, and
review other authors' applications and critiques of this theory. Among these
will be the position of James Sebenius on the significance of Haas' epistemic
community concept with respect to Sebenius' own negotiation analysis. In
6


particular, for Sebenius, the epistemic community theory has contributed to the
task of conflict resolution an expanded concept of negotiation. That expanded
concept of negotiation will later be pertinent in the design of Stage 2 of the
analysis.
This first section also will introduce literature by others offering
some commentary on the epistemic community theory. It will be seen that most
of this responding literature has concerned itself with longitudinal processes of
influence by groups on the policy process, rather than on specific qualifications
of those groups or on the shape of the theory itself.
The second section of Chapter II will explore related areas of theory.
Although this research is about international policy development, it will not
focus on power structures, or on transnational institutions and relationships as
presented by Robert Keohane and Richard Nye (1971). It is instead being
undertaken in the spirit of Harold Lass well's multidisciplinary approach to the
policy sciences (deLeon 1988), including attention to both psychological aspects
and policy content of the policy development process. And, although it is about
knowledge, it will not be focused on the sociology of knowledge explored in
depth by Richard Bernstein (1983). Rather, because Haas views knowledge as
accepted belief and focuses on beliefs in groups, the theory underlying this
exploration will lean toward those cognitive theorists who explore the acquisition
of beliefs and the nature of belief.
Looking first at structural issues, because Haas distinguishes interest
groups from epistemic communities, some basic literature on interest groups will
7


be introduced, including contributions from Harry Eckstein, Joseph
LaPalombara, David Truman, and Terry Moe. In connection with this, varying
meanings of the term "interests" in the literature will be noted. Behavioral
issues will then be addressed, with reference to literature in cognitive and social
psychology. Theories of belief formation will be introduced, including the work
of T. M. Newcomb, James Rosenau and Ole Holsti, Raymond Boudon,
Alexander George, James March, Herbert Simon, Jonathan Freedman, David
Sears, Curt Rausch, and John Odell, along with generalized insights of Irving
Janis and Robert Jervis. Next, ideas presented by Janis and Jervis will be
reviewed among those who apply theories of belief formation to a public policy
context, along with Hank Jenkins-Smith, James Clay Moltz, Irving Janis, and
Paul Sabatier. A final group of theorists includes five authors who incorporate a
heavy philosophical component into their work: Elinor Ostrom, Robert
Grafstein, Aaron Wildavsky, Charles Lindblom, and Jon Elster. Support for
this group's thinking from Robert Rothstein and Irving Janis will be noted, and
Janet Weiss' connection between beliefs and problem definition in a policy
context will be introduced.
Chapter II will include a brief review of a fundamental debate in
epistemology that is relevant to this research, between relative and positivist
concepts of knowledge. Kenneth Boulding's faith in the "outability the truth"
will be compared with Haas' confidence in knowledge convergence, and the
section with conclude with Donald Campbell's distinction between scientific and
social beliefs.
8


Literature on the value of analyzing the rhetoric that influences
policy development will be introduced, new material compiled by Frank Fischer
and John Forrester along with important contributions by the editors themselves.
This will be presented as an alternate way to learn about the competing groups
in complex policy arenas. This literature on discourse analysis and discourse
coalitions views the stated positions by competing policy groups as claims rather
than as knowledge. Serving as an implicit challenge to the epistemic community
theory, the discourse coalition literature presents both opportunities and
problems in the examination of the competing groups. This dissertation will
suggest that the contrast between the two theories raises questions about the
roles of deception and self-deception in the policy process, and has interesting
implications for validity.
Chapter in will present the policy context, the population policy
situation in UNCED. The origins of the UNCED conference will be described,
along with the expectations of the conference leaders for its contributions. The
disagreements about the appropriateness of including population on the agenda
will be described. Chapter El will also discuss characteristics of the population
subject that make it inherently repellent and sometimes difficult to address with
equanimity in the policy community.
Chapter IV will develop the principal analysis of the dissertation,
Stage 1, based on Haas' concept of epistemic communities. Since the groups
have been identified during the UNCED policy process on the basis of then-
most highly visible policy interests, the analysis will focus on a search for
9


patterns of consistent causal and principled beliefs. This chapter will explain the
use of comparative analysis in this exploratory research, based on work by
Barney G. Glaser and Anselm L. Strauss. The extent to which this analysis
should and should not be viewed as content analysis will be discussed in light of
characteristics of the population policy arena. The analysis will be
operationalized, with discussion on the organization of data, selection of
substantive categories of subject matter for analysis, rules of analysis that have
been developed, and the selection of data sources. The criteria for identifying
the five schools of thought in the population policy arena will be presented,
along with descriptions of these groups in terms of their historical origins. A
new "star model" of the five identified schools of thought will be presented, a
simple visual model designed to account for groups' different areas of emphasis
about population and also variations in people's intensity of affiliation with any
one of the five schools.
Lastly, validity issues concerning this research will be explored, with
recognition that this qualitative research based on the analysis of political
messages contains built-in hazards in terms of validity. It is hoped that the
acknowledgment of these hazards will help in the identification of what useful
information we can somewhat confidently derive from this research, as well as
the information about which we will necessarily be less sure.
Chapter V will present the results of Stage 1 of the analysis, in terms
of identified common beliefs in the data sources representing each of the five
groups. Following the five sections of Stage 1, representing the five identified
to


groups, there will be a discussion of problems encountered in this process,
problems that can be seen as opportunities for further exploration.
Chapter VI will present Stage 2 of the analysis and its results. In
Stage 2, groups' messages identified in Stage 1 will be rearranged in ways
suggested by the problems and opportunities discovered in the Stage 1 process.
The objective will be to view more clearly the relationships among the five
groups' positions. Analytic categories derived from James Sebenius' negotiation
analysis will be employed, due to their value in terms of revealing important
relationships among competing groups. Other categories will be rearranged,
and a new analytic category will be added in response to the observed self-
limiting mechanisms of information acceptance in the UNCED policy process.
These results are expected to show that the configuration of group relationships
in the population policy field is not simple, for it has been known from the
beginning that the groups do not line up on any linear continuum, nor in any
simple pattern of opposition. An asymmetry among positions will be described,
and a conceptual description of this asymmetry in a more generic sense will be
offered for application to other policy issue-areas as well. In addition, a model
of "acknowledged information" will be offered in an attempt to describe visually
the observed patterns of acceptance of segments of the information and concerns
of some groups by other groups.
Chapter VII will discuss the findings of the research in terms of the
theoretical contributions by the variety of authors introduced in the literature
review, Chapter n. It will be suggested that, to the degree that self-limiting
ll


mechanisms of information acceptance have been observed in groups' positions,
this situation appears to fit theoretical explanations found in some of the
literature on cognitive psychology, as well as the theories offered by Ostrom,
Grafstein and Lindblom for the convergence of beliefs.
Based on the observed configuration of the players in the
international population policy arena, an epistemic community model of that
arena will be presented, to account for the patterns of knowledge and areas of
interest outside the issue-area exhibited by the groups, within the framework that
has been presented by Haas. This model has been developed to account for
different concepts of expertise in this one very contentious issue-area, but with
the expectation that such a model will apply to other similarly configured policy
issue-areas.
This research has both practical and theoretical objectives. On a
practical level, it is hoped that understanding the complex configuration of group
positions in the population policy arena during UNCED will foster constructive
dialogue among the active groups themselves, and provide a clearer
understanding of the nature of the recent, and to some extent current, debates in
the population field in preparation for the population policy development
challenges in the near future. In the extended preparations for the upcoming
International Conference on Population and Development (ICPD), the United
Nations conference scheduled to take place in Cairo in September 1994, it has
not been uncommon for observers of the ICPD process to find perplexing or
incomprehensible some of the disagreements and apparently inconsistent aims in
12


this policy field. This research is designed partly to reduce that bewilderment
by revealing important patterns, thus clarifying the objectives and stands of the
major, influential participating groups. A more precise understanding of their
differences and of the structural relations of their positions may serve as an aid
to identifying their often unrecognized areas of common interest.
On the theoretical level, it is hoped that this research will draw
increased attention to the potential of Haas' epistemic community theory as a
useful analytical tool in policy situations characterized by the simultaneous
influence of groups holding radically different views of an issue-area and
claiming to have the authoritative expertise about the policy subject. In a decade
when international disputes are shifting away from traditional bloc configu-
rations, and where negotiations often concern transnational matters not defined
by country borders, it is advantageous to develop better insights into how major
interest groups of widely diverse form and organization perceive their roles in
the policy process and interact with the other groups with whom they compete.
This research suggests that the epistemic community theory's attention to the
knowledge held by influential groups makes that theory, with some adjustments,
a promising framework with which to view the positions of competing groups
that exhibit different and sometimes incompatible sets of knowledge in complex
policy disputes.
13


CHAPTER H
LITERATURE REVIEW
Epistemic Community Theory
The theoretical basis for this research is the epistemic community
concept as presented by Peter Haas. According to Haas, an epistemic
community is a
...network of professionals with recognized expertise and
competence in a particular domain and an authoritative claim to
policy-relevant knowledge within that domain or issue-area
(1992b, 3).
The epistemic community concept has been introduced by Haas to
demonstrate the important role that can be played by expert or knowledge-based
groups in international policy development, particularly in those areas
characterized by a high degree of uncertainty. They are seen as providing
assistance to policymakers in a number of ways, including
...articulating the cause-and-effect relationships of complex
problems, helping states identify their interests, framing the issues
for collective debate, proposing specific policies, and identifying
salient points for negotiation (1992b, 2).
He suggests that an epistemic community is influential in international policy
insofar as it aids in the development of consensual knowledge within the policy
community. Although he says that epistemic communities need not be
14


transnational, Haas is particularly interested in their potential for influence on
international policy development.
To the extent to which an epistemic community consolidates
bureaucratic power within national administrations and
international secretariats, it stands to institutionalize its influence
and insinuate its views into broader international politics (1992b,
4).
It is specifically this idea that serves as the theoretical basis for this study,
because development of population policy in UNCED was not characterized by
consolidation of power by any one epistemic community. Instead, several
disparate views about population remained influential throughout the process.
This research is expected to demonstrate that the nature of the disparity among
the groups' positions, a type of disparity evidently unanticipated bv Haas, made
it unlikely that consensual knowledge could develop in UNCED on the subject
of population, and that the configuration of interests in this policy case has
implications both for population policy development in the future and for
expanded application of Haas' epistemic community theory.
An important characteristic of most epistemic comunities is that they
are normally independent from formal organizational structure. The lack of a
requirement for organizational boundaries is evident in Haas' description of how
epistemic communities work in instances of international cooperation. They
.. .become transnational over time as a result of the diffusion of
community ideas through conferences, journals, research
collaboration, and variety of informal communications and
contacts (1992b, 17).
15


The community is seen here as possibly extending far beyond the boundaries of
any organization.
Haas describes the process of influencing the policy community with
reference to the diffusion of the knowledge of the epistemic community's experts
to the policy makers.
It is the political infiltration of an epistemic community into
governing institutions which lays the groundwork for a broader
acceptance of the community's beliefs and ideas about the proper
construction of reality. The result in turn may be the creation of
the proper construction of reality with respect to the particular
issue-area....(1992b, 27; emphasis by Haas).
He explains the influence of epistemic communities in international policy
development through their role of informing decision makers.
Members of transnational epistemic communities can influence
state [country] interests either by directly identifying them for
decisions makers or by illuminating the salient dimensions of an
issue from which the decision makers may deduce their interests.
The decision makers in one state may, in mm, influence the
interests and behavior of other states, thereby increasing the
likelihood of convergent state behavior and international policy
coordination, informed by the causal beliefs and policy
preferences of the epistemic community. Similarly, epistemic
communities may contribute to the creation and maintenance of
social institutions that guide international behavior. As a
consequence of the continued influence of these institutions,
established patterns of cooperation in a given issue-area may
persist even though systemic power concentrations may no longer
be sufficient to compel countries to coordinate their behavior
(1992b, 4).
16


He distinguishes members of an epistemic community from "other
social actors and groups" by their special qualifications:
The epistemic community members' professional training,
prestige, and reputation for expertise in an area highly valued by
society or [by] elite decision makers accord them access to the
political system and legitimize or authorize their activities.
Similarly, their claims to knowledge, supported by tests of
validity, accord them influence over policy debates and serve as
their primary social power resource. At the same time, the
professional pedigrees and validity tests set the community
members apart from other social actors or groups and not only
serve as a barrier to their entry into the community but also limit
the influence that these other actors or groups might have in the
policy debate (1992b,17-18).
Haas is careful to extend the epistemic community concept to fields
beyond the natural or "hard" sciences:
Epistemic communities need not be made up of natural scientists;
they can consist of social scientists or individuals from any
discipline or profession who have a sufficiently strong claim to a
body of knowledge that is valued by society. Nor need an
epistemic community's causal beliefs and notions of validity be
based on the methodology employed in the natural sciences; they
can originate from shared knowledge about the nature of social or
other processes, based on analytic methods or techniques deemed
appropriate to the disciplines or professions they pursue (1992b,
16).
He then gives examples of these other fields,
...communities involved in other efforts ...[with]... expertise
related to disciplines and professions such as economics and
engineering (1992b, 17).
17


Haas' view of knowledge is central to his concept of epistemic
communities. In the passage below, he states that "...knowledge is only
accepted belief," and yet he treats knowledge development as a convergence
toward consensus. This passage follows one in which he rejects relativism that
is based on the idea that, given the social construction of language, there is no
objective basis for observing material reality (1992b, 21-23):
Alternatively, those with a more essentialist or materialist view
argue that the world is a real and separate object of inquiry that
exists independently of the analyst and that although the categories
in which it is identified are socially constructed, consensus about
the nature of the world is possible in the long run. This limited
constructivist view informs the analyses presented by most of the
authors contributing to this volume. It also has implications for
evaluating the validity of a given body of knowledge, pointing to
the need for a consensus theory of a finite and temporally bounded
notion of truth, rather than a correspondence theory. Although
knowledge is only accepted belief, not correct belief, correct
beliefs may evolve over time, as progressively more accurate
characterizations of the world are consensually formulated. By
reference to internally formulated truth tests, contending groups
may collectively validate their conclusions and their beliefs may
converge intersubjectively in the medium run....The epistemic
communities approach focuses on this process through which
consensus is reached within a given domain of expertise and
through which the consensual knowledge is diffused to and carried
forward by other actors (1992b, 23).
His faith in the convergence of knowledge is demonstrated further in the
following passage:
18


If the consensus theory of truth is valid, then the fundamental
distortions to collective understanding may over time be reduced
or at least be detectable by informed study (1992b, 25).2
Epistemic Communities' Defining Characteristics
In addition to authoritative knowledge, or expertise, members of an
epistemic community, according to Haas, exhibit shared principled beliefs,
causal beliefs, validity criteria, and a policy enterprise. These four hallmarks
are defined by the author as follows (1992b, 3):
Common tor shared! policy enterprise.
...a set of common practices associated with a set of problems to
which their professional competence is directed, presumably out
of the conviction that human welfare will be enhanced as a
consequence (1992b, 3).
In his diagram (Exhibit 1), Haas substitutes the term "interests" for this shared
policy enterprise, and it is by the elements of interest most visible in the
UNCED policy process that the influential groups have been identified for this
research. It is important to note that, by this substitution in the diagram, Haas
has included atruistic objectives in the concept of "interest".
Shared causal beliefs.
...derived from their analysis of practices leading or contributing
to a central set of problems in their domain and which then serve
2 Haas' confidence in this convergence is similar to that of John W. Kingdon, who notes that
"specialists [in a given issue-area] do eventually see the world in similar ways, and approve or
disapprove of similar approaches to problems (1984, 140).
19


as the basis for elucidating the multiple linkages between possible
pobcy actions and desired outcomes (1992b, 3).
Principled beliefs.
...a shared set of normative and principled beliefs, which provide
a value-based rationale for the social action of community
members (1992b, 3).3
Validity criteria.
...shared notions of vahdity-that is, intersubjective, internally
defined criteria for weighing and validating knowledge in the
domain of their expertise" (1992b, 3).
Haas also speaks of the "unspoken societal assumptions experts transmit"
(1992b, 25). It appears that these assumptions constitute the validity criteria, the
common knowledge or experience shared by a group that serves in a validating
capacity in the acquisition of new beliefs. In his diagram (Figure 1), Haas
substitutes "knowledge base" for validity criteria, presumably referring to the
specialized knowledge shared by the group that is not necessarily related to the
pobcy issue-area and functions as a screen for new information in the issue-area.
In a footnote, Haas exphcitly differentiates between intercommunity
reahty tests, criteria forjudging the vahdity of information that are
acknowledged across groups, and intracommunity reahty tests, the criteria
appbed only within a single group. For his view of epistemic communities he
accepts the latter, the intra-group vahdity criteria. Acknowledging that these
3 Judith Goldstein and Robert Keohane provide a similar definition of principled beliefs, "ideas that
specify criteria for determining whether actions are right or wrong and whether outcomes are just or
unjust" (Ideas and Foreign Policy: Beliefs. Institutions and Political Change. Ithaca: Cornell
University Press, 1994, cited in Sikkink 1993: 412, note 2).
20


Exhibit 1
Distinguishing Epistemic Communities from Other Groups*
Causal beliefs
Shared Unshared
o u Epistemic Communities Interest groups and social
1 on movements
"O p Disciplines and Professions Legislators, bureaucratic
M agencies, and bureaucratic
s D coalitions
Knowledge base
Consensual Disputed or absent
"O Epistemic Communities Interest groups and social
a e/3 movements and bureaucratic
coalitions
o £
3 D agencies
*Source: Haas 1992b, International Organization 46:1, p. 18; reproduced with permission
from the MIT Press.
Note by author of thesis: The term "interests" in this diagram by Haas refers to "policy
enterprise" in his accompanying text, and the term "knowledge base" in this diagram
refers to "validity criteria" in his text. See pages 19 and 20 of this thesis.
21


internal truth tests are less ambitious, Haas states that he is skeptical about
universal validity claims in the broader policy community (Haas, 1992b: 17).4
Haas reviews other kinds of groups and asserts that they do not have
all of these shared characteristics. Disciplines and professions, for example, are
seen as having shared causal beliefs and knowledge bases (or validity criteria),
but not shared principled beliefs or interests. Bureaucratic agencies are seen as
focusing on preserving their missions and budgets. Interest groups are seen as
having shared principled beliefs and interests, but not shared causal beliefs.
Epistemic communities "tend to pursue activities that closely reflect the
community's principled beliefs," and its ethical standards are derived from "its
principled approach to the issue at hand," rather than, for example, a
professional code. Because of their special combinations of attributes, according
to Haas, an epistemic community, "if confronted with anomalies that
undermined their causal beliefs ... would withdraw from the policy debate,
unlike interest groups" (1992b, 18-19). There are indications that Haas may see
interest groups in a relatively narrow light, as motivated, for example, by
economically concerns.
4 Haas claims that the intracommunity validity criteria "contrasts with Ernst Haas's usage of the
concept of epistemic communities, in which he explicitly mentions that such communities 'profess
beliefs in extracommunity reality tests.'" His reference is to Ernst Haas' When Knowledge is
Power (1990) (Peter Haas 1992b, 17, note 39). Actually, from this author's reading of Ernst
Haas, it appears that the word "profess" is critical; Ernst Haas does not himself seem to
acknowledge or require the universal recognition of single groups' validity criteria. Instead, he
appears to be merely observing groups' beliefs that their validity criteria should be accepted by
other groups. There may be, then, less disagreement on this point between E. Haas and P. Haas
than the latter suggests. All further mention of "Haas" in this dissertation, except where
specifically labeled as "Ernst Haas", will refer to Peter Haas.
22


The Paradox
Haas' theory contains a paradox in the following combination of
elements: his faith in knowledge convergence between groups over time; his
view that knowledge is only accepted belief; and his view that groups' authority
in an issue-area is acknowledged or validated only internally. Implicit in this
combination of specifications for epistemic community influence in the policy
community is an assumption that, if they become better informed, competing
groups in an issue-area will develop or adopt the same, or at least compatible,
knowledge. Haas makes no allowance for a group whose internally-validated
knowledge may be completely incompatible with that of other groups in the
same issue-area, or for completely different definitions of the problem in the
issue-area. That this isn't seen as a conflict to Haas may be due to his evidently
preferred view of expertise as technical or relatively technical. The fields he has
mentioned (above) as outside the natural sciences are engineering and
economics, both of which might be seen as fundamentally technical areas of
inquiry.
It must be noted that Haas' definition of knowledge as accepted
belief, which is being adopted for purposes of this research, is not a universally
accepted definition. The competing groups in the population policy arena
displayed no tolerance for each others' sets of knowledge (accepted beliefs),
where they led to conflicts in policy objectives. No healthy skepticism about the
validity of current knowledge or eagerness to improve on it with new
information was exhibited by any of these groups, all of whom clung tenaciously
23


to their respective world views. This means that the situation noted above is
paradoxical within Haas' theory in Haas' terms, but not necessarily in terms
accepted by other theorists.
Evidence of a Preference for Technical Policy Subjects
Haas' defining essay on this subject serves as an introduction to a
special issue of International Organization containing a series of articles by
researchers applying the epistemic community concept to actual policy
situations. These articles all have relatively clear and unambiguous problem
definitions, due to the fact that the policy subjects addressed are relatively
technical in nature.
Among these articles, the issue-area most technical in nature is the
one where the single expert group appears to have had the most direct influence,
the ban of chlorofluorocarbons to protect stratospheric ozone. Haas describes
the
...ecological 'epistemic community', a knowledge-based network
of specialists who shared beliefs in cause-and-effect relations,
validity tests, and underlying principled values and pursued
common policy goals (1992a, 187).
In this situation, an epistemic community of atmospheric scientists and
concerned policymakers played an important role in developing consensus on
the need for the ban. This epistemic community had a "reputation for expertise
in atmospheric chemistry" (1992a, 196). Another example of direct influence is
an identified international epistemic community of economic development
24


specialists, agricultural economists, and administrators of food aid, all arguing in
a unified manner for longer range food security objectives in place of short-term
emergency food distribution (Hopkins 1992). While this policy issue-area
crosses disciplines, the policy problem can be seen as relatively clear-cut and
unambiguous, as in the technical policy issue-areas. The community based its
viewpoint, according to the author, on its members' shared knowledge and
experience.
An earlier article by Haas (1989) describes the success in
Mediterranean pollution control in those countries where technical experts were
able to influence their governments' policies. In this example, as well as the
other relatively technical and/or unambiguously defined policy subjects to which
the epistemic community concept has been applied, consensual knowledge was
eventually reached by the international policy community under influence of the
relevant epistemic community. In contrast to this example demonstrating the
development of consensual knowledge is an article by Emmanuel Adler (1992),
describing the influence of the arms control epistemic community. Adler notes
also that there were intellectual and policy rivals during the policy development
process, thus implicitly acknowledging the presence of several epistemic
communities.
Evidence of Haas' respect for the special status of traditional
university disciplines is seen in his acknowledgment of the distinctions among
groups involved in whaling policy in an article by M. J. Peterson (1992). The
three groups involved in this whaling policy study are "the epistemic community
25


of cetologists, the economic interest group of whaling industry managers, and
the issue-oriented lobbying coalition of environmentalists (Haas 1992b, 18). In
the Peterson article, the cetologists, seen as having "relevant claims to scientific
expertise" (Peterson 1992, 153), were unified in their objectives only until the
1970s, when differing interpretations of uncertainty in the data divided the
cetologists among conservationists, those among the environmentalists who
found some whaling acceptable, and the more radical preservationists who
supported a complete ban on whaling. The environmentalists are not considered
to have been an epistemic community because they were not unified in
objectives or knowledge base. The cetologists, on the other hand, are
considered by the author to have been an epistemic community even though they
were split.
Haas has noticed that when the experts in a given issue-area are split
into contending factions, which he suggests can happen in cases where scientific
evidence is ambiguous, then issues tend to be resolved less on their technical
merits than on their political ones. However, in his acknowledgment (above) of
Peterson's labels of groups in whaling policy, it appears that he does not notice
that these contending factions of cetologists might have developed opposing sets
of principled and causal beliefs, along with their separate interests or policy
enterprises, and possibly even different validity criteria, thus constituting
separate epistemic communities.
Further, the conservationist and preservationist groups were
represented by Peterson and Haas only as segments of the issue-oriented
26


lobbying coalition of environmentalists, rather than as two separate epistemic
communities, while it seems that each of these groups would have demonstrated
common principled beliefs, causal beliefs, interests and knowledge bases. There
appears, then, to have been a bias by both Haas and Peterson toward the
credentialed scientific community, even though Haas himself has stated that
natural science is not required as the knowledge base for an epistemic
community.
Applications and Critiques of the Haas Theory
There is no evidence that the epistemic community concept has been
widely adopted by policy analysts. It has, however, been applied experimentally
to policy development situations by a number of authors and commented on by
others. None of the existing applications of Haas' theory by other authors
focuses on groups' qualifications as epistemic communities. In each application
it is stated which groups are epistemic communities and which are seen as
interest groups or other categories, but the essays then proceed to focus instead
on the influence of the communities on the policy process. This is true both in
the International Organization issue in which Haas' theory exposition appears as
the introduction, and also in articles commenting on this concept in other
publications.
Sarah Mendelson has explored Haas' epistemic community concept
in conjunction with the dramatic changes in Soviet foreign policy in the late
1980s (1993). The article focuses on the political conditions under which Soviet
leadership learned from the epistemic community. Kathryn Sikkink, in the
27


context of a discussion of human rights in Latin America (1993), refers to the
epistemic community concept in terms of providing assistance to policymakers
who are uncertain about both what constitutes the national interest and how it
can be promoted. Again, the focus is on groups' influence in the policy process.
Referring to a 1992 study by Alexander S. Kamarotos, Lawrence T.
Woods (1993) questions the value of the epistemic community concept on the
basis of difficulties in applying it to the NGOs in the United Nations system
during the Earth Summit.
Abiding by the definition invoked by Peter Haas and company,
NGOs will rarely qualify as epistemic communities because
different NGOs and indeed different members of the same NGO,
while possibly agreeing on policy proposals, may disagree on
analytical perspectives or causal beliefs (Woods 1993, 11).
Woods evidently has not noticed that Haas does not require that epistemic
communities follow organizational boundaries. In contrast, the analysis in this
dissertation is being performed with the understanding that the five schools of
thought influential in population policy are free of organizational boundaries.
Any large environmental organization's membership, for example, can be
expected to exhibit a split between two perspectives on population growth, as
will be noted in Chapter IV.^
James Clay Moltz (1993) challenges literature that describes when
cooperative effects from consensual knowledge might replace competition in
international in policy development, including Haas' epistemic community
5 See subsection titled Other Groups, under Identifying the Groups, in Chapter IV.
28


theory. Moltz's policy example is an unsuccessful attempt at Soviet economic
reform, in which he identifies not convergent but divergent learning by
competing parties. In this policy situation, the politicians were observed to fall
back to positions consistent with the interests of their respective constituencies,
and exhibited selective use of available information.
[W]hile the process of learning may be knowledge driven, it is not
dictated by that new knowledge. At the level of policy, the
specific lessons "learned" by actors are affected in important ways
by their domestic political interests and the institutional context of
the learning process (Moltz 1993, 325).
One of Moltz's observations is that the subject matter of the transnationally
shared knowledge in the literature on international relations is usually technical
or scientific in nature.
James Sebenius' examination of the epistemic community concept
was not for the purpose of applying the theory but to explore the significance of
the theory in terms of his own work in negotiation analysis (1992a). Sebenius
suggests that an epistemic community can be understood "...as constituting a de
facto natural coalition seeking to build a 'winning coalition' of support behind its
preferred policy choice" (Sebenius 1992a, 352). He suggests also that this
theory has the effect of expanding the concept of negotiation. Rather than
limiting the negotiation process to that which lakes place at the table, it can be
seen as a tacit, even unacknowledged, ongoing communication process.
Sebenius interprets the "common policy project" promoted by an epistemic
community as
29


...an often tacitly proposed agreement...on a set of relevant issues
that reflects the community members' underlying shared interests
(values) by way of a common causal model (1992a, 352).
In this negotiation process, according to Sebenius, the effort to expand one's
own coalition or change others' perceptions becomes an action of competition
among groups.
Related Areas of Theory
Examining Haas' epistemic community theory in the context of
international population policy requires attention to a number of areas of theory.
Literature from a variety of areas will be reviewed in the remainder of this
chapter, with a focus on those elements of theory that might help to explain the
relationship between the observed population policy situation and Haas' theory.
This section is divided into five parts: Structural Issues, Behavioral Issues,
Epistemological Debates, A Rhetorical Approach, and Unavoidable Questions of
Degree. Ideas discussed in this section will contribute either to the analysis
(Chapters IV through VI), or to the interpretation of the results of the analysis
(Chapter VII).
Structural Issues
The title of this dissertation uses the term "interest groups" broadly.
This term has been selected because the agglomerations of people that were
observed as having been influential in international population policy
development do not fall uniformly under any single structural description other
30


than the very general terms "group" and "school of thought", and because these
groups were originally identified in the policy process by their highly visible
policy interests. In terms of structure, the groups themselves range in structure
from a completely unorganized school of thought, the market conservatives, to a
formally organized institution, the Vatican. To Haas, the presence of formal
organization is not a requirement for the presence of an epistemic community.
However, some of the theoretical literature in the section labeled Behavioral
Issues, below, contains assumptions about the presence of, variously, interest
groups, organizations or institutions. It seems helpful, then, to introduce briefly
a range of theory on those social categorizations, with some emphasis on the less
formal interpretations due to the unstructured character of all but one of the
groups in the international population policy arena.
Organizational status of interest groups. Haas views epistemic
communities as independent from formal organization, but interest groups have
not always been viewed in this way. Harry Eckstein (1968) saw "pressure
groups" in terms of their potential effectiveness, which was considered
dependent on attributes of the groups themselves and on the governmental
system they were attempting to influence. Writing in response to Eckstein,
Joseph LaPalombara (1968) described limitations of interest group theory,
claiming that it is impossible to measure the relative influence that groups exert
over administrative decisions. He declared that organization theory should
cover adequately the dynamics that Eckstein saw as requiring interest or
31


pressure group theory. Both of these authors considered the groups they
examined to be formal structures.
David B. Truman saw the interest group in a broader framework, as
...a segment of a public that shares a similar view of, or attitude
toward, the consequences under discussion. It may be merely a
potential interest group, of course, or an actual one if its members
attempt to "do something" about the consequences (1951, 218).
Such a group, according to Truman, could be large, diversified, and
disconnected. This is not unlike the advocacy coalitions described by Paul A.
Sabatier (1988), composed of organizations, individuals and sometimes sections
of agencies with a common set of core beliefs and values, exhibiting some
degree of coordination toward support for a policy position. James K. Sebenius
describes Haas epistemic community roughly in the same light as Sabatier's
advocacy coalition, as a
...de facto natural coalition of "believers" whose main interest lies
not in the material sphere but instead in fostering the adoption of
the community's policy project (1992a, 325).
While Haas, unlike Sebenius, carefully distinguishes epistemic communities
from interest groups, it is not clear whether he considers interest groups to be
formally organized or even coordinated.
Defining institutions. Terry M. Moe (1980) discusses interest groups
in terms of organization, but not as being identified by their organizations.
Instead, he sees them as behaving like organizations. Members may join,
contribute, and participate for reasons that have nothing whatever to do with the
group's political goals or activities.
32


In the internal life of an economic interest group, the role of
politics is problematical, its importance varying with a host of
factors bearing on the incentives of individuals and the context in
which they make their decisions (1980, 72).
This position Moe contrasts with the theory of Mancur Olson (1965), who
applies rational choice theory to groups of people who are all working toward
the same goals, seen as collective goods. Moe sees individuals' reasons for
joining organizations in terms of the self-interest of those individuals. This self-
interest is not necessarily economic; it might instead be measured in terms of
satisfaction derived from participation. Moe's view is based on an individual's
presumed assumption that he personally can make a difference in the outcomes
for the organization as a whole.
Just as interest groups can be seen as formally organized or more
loosely structured, institutions can be seen in a variety of ways. James G.
March and Johan P. Olsen (1989) have described political institutions as a
"systems of rules and structure of meaning". Similarly, Elinor Ostrom has
described institutions in terms of sets of rules (1986), and sets out ways to
classify rule configurations. Ostrom points out that there are a number of
"referents" for the meaning of "institutions", and an answer to the question,
"What are institutions?" would require a decision about the "right" or preferred
referent. She mentions also some views of institutions given as possibilities by
C. R. Plott (1979), including preferences and opportunities as defining the
structure of institutions, institutions as customs and ethics, and institutions as
coalitions.
33


Meanings of "interest". Because interests are fundamental to interest
groups, the presence of a variety of interpretations of this term should be
acknowledged. In particular, the literature contains differing positions on
whether interest should be interpreted as self-interest. Paul Sabatier (1988) sees
interests in material or economic terms. At the same time, however, he includes
value issues and preferences in "belief', including preferences characterized by
altruistic objectives. Therefore Sabatier's working definition of "interests" is
simply narrowly defined, rather than a commentary on human behavior as being
only self-interested. Aaron Wildavsky (1987), in the process of discussing how
preferences are chosen, distinguishes preferences from interests, gives
indications that he views the latter as self-interest.
Taking the other side, Anthony Downs comments that his economic
theory of democracy, exploring patterns of choice analogous to economic choice
in democratic processes, does not adequately account for altruism (1957, 9,
quoted in Almond 1991; personal conversation with author, 1993). Barry
Hindess (1989) writes of the inadequacy of the assumption of self-interest in the
rational choice tradition. Herbert Simon (1993a) bemoans the treatment of
economic gain as the primary human motive in economic theory, and calls for
"an empirically grounded theory [that] would assign comparable weight to other
motives, including altruism and the organizational identifications associated with
it" (1993a, 160). He suggests that in public choice theory, we cannot predict
voting behavior without knowing with what groups people identify, and that
assuming self-interest is inadequate for this purpose. James Sebenius, in
34


discussing negotiation between competing epistemic communities, specifies that
interests are not limited to self-interest:
Note that these interests need not be material but may be entirely
consistent with the subjectivist orientation of negotiation analysis:
anything that enters a utility function, whether a commodity or an
altruistic project, can have the analytic standing of an interest
(1992b: 352).
Behavioral Issues
Haas' emphasis on epistemic communities' shared beliefs, and this
author's observation of evidently consistent beliefs in the international
population policy arena, raises questions about how beliefs develop in policy-
interested groups. Christer Jonsson describes a growth of interest in cognitive
approaches to studies in international politics, based possibly on increasing
complexity of relationships among states and the "perceptual lag" of individuals'
information in fast-moving events (Jonsson 1982, 2). Robert Axelrod and
Robert Keohane found that contributors to the World Politics issue on
"Cooperation Under Anarchy" "did not specifically set out to explore the role of
perception in decision making, ... [but] the importance of perception ... kept
asserting itself" (1985, 247). Stephan Haggard and Beth A. Simmons, in
defining international regimes, suggest that "historical episodes of cooperation
may be inexplicable without reference to shared knowledge and meanings," in
particular "the degree of ideological consensus and agreement over causal
relationships" (1987, 511). Ambiguity in defining the issue-area of cooperation,
they suggest, makes a cognitive approach important, for "structural, game-
35


theoretic and functional theories assume that cooperation operates within an
issue-area which is relatively unambiguous" (1987, 510).
The subject of belief appears in both cognitive and social
psychology, and in Herbert Simon's view, there is no clear distinction between
these two closely related fields.
Since adaptive behavior is a function of strategies and knowledge,
both largely acquired from the social environment, there can be no
sharp boundary between cognitive psychology and social
psychology. The context in which knowledge is acquired and
used, an exogenous variable in cognitive psychology, provides the
endogenous variables for social psychology and sociology (H.
Simon 1992b, 157).
Simon points to our inability to receive or handle all of the information we need
to optimize outcomes, and criticizes rational choice theory on this basis. He
applies this limitation to situations "that are complex and in which information is
very incomplete (i.e., virtually all real world situations)", and says that in these
situations even approximating maximization of utility is not possible (H. Simon
1987, 39). He has suggested,
Even before the development of the modem interest in rational
behavior under uncertainty, it should have been obvious (at least,
so it seems in hindsight) that the classical economic model of
rational man makes preposterous demands on the ability of the
entrepreneur to acquire and use information. The additional
burdens placed on the decision-maker by uncertainty (e.g., the
need to estimate joint probability distributions of future events)
made the preposterous doubly preposterous, and resulted in a
certain minimal attention to the information-processing problem
(H. Simon 1982, 236).
36


Social psychologist's views of belief. Definitions of belief vary
somewhat among texts in social psychology. Beliefs are generally seen as one
of three components of attitudes. Attitudes are defined as having a belief
component (cognitive), a feelings component (affective), and an action
component (conative) (Pratkanis and Greenwald 1989; Zimbardo 1976; Finlay,
Holsti and Fagen 1967). In this context, there is rarely much discussion about
how the belief component is defined, and differences do appear. Pratkinis and
Greenwald as well as Finlay describe belief as knowledge, or what we know,
whereas Zimbardo includes a value factor in belief. Just as psychologists use
the term "belief" in inconsistent ways, so do authors whose theories enter into
this examination of epistemic communities. James Sibenius (1991a,b; 1992a,b)
appears to view beliefs as acceptance of fact, as does Lindblom (1990). In
contrast, Paul Sabatier (1988) includes values in beliefs. Haas sees beliefs as
containing values, but makes a distinction in labeling, differentiating those
beliefs containing values as "principled beliefs" from the beliefs about accepted
fact, which he labels "causal beliefs" (1992b).
Belief formation in the cognitive model. A number of social and
cognitive psychologists have searched for explanations of the ways people
choose information, with implications for the way groups form common beliefs.
The basic premise of cognitive psychology is that people develop their factual
views of the world (beliefs in the factual sense) on the basis of previously held
positions, beliefs, or values. An early social psychology textbook by T. M.
Newcomb describes influences on perception:
37


...Perception involves a good deal of omitting ... supplementing
... and structuring whatever is available to be perceived ... Any
object, event, or situation is necessarily sized up in relation to
something, and this "something" constitutes the frame of reference
.... (1950, 90-95, 212, in Wildavsky 1962, 723.)
Holzner and Marx describe this frame of reference as
...a structure consisting of taken-for-granted assumptions,
preferences for symbol systems, and analytical devices within
which an observer's inquiry proceeds (1979, 99).
Cognitive psychologists typically see beliefs as operating through a belief
system, which is defined by Finlay as "all the beliefs, sets, and expectancies,
conscious and unconscious, that are accepted as true of the world" (Finlay,
Holsti and Fagen 1967, 19). The concept of belief system is explored by Milton
Rokeach (1960), who suggests the presence of both belief and disbelief systems.
James N. Rosenau and Ole R. Holsti describe a belief system as [enabling] its
holder to explain the course of events and to interpret new developments as
confirmations rather than negations of their convictions" (1981: 375). A belief
system is, they say, internally consistent and tied together by its own logic.
Raymond Boudon has introduced the concept of subjective
rationality, in which false beliefs can be grounded and consolidated (1992), and
by this mechanism primitive beliefs, for example, can be seen as rational. The
importance of this, and also of the seemingly purposeful design of belief systems
by Rosenau and Holsti (1981), in connection with the current research into a
policy issue, is that different groups can develop different sets beliefs from the
same information.
38


Alexander George describes an "operational code" as a set of
existing beliefs (of political content, in his case) determining in part the
acceptance of new information (1979, 96-104). George observes that a "master"
or "first philosophical belief concerning the nature of the political conflict, and,
related to, the image of the opponent" affects other beliefs related to that issue
(1979, 101). He suggests further that
...a person's operational code beliefs structure and channel the
way in which he copes and deals with the cognitive limits on
rationality; they serve to define his particular type of "bounded
rationality" (1979, 103).
Other authors taking the cognitive or generally behavioral approach to
belief formation point to the shaping of beliefs by group membership. Herbert
Simon suggests that group loyalties have a cognitive component, for they "cause
particular variables and simplify world models to govern the thinking of group
members" (1993a: 159). James March and Herbert Simon say,
Organization members are social persons, whose knowledge,
beliefs, preferences, loyalties are all products of the social
environments in which they grew up, and the environments in
which they now live and work (1993, 2).
Leon Festinger's cognitive dissonance theory (1957) proposed that
individuals presented with dissonant information will find ways to reduce the
dissonance, including avoiding exposure to the dissonant information. Much of
the research that has been conducted on this theory has not confirmed the
expectation that people will limit their exposure to information, instead
producing conflicting answers as to the importance of an actor's choosing to
39


avoid exposure to information dissonant with his beliefs. Jonathan L. Freedman
and David O. Sears' have observed, based on a review of a series of
experiments in this area, that the mechanism for selecting among available ideas
may not be a predictable reduction of exposure to dissonant information, but
instead other mechanisms in the form of resistance, such as challenging and
disparaging the material (1965), and exposing it to "mercilessly unsympathetic
scrutiny" (1967, 213). These mechanisms at the level of evaluation, they
suggest, serve the same purpose as selective exposure to information, and fit
better the available evidence. Irving Janis and Curt Rausch (1970) support the
Freedman and Sears findings with a variety of suggested reasons for
disparagement rather than selective exposure to information in a range of
situations.
John S. Odell (1982) distinguishes between specific and general
beliefs, describing the specific variety as "the individual's causal map of the
immediate situation", and the general beliefs, along with values, appearing as
predispositions shaping and coloring the way new information is processed
(1982, 62-63). Odell's comments are in the context of observations of United
States international monetary policy development. Robert Jervis concurs with
this view, saying that existing theories and images determine, to a large extent,
what people notice (1976). This has been confirmed by psychologists Anthony
Pratkanis and Anthony Greenwald (1989), who report on an experiment in
which subjects, in describing other individuals, made internal attributions that
were consistent with their own attitudes.
40


Theories of belief formation in a policy context. The common
beliefs held in common by members of epistemic communities, and tbe shared
bebefs expressed by tbe groups influential in international population policy
during UNCED, warrant attention to theories of group belief formation in
connection with public policy. Cultural anthropologists have long seen the
promotion of common bebefs as having served as an organizing influence in
societies throughout human history (Plog and Bates 1976).
Several authors have suggested that behef development plays a role
also in organizing or defining interests in the pubbc pobcy development. Hank
Jebkins-Smith (1990) suggests that bebefs are developed to support positions
when people are chahenged by others' problem definitions with potentiaby
adverse effects on them. In his model, constructed bebefs appear to serve as
problem definitions in pobcy debate, which must take place when the threat is
present. Jenkins-Smith's observation that once a set of bebefs or causal relations
has been accepted in and has become central to this process, "there is often
strong resistance to change even in the face of conflicting evidence for such
change" (1990, 91). Moltz (1993), in his article on divergent learning, observed
resistance to available knowledge by participants in the Soviet economic reform
efforts when they found it more practical to adopt ("borrow", for Moltz) ideas
from behef systems consistent with the positions of their respective pobtical
constituencies. Irving Jams' research developing the concept of "groupthink"
(1982) provided additional support for the notion of a groups construction of
factual bebef in the process of developing a preferred position that was in turn
41


consistent with policy needs. In his examples, the group conformity
strengthened adherence to the belief.
A concept of secondary beliefs that serve as strategic support has
come from Sabatier's description of an advocacy coalition framework of policy
change. Sabatier (1988) differentiates between core beliefs, those which are
closely held and tightly defended by members of a coalition, and lower levels of
belief, where there is more possibility for "policy-oriented learning", or belief
change based on experience or observed inconsistencies between those beliefs
and observed phenomena. Jenkins-Smith (1990) similarly refers to core and
peripheral values imbedded in the belief systems of participants in the policy
process, affecting the stability or strength of the adherence to a given belief. In
place of Sabatier's policy-oriented learning, Jenkins-Smith offers the concept of
intractability of beliefs. Those beliefs based on core rather than peripheral
values are hard to alter, or intractable, while those in the peripheral area,
provided they are not at too low a level to warrant serious concern, stand a
greater chance for change.
Robert Jervis, in writing of behavioral components of political
action, describes constructs by persons desiring to present the optimally coherent
case for a desired policy action. In social psychology this is known as irrational
consistency, in which "people who favor a policy usually believe that it is
supported by many logically independent reasons" (128). Jervis illustrates this
behavior with historical examples in international defense-related policy
42


development, pointing out that belief systems thus multiply supported often
display overkill.
Cross-cutting theorists' views on conformity of group beliefs. One
group of authors, crossing boundaries between political science or policy
studies, psychology and philosophy, focuses on convergency or conformity of
beliefs within groups or institutions. Elinor Ostrom provides an explanation for
uniform beliefs within a group in her theory on institutional rational choice
(1991). Institutional rational choice theory is presented as a combination of
rational choice theory and institutional analysis. Rational choice theory assumes
that our preferences are expressed according to our individual self-interest, or in
economic terms, utility. It assumes that everyone is rational, and knows what
his best interests are, and choices and decisions are made according to expected
consequences (Elster, 1984; Ostrom, 1991). Institutional theory, in contrast to
this thinking, places greater weight on the cultural influences on choices and
decisions.
Ostrom points out that a problem arises when one attempts to explain
the formation of interest groups under rational choice theory, which would
require that all members of a coalition develop a set of positions consistent with
their own interests but similar to those of all the others in the coalition. She
suggests that the applicability to this situation of rational choice theory can be
improved by merging it with institutional analysis or theory, accounting for
rational decisions within the constraints present in the cultural or institutional
framework. She writes of the ruling-in and ruling-out of actions, which serve to
43


limit choices for these constrained rational decision makers. By joining an
organization (or institution, or interest group) that has ruled out behaviors and
ideas, one expresses one's rational choice among ideas within the ruled-in set of
actions. This narrowing of options, Ostrom suggests, fosters the cohesiveness
of the institution.
Robert Grafstein (1991) produces an explanation of conformity and
institution formation or maintenance, based on rational agents' views of
improvement of self-interested outcomes through their own participation in that
conformity. In this view, if an individual perceives that the objectives of a
coalition are those with which he agrees, and also that the chances of success are
dependent on his own participation in that coalition, he will adopt the positions
of the group. Presumably those positions include the group's beliefs. Grafstein
is describing behavior amounting to a social interpretation of utility.
Aaron Wildavsky looks closely at how individuals develop their
preferences, such as political choices, based on very little information. His
answer to this puzzle is given in a framework of cultural theory, in which
individuals' preferences are formed in the process of developing social relations.
Wildavsky suggests that group preferences "emerge from social interaction in
defending or opposing different ways of life" (1987, 5). They serve the purpose
of justifying or rationalizing the group's social practices. These practices are
described as social arrangements.
The rational act on the part of the individual, according to
Wildavsky, is the selection of the appropriate group to join, except in cases
44


where the decision is made under coercion. The joining of a culture is the
acceptance of a form of social constraint. An individual can quickly detect and
adopt the pattern of world-views and political positions of the group joined. The
implication is that in joining a group one is compelled to accept its constraints.
This helps to explain Wildavsky's observation that individuals develop "miles of
preferences" from "only inches of facts" (1987, 8) in the process of social
interaction within a group. The adoption by individuals of a conformed package
of beliefs could be inferred from this, although Wildavsky does not offer us this
possibility directly, since beliefs and preferences are different animals.
Charles Lindblom describes belief convergences as characterized by
patterned impairment of learning. Lindblom (1990) builds a case for asserting
that all culture is dependent on impairment of the learning process. This
concept of culturally determined impairment is consistent with Ostrom's ruling
in and ruling out of options in making choices based on institutional
requirements, but Lindblom has developed more fully the mechanisms for this
activity. Lindblom's concept of purposeful impairment bears similarities also to
Wildavsky's constrained cultures, but goes further than the cultural theory by
focusing on the constraints as limits on learning, not only on preferences
expressed.
Lindblom, like Herbert Simon in acknowledging that we cannot
know everything in complex social situations, focuses on the role of belief in the
place of knowledge. He looks at the conditions under which convergence of
beliefs occurs, and observes that increased knowledge is not the driving factor,
45


for in a complex social system knowledge can create more diversity in views of
causality rather than fewer. He then suggests that the impairment of learning,
producing the limited nature of group knowledge, provides the means of belief
convergence. To Lindblom, that convergence itself narrows the range of
acceptable beliefs. This suggests that groups' belief systems may be internally
consistent not necessarily because people are drawn to them due to their
consistent beliefs; but instead, that some coalitions' belief systems are consistent
because they become consistent after the groups with common objectives are
identified, and new members are drawn into the vortex of convergence,
conforming their beliefs en route.
Such an inference, that coalitions form around common objectives or
interests and not only around common beliefs, and that belief systems become
consistent after those common goals are identified, in turn implies that beliefs
can be developed for rational reasons. 6 Jon Elster has explored the possibility
of rational belief and rejected it (1984). His rejection is based on his
observation that the decision to believe would have to be accompanied with a
decision to forget that which one has already believed, which he deems
impossible. Elster suggests, however, that an abdication of reason is
conceivable, and such an abdication of reason does seem consistent with
Lindblom's description of impaired learning, where one willingly separates
oneself from a body of knowledge in the process of social conformity, thus
6 Rational here must mean with reason, with an eye on consequences, not necessarily self-interest.
This is consistent with Herbert Simons definition of rationality, meaning consistent with goals
(1993).
46


adopting the beliefs consistent with those of the conformed group. It also
resembles Ostrom's observation of relinquishment of ruled-out ideas in joining
an institution.
Robert Rothstein has presented an example of what he has seen as
politically tailored beliefs, in an article on commodity negotiations. Rothstein
describes the United States' failure to acknowledge relationship of commodity
trade issues and prices to other parts of the economy and to characteristics of
developing country economies, and says that "political leaders on both sides
remain tied to perceptions and attitudes that are impervious to analytical
considerations when those consideration threaten valued goals (1984, 748.)
Similarly, Irving Jams' "groupthink" concept is based on the possibility of every
cohesive group's purposeful (although not always consciously purposeful)
perceptions, such as shared negative stereotypes and "a set of policy doctrines,
derived from the members' subculture, that provides the members with a
cognitive map for conceptualizing the intentions and reaction of opponents,
allies, and neutrals" (1982, 257).
Connecting causal beliefs and problem definitions. Problem
definitions and causal beliefs appear to be connected through their roles in
describing or accounting for causal relationships in the policy areas being
addressed. Janet A. Weiss has made this connection in describing the powerful
role of problem definition in policy formulation. She suggests that the definition
of policy problems, while commonly seen as the beginning of the policy
process, is actually at the heart of it, often remaining an open question through
47


the entire process. "Much policymaking, in fact, is preoccupied with whose
definitions shall prevail" (1989, 98). Weiss suggests that coalitions form around
competing definitions. Sometimes a common language can be found to make
cooperative progress possible between the competing coalitions. At other times,
multiple definitions survive the policy formulation process "to haunt the
implementation process, as differing conceptions of the problem guide the
various actors who work to turn the official rhetoric into many realities" (98).
Weiss suggests fiirther that problem definition "is a package of ideas that
includes at least implicitly an account of the causes and consequences of some
circumstances that are deemed undesirable, and theory about how a problem
may be alleviated" (Gery 1984 and Gusfield 1981, in Weiss 1989, 97).
Epistemological Debates
Jonsson has pointed out that a cognitive approach to politics,
acknowledging that people's reception of information is influenced by pre-
existing images and ideas, inevitably has epistemological implications and raises
some philosophical questions (1982). The dominant debate in epistemology
concerns what we can and cannot know, or, in effect, what constitutes
knowledge. On one side of this debate are those sometimes referred to (often
pejoratively) as relativists, those people who say our knowledge, even scientific
knowledge, may always be an imperfect representation of reality and is subject
to change when new evidence appears. Thomas S. Kuhn (1970) describes
paradigms as never necessarily final, and within which all scientific progress is
48


made. A paradigm is eventually overthrown if anomalies revealed in the
process of normal science warrant change, and only after a period of
competition between old and new paradigms. Richard J. Bernstein (1983)
discusses Kuhn's view of knowledge at length in view of the potential for
shifting paradigms, suggesting that the shifts are between incommensurable
perspectives, perspectives that cannot be compared directly because they address
different problems and standards rather than negating each other's information.
Odell (1982) suggests that differing schools of thought appear
according to information accumulated by persons guided by different belief
saliences, defined as different subjects dominant in their minds and important to
them. The information processed by people operating on different perceptions
of what is important will produce different answers. This model appears to
differ from Kuhn's in its recognition of concurrently present paradigms or
schools of thought. Kuhn seems to imply that a single paradigm is dominant at
any given time.
Yaron Ezrahi asserts that knowledge is dependent on reference
groups.
Analytically, the sense that knowledge is context-free and
universally compelling is of course an "optical illusion" that
emerges from the assumption that scientific facts and truths are
defined in reference to universal cognitive and normative frames
of reference or to an all-inclusive social reference group (1974,
226).
Along this same line, Ludwig Fleck has introduced "thought collectives",
explaining,
49


...once a structurally complete and closed system of opinions
consisting of many details and relations has been formed, it offers
constant resistance to anything that contradicts it (1979, 38).
To Fleck, cognition is a socially-conditioned activity and knowledge is a social
creation. Michael Mulkay (1979) sees scientific knowledge as subject to a
continual process of cultural reinterpretation, by means of which scientists
develop their views of the physical world. Holzner describes knowledge as
follows:
..."knowledge can only mean the "mapping" of experienced
reality by some observer. It cannot mean the "grasping" of reality
itself. In fact, philosophical progress has produced the conclusive
insight that there can be no such thing as the direct and "true"
apprehension of "reality" itself. More strictly speaking, we are
compelled to define "knowledge" as the communicable mapping
of some aspect of experienced reality by an observer in symbolic
terms (1979, 93).
Kenneth Boulding expresses more faith than Mulkay in the ability of
humans to know the truth about real conditions. Boulding thinks of the growth
of knowledge "as the growth of structures within the human organism that
correspond in some way in their patterns to structures outside it, ... [and] the
structures outside it ultimately dominate the growth of the structures inside it"
(1978, 237). Boulding calls this the "outability of truth", a concept that
corresponds somewhat to Haas' faith in the ultimate convergence of knowledge.
Donald Campbell distinguishes between scientific and social
knowledge, firmly backing Karl Popper's conviction that, even though evidence
can cause us to reject earlier beliefs, scientific knowledge does bring its
50


scientists ever closer to objective reality. Donald Campbell (1988) excoriates
Kuhn and other supporters of the social nature of science by describing a "tribal
model" of the social system, in which groups' social beliefs and superstitions are
maintained to constitute self-perpetuating belief systems. The scientific
community, according to Campbell, is not characterized by this tribalism. In
science, challenging an assumption is usually acceptable and admired. The
scientific world is the only place, according to Campbell, where "competent
arrogance" is rewarded. Accordingly, he says,
.. .the norms of science are uniquely anti-tribal, and this may help
make science more able to minimize the interference of authority,
conformity, and loyalty mechanisms in distorting the content of
scientific beliefs (1988, 503).
Campbell, then, would appear to have provided the key support for Haas'
epistemic community, in his acknowledgment of the special status granted the
knowledge-based group for its technical competence. However, the research
for this dissertation is expected to show that, in the population policy
development process during UNCED, it would have been unwarranted to expect
that a correct body of knowledge would be reached by convergence of ideas
among the influential groups.
The Communication Approach
In contrast with Haas' interest in beliefs and knowledge, an alternate
view of the policy development process is a focus on the communication,
including the rhetoric of debate and presentation. Jurgen Habermas recognizes
51


the importance of distortion in communication and has focused on moral issues
derived from this distortion. He has specified the "ideal speech situation" which
is "free of distorting influences" (Habermas, in Farrell 1993). An objective is to
"uncover the influences that, unknown to individuals involved, are distorting
their beliefs and actions" (Bobrow and Dryzek 1987, 71).
Frank Fischer and John Forester (1993) have taken the examination
of communication a step further, to examine rhetorical content of political
communication, exploring the role of language in the shaping of views in the
policy arena. Unlike Habermas, they do not focus on unintentional distortions
as much as on rhetorical design and it effect. In their new book focusing on
argumentation in policy analysis,? the editors tell us:
If analysts' ways of representing policy and planning issues must
make assumptions about causality and responsibility, about
legitimacy and authority, and about interests, needs, values,
preferences, and obligations, then the language of policy and
planning analyses hot only depicts but also constructs the issues at
hand (Fischer and Forester 1993, 1).
The editors carefully separate these explorations into the nature of
policy and planning discourse from cognitive issues and epistemological
questions, and more particularly from knotty problems of epistemological
relativism, by stating in their introduction that the statements made by analysts
1 While this statement and the book's title suggest that policy analyses serve as the contexts of the
influential language, the chapters of the book look beyond formal analyses to verbal exchanges in
meetings and other forms of discourse in policy development, such as interest group political
commentary. Two examples of this are differing policy belief systems concerning Dutch ethnicity
policy (Robert Hoppe, page 77-100) and rhetorical patterns revealed in the disputes over electric
power planning arguments in Chicago (J. S. Throgmorton, pp. 117-144).
52


(and presumably political competitors as well) are claims, not taken by these
editors as "truth". Fischer and Forester do not consider every claim to be as
valid as every other claim, and they point to Bruce Jennings' suggestion in his
chapter on health policy discussion in Oregon (1993, 101-116) that ensuring
relatively dialogic democratic practices will make it possible to avoid endless
cycles of debate between seeming intractable belief systems. The editors point
to the success of continuing dialog in achieving gains in knowledge in the
scientific community.
The concept of a "discourse coalition", so dubbed by Maarten A.
Hajer (1993 43-76) is potentially a rich one to apply to the population policy
arena. The language of the competing groups does indeed help to explain the
seemingly disciplined but actually unorganized convergence of beliefs within any
one of the groups. Of special importance is Hajers observation about the way
to view policy influence through language change:
A discourse coalition becomes dominant if (1) the central actors
are persuaded by, or are forced to accept, the rhetorical power of
a new story line (discourse structure); and (2) this is reflected in
the organizational practices of a given domain (discourse
institutionalization) (1993 66).
Fischer and Forester recognize the importance of subjects not
included in a policy discussion.
In some cases, what analyses do not say matters more than what
they do say. Analysis focuses attention selectively and
deliberately, enabling a more focused consideration of some
alternatives and excluding others from practical consideration
altogether (1993, 6).
53


The editors explain that by recognizing "policy analysis and planning as
argumentative, we can understand immediately how they can be complex
exercises of agenda-setting power" (1993, 6). This could be a useful and
interesting way of viewing the competing groups in the population policy arena.
In their reluctance to acknowledge the significance of their
contributing authors' work in terms of knowledge itself, the subject they
consciously avoid, Fischer and Forester have made a decision with implications
for validity in political research, and also with implications for the role of
deception in political communication and policy development. In terms of
validity, by examining the rhetoric and not people's beliefs, they are reporting
only what they can see and verify, a process that probably suffers fewer validity
problems than assuming, as in epistemic community theory, that the political
statements represent people's beliefs. At the same time, it must not be
overlooked that a fair number of the rhetorical statements must represent beliefs
of those giving them, because of the speaker's own learning processes
influenced by the rhetoric of others.8
The implications for deception are potentially interesting. If the
rhetoric is composed of claims, it must be assumed that such political
communication is heavily characterized by lies. Lying, according to Sissela Bok
(1989), is stating information that one knows is riot true. A claim must be a fie
8 That is, messages given as claims not believed by the speaker are not packaged with a disclaimer,
"Here is my claim for you to use, but don't believe it." It is likely that many listeners believe
others' stated claims, and therefore, presumably only a small percentage of speakers are at the
beginning of the "chain" of messages carrying such claims not representing actual belief.
54


if it is not believed by the speaker. If, on the other hand, the rhetoric represents
the speaker's beliefs, and if those beliefs were obtained by selective attention to
available information or disparaging treatment of information not consistent with
one's interests, self-deception is at work.9 These are only two possibilities; an
entire matrix of combinations of speakers' and receivers' self-deception, beliefs
and lies is conceivable.
Fischer and Forester's volume stands as a significant challenge to
epistemic community theory, largely because of the importance of recognizing
the role of language in policy change, and also because of the increased validity
of research that does not assume that the speakers believe what they say. In this
dissertation research, however, it has seemed that the activist members of the
five groups in the international population policy community have indeed
believed what they were saying. To view their words as representative of belief
systems, or sets of knowledge separate from the sets held by other groups,
appears to be a productive way to understand the divergent opinions in the
policy community. According to the cognitive psychologists and the policy
experts who take a cognitive approach to understanding policy development,
prior beliefs and values serve as a basis for acquiring beliefs, and this selective
acquisition process can account for the persistent differences among groups'
beliefs and the convergence of beliefs within them.
9 Bok (1989) provides a useful discussion into the differences between deception and lies; Jon
Elster (1989a) discusses the differences between self-deception and wishful thinking, in the context
of rationality and irrationality.
55


A core problem here is that, in research on peoples' political
positions, we cannot know whether a speaker's words represent his beliefs.
Robert Axelrod's (1976) approach to this problem is a practical one. In
developing a system of cognitive mapping for describing beliefs of participants
in policy decisions, 10 he has been careful not to attempt to draw distinctions
between actors' causal assertions (claims) and "true" or "inner" beliefs, because
of the difficulty in determining the latter, and he simply accepts their statements
as their beliefs. Axelrod reminds readers that even the actors themselves
sincerely describing their decisions cannot always distinguish between their true
motives and their own rationalizations for their actions. It seems that the safest
position on whether political statements represent beliefs or are merely claims, is
that distinguishing between them in any policy community cannot be definitively
achieved. That is the safe position of the discourse coalition literature, but not
of this dissertation research.
Unavoidable Questions of Degree
It is not uncommon for attempted categorizations be plagued with the
problem of gradual change from one category to the next. Fuzzy set theory
(Laughlin 1993) addresses this problem, asserting that it must not be assumed
that categories can't be made because clear lines don't separate them. E. E.
Schattschneider reminds us of this problem in a discussion about organization.
10 Drawing on a wide range of sources for clues, Axelrod creates the graphic maps based on actors'
assertions of causality.
56


£0Z
We do not dispose of the matter by calling the distinction between
organized and unorganized groups a "mere" difference of
degree[,] because some of the greatest differences in the world are
differences of degree (1960, 28).
Miles and Huberman (1984) recognize also the difficulties of deriving
distinctions from a large, shapeless set of qualitative data, but assert that in the
social world there are patterns to be uncovered by such exploration, not only
patterns to be imposed by researchers. This dissertation research explores a
political arena characterized by a wide variety of combinations of beliefs and
positions on the policy issues. The "sets" or groups discussed are expected to
have "fuzzy" edges, but it is assumed as a basis for this research that the
patterns revealed will have greater significance, in terms of both policy and
theory, than the continuity between the groups observed.
57


CHAPTER m
THE INTERNATIONAL POPULATION POLICY ARENA:
LABORATORY FOR THE STUDY
Haas has pointed out that it is particularly in those policy
development situations characterized by uncertainty that policymakers turn to
expert groups, or epistemic communities, for advice. The international
population policy arena is undeniably characterized by uncertainty for
policymakers, in that any damage done by population growth is hard to
measure, and broad disagreements about solutions persist. Establishing
causality in a complex social system is difficult. Adding to a complex social
context a global dimension and a large dose of ethical content, the uncertainty is
greater still.
The laboratory for this study is the United Nations Conference on
Environment and Development (UNCED), which formally began with an
organizing session in March 1990 and culminated at the Earth Summit in Rio de
Janeiro in June 1992. In this process, many people sought to increase the
conference's focus on population growth as a factor contributing to
environmental decline or hindering development, and hoped that the Rio
conference would produce a clear statement affirming these relationships. At
the same time, many other people sought to reduce attention to population
growth in the UNCED proceedings (Lewis 1992).
58


In addition to these fairly transparent objectives, the extended
conference process revealed among the participants in population discussions a
variety of inconsistent beliefs about population growth, and differing views
about what would constitute appropriate policy content. It appeared that a
variety of interest groups were bringing some variety of perceptions of the
global population situation into the policy formulation process, confusing
observers and even some participants with the competition over establishing
objectives and appropriate means to the (various) desired ends. This situation,
observed by this author at the fourth Preparatory Committee meeting (PrepCom
IV) at the United Nations during March, 1992, and at the Earth Summit in Rio
de Janeiro in June, 1992, was not inconsistent with long-standing confusion in
the population field as a whole.
The Population Subject:
Disputes. Confusion and Sensitivity
Defining the Problem
There has long been considerable disagreement over two basic
questions about population among policy-makers, their critics and other
observers. The first of these questions is whether population growth on an
international or global scale is a problem worthy of policy attention. Many
people consider population growth to be a critical problem that, if left unsolved,
stands as a major threat to the environment and will lead to increased human
suffering in the future. Many others feel that these worries are overblown and
that the earth has infinite capacity to absorb human population growth, given the
59


creative capacities of people to develop technologies for more efficient use of its
resources.
An overriding, fundamental problem contributing to uncertainty and
disagreements in the population field is the difficulty in demonstrating causality
in a complex social system. For any given environmental change or alteration in
human conditions, there are usually several possible causes and evidence is
available to back up everyone's claims. For example, it is difficult to make a
definitive connection between population growth and forest depletion when
someone produces persuasive evidence that land use policies and practices also
determine how forests are treated.
The ability to establish causal connections with a high degree of
confidence is important to policymakers, the people ultimately responsible for
the allocation of common resources. Lawrence Summers, speaking as chief
economist of the World Bank (1991), noted the difficulties of planning in this
area when the causal relationships are unclear concerning population growth and
environmental degradation, determinants of migration, links between population
growth and inequality in the wage structure, and the fiscal implications of
population growth.
Nathan Keyfitz (1993) has demonstrated that biologists, who
typically view population growth as a problem, and economists, who often
don't, exhibit fundamental differences in methods and perspectives that, until
now, have gone largely unrecognized. He observes that biologists see humans
as only one species among many, see life in terms of the eons of evolutionary
60


time, and deal with absolute size in relation to the biosphere, while economists
see humans as the only terrestrial actor, view action on a time scale of years or
decades, and care more about proportions and their allocation than about scale.
Central to economists' thinking is indefinite market-driven substitutability of
resources, while biologists are not satisfied with the substitution of one fish
species for another. These are among the eight "axes of difference" that,
according to Keyfitz, go far to explain diametrically opposed positions on
whether there are limits to growth of the human population and its activities, or
whether there are virtually no such limits.
Keyfitz (1993) discusses the careful demonstrations by economists
that population growth has been closely correlated with increases in per capita
Gross Domestic Product (GDP) throughout history. These economists tend to
assume that the former was causing the latter, overlooking the possibility, he
says, that the larger GDP might be promoting the larger population. The
empirical studies of economists, Keyfitz points out, result in large collections of
data but with uncertain conclusions, while the empirical studies of biologists are
often not expressed in quantified data from which useful conclusions can even be
attempted.
Remarkably few studies have been carried out in an attempt to
ascertain whether population growth in developing countries actually has
produced undesirable effects. The interdisciplinary nature of population growth
effects has made them difficult to study, even after some recent gains in
interdisciplinary studies on purely environmental subjects. The disciplines
61


involved in the population subject are diverse, including biology, economics,
social psychology, demography (normally within sociology, but sometimes
practiced by economists), cultural anthropology, and agriculture. The
scientists and social scientists use different vocabulary and publish in different
journals (Meeting the policy challenge 1992).
Defining the Solutions
The second key question or area of dispute concerns what factors or
policy actions reduce human birth rates and thus slow population growth.
Some groups insist that increased family planning programs are required to
reduce human fertility, while others point out variously that economic
development or opportunities and greater equality for women will bring birth
rates down. Numerous studies have been produced to demonstrate the
correctness of both assertions (Hodgson 1988; Bongaarts 1990; Hernandez
1984; Repetto 1979).
The argument that family planning programs are required to reduce
human fertility, or even that they will effectively reduce fertility in the absence
of other societal changes, has been countered for several decades by two forms
of demographic transition theory. Demographic transition theory was
developed in its modem form by Frank Notestein (1945) and Kingsley Davis
(1945). The basic shape of the theory is that improvements in nutrition,
hygiene and medical care in traditional societies cause high mortality rates to
drop, whereupon the population grows rapidly, and after a lag fertility rates
62


drop and population growth slows. The mechanisms for these respective drops
are complex, and are explained in at least two different ways (Cleland 1993),
but variations in the theory have in common economic and social development,
and reduced need for many children (including, in one variation, dependence
on a drop in infant mortality), along with increased cost of supporting children
and increased desire for smaller families. This paradigm has been widely
accepted without dispute for several decades, until the very recent appearance
of research demonstrating the theory's inability to explain the dramatic fertility
declines in much of Europe between 1880 and 1930, and also the more recent
fertility declines in a large number of Asian and Latin American countries
(Cleland 1993).
A complicating factor leading to confusion in this area has been
inadequate attention to the distinction between what is necessary and what is
sufficient (M. Campbell 1992a). This is especially the case in descriptions of
situations where birth rates have dropped under conditions of economic
improvement, where modem methods of family planning were available but
were not considered to have been the impetus for the decline (J. Simon 1981).
While history has shown some reductions in birth rates before the advent of
modem contraception, this has generally taken place with the use of crude
contraception supplemented by widespread abortion and infanticide, the latter
not widely accepted today in many societies. Modem birth rate reductions
have all taken place with the help of modem contraceptive methods, with the
single exception of the island of Mauritius (Engelman 1993).
63


\oz
More confusion arises also when the two dominant questions are
mixed. It is not uncommon to see the first question, whether population
growth is a problem, elicit responses to the second, what reduces fertility (M.
Campbell 1992a).
Sensitivity
Contributing to the elusiveness of problems in these areas, let alone
answers, is the perceived sensitivity of the subjects of population growth and
human fertility. Population has long been regarded a delicate subject because
of its close association with matters of life and death, sex, cultural values,
religion, race, political power, and the uneven distribution of wealth, resources
and opportunities (Lassonde 1992; M.Campbell 1992b; Teitelbaum 1992-93).
These all are value-laden areas of discourse, none of them easy to deal with in
a policy context.
The sensitivity may be due also to the population subject's close
relationship with well established causal beliefs serving to support values and
established social and economic systems that some groups would rather see left
unexamined and undisturbed. In particular, the close relationship between high
fertility and poverty easily raises questions of systemic inequities. In some
traditional societies, resistance to changing inequitable systems with respect to
either the poor or women may have served as a disincentive to address the
social and economic roots of population dynamics. Paradoxically, some
groups promoting improved equity have also exhibited a preference for
reducing attention to population, which they see as detracting from the problem
64


of persistent poverty, or from the need for individual or women's rights. All
of these different approaches to the population subject were exhibited during
UNCED.
The Ethical Mine Field
Values issues lend additional complications to the population field.
At least four fundamental ethical questions, especially when left unexamined,
tend to hinder progress toward mutual understanding of the issues. These are a
question about obligations to present versus future generations, a question
about individual versus community rights and benefits, a question about quality
versus quantity of life, and a question about the distribution of resources and
opportunity (Bayles 1980; Parfit 1990). The first question is whether one has
the right to require people to sacrifice today for the benefit of future
generations. The second question is whether one has the right to ask
individuals to reduce their welfare, or to go against their perceived self-
interest, for the welfare of the community (or the country, or the planet).
Derek Parfit (1990) explores the third question, concerning quality
versus quantity of life, in terms of average quality of life versus total quality of
life. Promotion of the higher average measure supposes that having fewer
people living with better nutrition, health, possessions and opportunities is
desirable. Promotion of the higher total measure supposes that the presence of
more people living with lower levels of these benefits is preferable. Parfit
takes this puzzle to its extremes, asking whether many millions of additional
65


people living with very limited resources, such as, in his example, nothing but
potatoes and muzak, might be considered preferable to fewer people with more
resources, and he raises the question of who is to decide.
The fourth question on the distribution of resources and opportunity
comes up in discussions of carrying capacity, a concept not infrequently
debated in population and environmental circles. Carrying capacity is defined
in the 1984 World Development Report as follows:
The carrying capacity of a particular region is the maximum population
of a given species that can be supported indefinitely, allowing for
seasonal and random changes, without any degradation of the natural
resource base that would diminish this maximum population in the
future. The concept of carrying capacity is familiar to biologists and
wildlife managers....with modifications, it is also an important measure
of the ability of regions to support human populations (World Bank
1984, 224).
Problems typically arise in the application of this concept to human populations
due to a country's or community's liberal consumption of resources from
beyond its own borders, as in the cases of the Netherlands or Hong Kong, due
to their ownership of capital and their ability to trade for goods. A citizen
from a relatively poor developing country with a rapidly growing population,
such as Ghana, may well be justified in suggesting that its carrying capacity is
far from exceeded when the Netherlands, with seven times Ghana's population
density, is not viewed as having exceeded its carrying capacity. A pertinent
question in this case concerns the perceived unfairness of distribution of wealth
and opportunity among countries, an issue often raised by developing country
voices.
66


George Stolnitz (1985) has posed a series of more specific ethical
questions facing policy makers dealing with rapid population growth, including
the appropriate degree of freedom of individual choice concerning fertility, and
whether the state, religions, individuals, or the free markets should decide on
the family planning information and methods to be made available to
individuals. These more specific questions and the four fundamental questions
described above served as largely unrecognized elements of the disagreements
about population during UNCED.
The Policy Context: The United Nations Conference
on Population and Development (UNCED)
The United Nations Conference on Environment and Development
(UNCED) was established by General Assembly resolution 44/228 in
December, 1989. The conference process covered two years, and included
four preparatory committee meetings, known as PrepComs, and the Earth
Summit in Rio de Janeiro, Brazil, June 3-14, 1992. PrepCom I was held in
Nairobi, August 6-31, 1990; PrepComs U and El were held in Geneva, March
18-April 5, 1991 and August 12-September 4, 1991; and PrepCom IV was
held at the United Nations in New York, March 7-April 4, 1992.
The Earth Summit took place at a conference facility twenty-five miles
from central Rio de Janeiro. This mammoth structure, built partially open for
a warm climate, had many acres of interior space. A number of the large
indoor rooms were fitted out to become proper United Nations' meeting
rooms, with seating and table space for several delegates from each of the 178
67


countries, facilities for simultaneous interpretation in several languages, and
seats for the press and NGO observers. The Global Forum, running
simultaneously with the Earth Summit, was held in a large number of sturdy
tents and temporary buildings set up along a half-mile section of Flamingo
Park in downtown Rio de Janeiro. This gathering of 30,000 representatives of
non-governmental organizations (NGOs), approximately ten times the size of
the Earth Summit, was considered an integral part of the UNCED process.
NGO participation in UNCED was in the original plan for the two-year
conference process. In addition to their attendance at the Global Forum,
selected NGO representatives participated both in country delegations and in a
lobbying capacity at the conference sites during the PrepComs and the Earth
Summit. ^ 1
The Earth Summit took place on the twentieth anniversary of the
first world environmental conference in Stockholm. The establishment of
UNCED was based on the studies produced by the World Commission on
Environment and Development, whose work had been commissioned by the
General Assembly in 1983 and was completed in 1987. This commission was
chaired by Mrs. Gro Harlem Brundtland, Prime Minister of Norway, and
became known widely as the Brundtland Commission (Strong 1991). The
A common understanding at PrepCom IV regarding the decision to include NGO
representatives in UNCED was that much of the expertise in environmental areas resided in the
NGOs, and that UNCED leaders recognized that many governments would not have created
environmentally beneficial policies without pressure from the environmental NGOs. These
reasons for the NGO relationship to the formal process were discussed briefly in a meeting for
NGOs in New York at the beginning of PrepCom IV, on March 11, 1992, attended by this
author.
68


Brundtland Commission's final report, Our Common Future (1987), presented
a compelling case for sustainable development, the integration of
environmental protection with economic development, seen as the only way to
guard against destruction of the environment, and also as the only way to
achieve on a lasting basis the development desired by the Third World
countries. Population was one of eight subjects formally addressed by the
Commission, and its report contains a statement that a reduction of current
population growth rates is imperative for sustainable development (Our
Common Future 1987).
The mission of UNCED, accordingly, was to find globally
acceptable ways to halt destruction of the environment and achieve sustainable
economic development for the less-developed countries. Sustainable
development became the operative term for the conference. The hopes of
UNCED Secretary-General Maurice Strong included a new level of support for
the developing countries (alternatively known as the South, or the Southern
countries) by the industrialized countries (the North, or Northern countries), to
provide the means for enjoying economic advancement without irreparable
damage to land, forests, water, and the fragile balances of life systems. The
results sought from UNCED included technical assistance, transfers of
technology, and "capacity-building" in terms of technology, professional skills
and institutions. In addition, financial assistance from the North to the South
was sought to support these envisioned massive changes. Secretary-General
Strong viewed these innovations in terms of a changed relationship:
69


The Earth Summit must establish a whole new basis for relations
between rich and poor, North and South, including a concerted
attack on poverty as a central priority for the 21st Century. This
is now as imperative in terms of our environmental security as it is
on moral and humanitarian grounds. We owe at least this much to
future generations, from whom we have borrowed a fragile planet
called Earth (Strong 1991, 3).
The United Nations resolution establishing UNCED listed eight
major issues to be taken up by the conference:
protection of the atmosphere by combating climate change, depletion of
the ozone layer, and transboundary air pollution;
protection of the quality and supply of freshwater resources;
protection of the oceans and other seas;
protection and management of land resources, including combating
deforestation, desertification, and drought;
conservation of biological diversity and environmentally sound
management of biotechnology;
environmentally sound waste management;
improvement of the environment of the poor in both urban and rural
areas; and
protection of human health and improving quality of life (Fletcher
1992). 12
Population was not on this list, having been purposely excluded as peripheral
and potentially divisive (Kalish and Carty 1992). Objections to its presence on
the agenda included fears by some delegates from the South that the financial
assistance they sought might be predicated on a slowing of population growth
rates, and also a reluctance by other Third World delegates to link population
12 For a thorough discussion of UNCED processes and outcomes, see this report on the conference
produced by Dr. Susan Fletcher (1992) at the Congressional Research Service.
70


growth and environmental problems, given the environmental impact of the
well known high level of consumption in the North.
During and following PrepCom I, representatives of various
nongovernmental organizations (NGOs) voiced their objections to UNCED's
neglecting the population component of environmental concerns. Maurice
Strong accepted population as a "cross-cutting issue," and it was decided that
population would be brought up in conjunction with every relevant discussion
topic. In a plenary session of PrepCom n, Ambassador Curtis Bohlen,
speaking as head of the United States delegation to UNCED, presented a case
supporting the inclusion of population on the conference agenda and as a
separate chapter in Agenda 21. the action statement to be approved at the Earth
Summit (Ryan 1993). Advocates for attention to population growth attended
PrepCom HI with the intention of winning this place on the agenda (Kalish and
Carty 1992). The population subject was put on the agenda only when the
Northern countries, and in particular the United States, ultimately agreed to the
inclusion of consumption issues as well. By the end of PrepCom IE, it had
been agreed that both subjects, population and consumption, would be included
as separate and equal chapters in Agenda 21. The UNCED Secretariat staff
drew up a draft agenda and circulated it among organizations known for their
attention to international population issues,^ as well as a number of individual
13 Organizations solicited for text contributions to the negotiating document for the demographic
chapter of Agenda 21 included the United Nations Population Fund (UNFPA), the Sierra Club,
the Population Crisis Committee, Zero Population Growth, the Futures Group, the Worldwatch
Institute, and the National Wildlife Federation (Kalish and Carty 1992).
71


population experts, for contributions to the text (Lassonde 1992). The
resulting, expanded text became the negotiating document for Chapters 3, 4
and 5 of Agenda 21. covering poverty, consumption and population.
Throughout PrepCom IV and the Earth Summit, these chapters were referred
to in plenary sessions as "the poverty cluster".
It was important to many delegates from developing nations not to
place the blame for environmental decline in their own countries' population
growth, but to stress instead the role of Northern countries' high levels of
consumption. This was consistent with their fears that assistance from the
North might be tied to their levels of success in reducing their population
growth (Kalish and Carty 1992). It was widely understood at the preparatory
meetings that these countries had high expectations that they would receive a
substantial amount of financial support from the industrialized countries as a
subsidy for instituting technologies consistent with sustainable development in
place of cheaper, more damaging, less sustainable methods of production
(Lassonde 1992; Harkavy 1992; The World Commission 1992; Deen 1992;
Rolston 1992; Adams 1992). Accordingly, conference attention was
repeatedly diverted from Southern population growth to Northern
consumption, and also to conditions limiting developing countries'
opportunities to reduce their persistent poverty, such as trade arrangements
favoring the North, commodity pricing, debt burdens and structural adjustment
during the 1980s designed to relieve these debt burdens (Weisskopf and
Robinson 1992; Adams 1992; Earth Summit Bulletin 1992). The concept of
72


international distributional inequity supported the need for financial support
from the North.
Population later became a leverage item in an effort by the Group
of 77 (G-77) 14 t0 obtain this funding. This connection was clarified during a
press conference by Ambassador Jamsheed Marker during PrepCom IV at the
United Nations. Ambassador Marker, who represents Pakistan at the United
Nations and chaired the meetings of the Group of 77, was asked when that bloc
would be ready to "move" on population. His reply: "The G-77 will be ready
to move on population when the North is ready to move on finances. In the
end, the massive North to South transfer payments were not forthcoming.
This Southern resistance to focusing on population growth stood in
contrast to those countries' interest in the population subject in their UNCED
country reports. All countries participating in UNCED were asked to submit
to UNCED offices in Geneva before the Earth Summit country reports
containing their governments' appraisals of their environmental condition,
problems, and prospects. Nine out of ten developing countries referred to
population growth and distribution among their major concerns. Many of them
put population near the top of their agenda (Sadik 1992a).
There may have been three separate reasons for Southern resistance
to attention to the population subject at UNCED. First, the resistance served
14 The Group of 77 is the large bloc of developing countries, by the time of UNCED numbering
129, that coordinates on many policy approaches. During PrepCom IV, the G-77 countries'
delegates met daily behind closed doors.
15 Press conference in the United Nations, March 11, 1992, attended by author.
73


as leverage to achieve other objectives, in particular financial assistance from
the North to assist in their economic development efforts in conjunction with
environmental sustainability. A second reason may have been a sovereignty
issue, wherein addressing one's own country's population growth problems is
considered more palatable than placing such concerns into international
agreements, where a country's human fertility rates implicitly become the
business of other countries. A third reason, or justification, for a relatively
low level of attention to the population subject appeared to be the United
Nations' own plans for future conferences. It was well known that the United
Nations was planning an International Conference on Population and
Development (ICPD) only two years away, scheduled for September 1994,
when more definitive agreements on population would presumably be reached
(Stevens 1992). It was not uncommon during PrepCom IV and the Earth
Summit to hear country delegates commenting informally that any statement on
population produced in UNCED would not be critically important because the
more definitive population agreements would be produced only two years later.
Those Southern delegates who provided the resistance to the
population subject in UNCED represent one of the groups studied in this
research. A second group preferring reduced attention to population growth in
UNCED was the Vatican, seated in United Nations meetings with country
delegates as the Holy See, Permanent Observer State. 16 The Holy See's first
active involvement in the policy development process of UNCED was in
16 Switzerland has this same status in the UN.
74


PrepCom IV in New York (March 2 to April 4, 1992). Its delegates tended to
shift attention away from population concerns in developing countries in favor
of attention to poverty and inequity issues, and they held firmly to the Catholic
Church's traditional stand against artificial means of birth control.
Vatican thinking was reportedly influential over two points in the
negotiations. In a contact session, 17 delegates of the Holy See, backed by the
Philippines and Argentina ("The nearly forgotten factor..." 1992; Allen and
Gutfeld 1992), insisted that language respecting individual choice regarding
means to determine family size be supplemented with language allowing for the
control of individuals by cultural values (Carter 1993). This language, "taking
into account ethical and cultural considerations," is found at the end of the
following passage:
Governments should take active steps to implement, as a matter
of urgency, in accordance with country-specific conditions and
legal systems, measures to ensure that women and men have the
same right to decide freely and responsibly on the number and
spacing of their children, to have access to the information,
education and means, as appropriate, to enable them to exercise
this right in keeping with their freedom, dignity and personally
held values taking into account ethical and cultural
considerations (Agenda 21. Chapter 5, paragraph 5.50).
17 "Contact sessions were the closed-door negotiating meetings throughout UNCED, each
attended by approximately twenty delegates, usually self-selected. No minutes were taken, and no
written record of these meetings exists. There are no standardized rules for this process in the
UN. Each contact session leader or convener is free to establish procedures. Information about
these meetings is obtainable only through conversations with participants willing to share the
experience. On population, three contact sessions took place during PrepCom IV, resolving all
major issues so that none would be carried over to the Earth Summit.
75


According to Ryan (1993), the Holy See delegates were ready, after some
extended discussions in the contact session, to back down from their insistence
on inclusion of this passage, but the delegates of the Philippines and Argentina,
whom the Holy See had enlisted to support them, were unwilling to give up on
this point.
The second Vatican-inspired change took place in a plenary session
in the last days of PrepCom IV, when the Holy See said it could not accept the
terms "family planning" or "contraception" in the language of the Agenda 21
chapter on demographics. The words "family planning" were changed to "the
responsible planning of family size" (Holy See 1992; Carter 1993) in the
following passage.
Governments should take active steps to implement programmes
to establish and strengthen preventive and curative health
facilities that include women-centred, women-managed, safe and
effective reproductive health care and affordable, accessible
services, as appropriate, for the responsible planning of family
size, in keeping with freedom, dignity and personally held values
and taking into account ethical and cultural considerations
(Chapter 5, Agenda 21. paragraph 5.51).
The Holy See subsequently was subjected to criticism in a variety of news
reports and NGO accounts of these events (Allen and Gutfeld 1992; Oliver
1992). A two-page, undated statement was issued from Rome in time for the
Earth Summit, denying reports that it had removed population from UNCED,
and clarifying its position.
The Holy See has not attempted to eliminate any wording
relating to population, but only to improve it, reaffirming respect
76


for liberty and for the conscience of the human person,
defending the poorest of the poor from the unjust supposition,
that due almost to the very fact of their very existence and that
they are numerous, that they are the cause rather than the victims
of their underdevelopment and of ecological degradation .... If
instead of "family planning" the document now reads
"responsible planning of family size," who can honestly object to
the choice of spouses being responsible? (Holy See 1992)
The critical observers who identified with the population-concerned school of
thought were not satisfied, knowing that the word "responsible" in the context
of human reproduction had long meant, in Vatican usage, avoiding the use of
modem methods of family planning (Nagel 1978). The Holy See's mixing of
the euphemistic and general meanings of "responsible" in a single sentence was
seen as tactically clever. 18
At the end of PrepCom IV, only hours before the passage of the
demographic chapter of Agenda 21, Archbishop Renato R. Martino, head of
the Holy See delegation to UNCED, said in plenary session that the Catholic
Church "opposed the imposition of demographic policies and the promotion of
methods for limiting births" (Oliver 1992, 11). Observers were somewhat
confused, then, when Archbishop Martino said in the opening session of the
Earth Summit, "The position of the Holy See regarding procreation is
frequently misinterpreted. The Catholic Church does not propose procreation *
*8 The term "responsible played a prominent role in another population policy situation. The
Mexican government, in instituting its new program of population control (their chosen term) in
1972, made use of the term "responsible parenthood", long familiar to them in the Catholic
Church as a restriction on the use of modem birth control methods, by redefining the term for
billboards and other publicity to mean providing good nutrition and education for a small, healthy
family (M. Campbell 1989).
77


at any cost" (Martino 1992). These were new words for the Vatican, and
raised some hopes and a great deal of curiosity at Rio about a possibility of
policy change within the Church with regard to contraception.
Organized feminists constituted a third group resisting attention to
population growth in UNCED. Although only thinly represented in the official
delegations to both PrepCom IV and the Earth Summit, concerted lobbying
efforts at PrepCom IV resulted in incorporation of language about women's
rights and participation into many chapters of Agenda 21. including the
demographic chapter. 19 According to Nancy Carter, Population Coordinator
in the United States Department of State and a negotiator for the United States
throughout UNCED, some of the key language in the demographic chapter was
drawn directly from the chapter of Agenda 21 devoted to women, benelitting
from the negotiations that had produced the language for that chapter (Carter
1993).
During the two weeks of the Earth Summit, feminists organized by
the Brazilian Women's Coalition gave speeches and held discussions in the
Women's Tent, the largest tent at the Global Forum. Two days of these
speeches and discussions were devoted to population and population policies.
Leaders of this activist group were from Brazil, India, Barbados, Peru and the
19 The active lobbyists for women's interests were led by former U.S. Congresswoman Bella
Abzug of New York, chairman of the Women's Environment and Development Organization
(WEDO). An NGO, WEDO produced during PrepCom IV a 68-page document commenting on
and making recommendations for the contents of Agenda 21. section by section. This document's
recommended changes in language supported both the inclusion of women as participants or
beneficiaries of actions and systems, and opportunities for developing countries, children, and
indigenous peoples.
78


United States, and women from Nigeria, Bangladesh, Norway and Canada
participated as speakers in the program as well. Repeated themes through
these talks were a flat denial that population growth causes environmental
decline and an abhorrence of family planning programs designed to reduce
fertility, seen as insensitive and often abusive of women. During these
sessions, an NGO alternative "treaty" on population was developed in response
to the Agenda 21's demographic chapter,20 which was viewed by the more
vocal participants in the Women's Tent discussions as the illegitimate outcome
of a patriarchal system. These activists tended to view the predominantly male
diplomatic corps in the UN processes as unqualified to speak on population, a
subject seen by the women as fundamentally about women's reproduction.
UNCED Secretary-General Maurice Strong, not having been
previously a vocal advocate for attention to population concerns, spoke
vehemently on the subject in his opening address at the Earth Summit, most
memorably with the statement, "The world's population must be stabilized. If
we do not do it, nature will, and much more brutally" (Strong 1992). Prime
Minister Brundtland's remarks followed, with more support for slowing the
growth.
Unless poverty is alleviated, there is no chance that we will be
able to stabilise the world population. It has grown by 500
million since the Commission last met five years ago. We must
deal with population growth through an integrated human rights
approach, including education and the enhancement of the status 20
20 The women's alternative treaty is Appendix B of this thesis.
79


of women, improved public health, and family planning
(Brundtland 1992).21
Two days later, Captain Jacques Cousteau delivered a forty-five minute
address in the Press Room, all of it on the need to reduce population growth.
The sudden high visibility of this subject seemed a startling change to people
not accustomed to hearing from the leadership of the conference on this
subject. The language for the demographic chapter of Agenda 21. however,
had already been decided upon during PrepCom IV two months earlier,
without the benefit of this kind of publicity for the subject. The chapter as
finally agreed upon22 was reasonably comprehensive but also less ambitious in
stated objectives than the language that had been agreed upon eight years
earlier at the United Nations 1984 International Conference on Population in
Mexico City. As described by an analyst from the Energy and Environmental
Study Institute in Washington, the chapter on population
.. .fails to reflect the importance of the issue to global sustainable
development. It avoids the term "family planning" and makes
only oblique reference to contraceptive technologies and
information. There is no reference to targets for stabilizing
global population growth already agreed to by the world
community, and no specific recommendations for action. A
provision calling for equal rights for women in deciding on the
number of spacing of their children is qualified by the phrase "in
21 Two days after this speech, Prime Minister Brundtland spoke in the Women's Tent at the
Global Forum, and, having to leave to return to the Earth Summit before an opportunity to take
questions, received heavy criticism in her absence from the organizers, who objected to her
attention to population growth and to her desire to find ways to reduce that growth (Women's
Tent tapes, June 6, 1992).
22 See the demographic chapter, Chapter 5, Agenda 21. Appendix A.
80


accordance with country-specific conditions and legal systems"
(Porter 1992).
The Agenda 21 chapter might best be seen as additive, containing statements of
the general wishes of each faction involved. Many deletions were made as
well: much of the language of the original negotiating document was removed
in the negotiating process by various delegates' objections. The word
"population" was changed, at the request of developing country negotiators, to
the less sensitive term "demographic" virtually throughout the chapter (Kalish
and Carty 1992). The loss of the term "family planning from the document
reversed language that had been accepted in United Nations agreements for
twenty years.
According to Carter, a perceived responsibility of the contact group
was to reach agreement on language with as few brackets^ as possible by the
end of PrepCom IV (Carter 1993). It appears that consensus was defined as
unanimity during that process. Under that condition, in which everyone must
be satisfied, each participant effectively has veto power, because the last hold-
out can force the group to shift to satisfy his or her wishes. Accordingly, the
chapter has been widely seen as a weak, lowest-common-denominator
agreement, not destined to serve as an influential document shaping global
thinking on population issues in the future (Porter 1992).
One view of the significance of the population policy formulation
process in UNCED is that the circumstances leading to the relatively bland, 23
23 The term "brackets" in the UN context means bracketed language within agreements that are in
the formative process, those passages on which consensus has not yet been reached.
81


additive policy statement in Agenda 21 provided an interesting view of
maximum divergence of opinion in the field, in the form of relatively undiluted
positions of differing groups. This has presented an opportunity to examine
raw, uncompromised sets of beliefs and policy preferences on population
before the preparations for ICPD, the upcoming International Conference on
Population and Development, were in full swing.24
Tying the Policy Context to the Theory
Looking back at Haas' statement about conditions leading to the
dominant influence of an epistemic community ,25 it is clear that no one
epistemic community succeeded in having its one brand of knowledge about
population accepted by the policy community as the consensual knowledge in
UNCED, and thus no one group could claim dominant influence over the
policy development process. The competing views about what would
constitute appropriate population policy were far apart, but also different in
terms of how the groups saw the issue-area. In his theory, Haas did not appear
to recognize the possibility for this type of configuration of competing interest
groups. This research will take a closer look at the five groups that have been
identified, with the expectation that this policy example will have implications
for Haas' theory. In the process, it might also be possible to develop a better
sense of what it might take for these influential groups to reach agreement on 24 25
24 The first of the regional conferences for ICPD opened in Bali, Indonesia, only six weeks after
the close of the Earth Summit.
25 See page 15, in Chapter II of this thesis.
82


policy objectives, or at least a clearer view of the obstacles to that ultimate
convergence in the policy issue-area envisioned by Haas.
83


CHAPTER IV
DESIGNING THE ANALYSIS
Purpose of the Analysis
If it can be shown that several of the groups active in international
population policy development during UNCED qualify as epistemic communities
according to Haas' specifications, their presence in the policy arena will have
implications for Haas' theory. First, such a finding would be consistent with
Haas' observation about influence, as stated in Chapter H:
To the extent to which an epistemic community consolidates
bureaucratic power within national administrations and
international secretariats, it stands to institutionalize its influence
and insinuate its views into broader international politics (1992b,
4).
This statement might be confirmed with the UNCED example in the form of an
approximated contrapositive (reversed and negative) form of this statement: To
the extent to which no epistemic community has been able to institutionalize its
influence and insinuate its views into broader international politics during
UNCED, no one epistemic community has succeeded in consolidating
bureaucratic power in ... the international secretariat. In this policy case, no
one epistemic community succeeded in consolidating power among the hundreds
of delegates from the 178 countries taking part in the conference.
Secondly, a demonstration of the presence of several epistemic
communities whose experience and expertise were outside the issue-area as it
84


had traditionally been defined would suggest an extension to Haas' theory is
needed: that a policy issue-area can attract the involvement of authoritative,
knowledge-based groups representing expertise in subjects other than the subject
originally defining the issue-area, but who have a keen interest in the policy
outcomes of that issue-area. In bringing to the issue-area an alternate definition
of the problem in terms of their expertise on other subjects, they must be
accepted as epistemic communities with authoritative expertise in the issue-area.
Such communities would normally have a primary interest completely outside
the issue-area, and it is possible for one or more of them to have a secondary or
instrumental interest in reducing attention to the original or traditional
conception of the issue area.
Haas' liberal specification requiring only internal validation for a
group to be considered authoritative or expert has opened the door to such an
extension to his theory. That this is applicable to other policy issue-areas seems
entirely reasonable in light of the current configuration of the health care reform
issue-area in the United States. 26 According to Haas' specification, we might
decide that holistic medicine could qualify as a genuine epistemic community in
the health policy issue-area, for example, because practitioners of holistic
medicine recognize their own body of knowledge about health as authoritative
and recognize expertise within their own group. Such an example is not
qualitatively different from the situation in the population policy arena: clearly, 26
26 in health care reform, the problem is defined differently by physicians, the disabled, the
uninsured, hospitals, insurance companies, small businesses, etc. Physicians who speak out on
health care reform, for example, tend to view the issue-area in terms of quality of care, which in
turn is seen as dependent on Americans' ability to select their own doctors.
85


the organized women's feminist history, shared experience and philosophy as
that interest group's knowledge base and subject of expertise, and the Vatican's
extensive body of knowledge in moral theology as that group's common
knowledge base and subject of expertise, must be acknowledged as legitimate
expertise in the population issue-area once these groups have brought their
perspectives on population into the policy arena.
While tying their expertise to the population subject in terms of its
well established beliefs about the nature and importance of human life (the
Vatican), and in terms of women's rights and reproductive health issues (the
organized women), these active groups have, in effect, redefined the issue, or
the policy problem, in their own respective terms, and they have in that process
become, in Haas' internally validated sense, experts on the population subject.
Similarly, the Southern country delegates' position resisting attention to
population growth issues might indicate the presence of another epistemic
community, as might the market conservative position, representing many
economists who consider population growth either beneficial or unimportant.
Haas appears not to have seen this possibility, probably because relatively
technical issues engender relatively little debate about how the problem is
defined. In contrast, social, complex, value-laden, multi-cultural issues may
have a number of problem definitions, each one supported by its own
authoritative epistemic community.
A central activity of this research is to explore one of the reasons
why there was no definitive population policy formulated at Rio. According to
theorists like Haas, when a consensus is reached among acknowledged experts
86


in an issue-area, policy development can move ahead unhampered by conflict
and with considerable support. It will be argued here that there was not one
consensual group in the population policy issue-area during UNCED, but instead
multiple groups whose differing objectives were based on largely contradictory
knowledge; and that this lack of consensus among groups may be one of the
reasons why no cohesive population policy was forthcoming.
Accordingly, a proposition of this research is that several of the
groups qualify as epistemic communities, each with its own brand of
authoritative knowledge and expertise on the population subject. An objective,
then, then, is to see which of the groups influential in the international
population policy development at UNCED one might be justified in labeling as
epistemic communities. The proposition underlying the following analysis is
that several of these groups will qualify as epistemic communities. If some do
qualify, there will be implications that might appropriately be viewed in this
research as subsidiary propositions. First, the professional pedigrees that Haas
sees as necessary for members of an epistemic community might not in fact be
required. Second, distinguishing between interest groups and epistemic
communities in the way that Haas has proposed may be, in at least this policy
case and perhaps in some other issue-areas as well, unwarranted and not
especially useful. Third, Haas' observation that members of an epistemic
community will, when confronted with an anomaly undermining their beliefs,
will withdraw from the debate. This third proposition is offered in light of the
theoretical literature in cognitive psychology reviewed in Chapter II of this
thesis, showing that exposure to dissonant beliefs does not necessarily reduce the
87


strength of an individual's original position, for people exhibit other behaviors to
protect their beliefs.
Research Methodology
The qualitative method to be used in this exploratory research is
comparative analysis, described by Glaser and Strauss (1967) as a means to
develop grounded theory. The authors recommend taking a comprehensive
view of comparable groups, situations or sets of data to see what emerges. They
stress that the groups must have some characteristics in common, and that one
must include all groups within the perceivable set of those characteristics.
This method views theory as an ever-developing process, not as a
perfected product (Glaser and Strauss 1967). The authors stress the importance
of starting with a formal theory on the basis of which to collect data, and point
out that it is the appearance of a case that does not fit that theory adequately that
can serve as a point of departure toward explanation in the form of theoretical
adjustments beyond the original theory. This situation is precisely the point of
departure for this dissertation research.
According to Glaser and Strauss, while the subject is examined
within a theoretical framework, this form of comparative analysis differs from
deductive, proposition-based analysis in its freedom to move beyond the limits
of the original theoretical framework. The new theory thus developed can be
somewhat speculative or tentative, but derives its legitimacy from its grounding
in observed conditions.
88


The approach to the research in this dissertation is consistent with
interpretive research in Public Administration as described by J. D. White.
According to White,
Interpretive theory seeks an enhanced understanding of social
situations, not only for the researcher but also for those involved
in the situation....Interpretation seeks to understand the meanings
that actors attach to their social situations, to their own actions,
and to the actions of others....The basic aim of the interpretive
model is to develop a more complete understanding of social
relationships and to discover human possibilities (1986, 16).
White explains further that this research makes possible generalizations derived
from observations of patterns of behavior, and distinguishes it from positive
research which "requires the development of a collection of related and testable
laws that state causal relationships among relevant variables":
Interpretations tell us what is the case or "just what's going on
here"....Interpretive research is concerned with the meanings that
people attach to the norms, rules, and values that regulate their
interactions. Care is taken not to impose a prior understanding of
norms, rules, and values upon others, but rather to understand
their beliefs and actions from their point of view (1986, 18).
The comparative analysis to be performed can be seen also as content
analysis, but without a study of frequency of messages, words or phrases. Klaus
Krippendorff has noted that, "When historians make an effort to infer events
from documents, they are, by definition, involved in content analysis".^
Content analysis is often seen as a frequency-driven process, but, Krippendorff 27
27 This has been stated by Krippendorff (1980, 24) with reference to V. K. Dibble, "Four Types of
Inferences from Documents to Events," History and Theory 3, no. 2, 1963: 203-221.
89


warns, the quantitative approach to analyzing documents of a political nature
tends to be inadequate. He explains:
...quantitative indicators are extremely insensitive and shallow in
providing political insights. Even if large amounts of data are
available as required for statistical analyses, they do not lead to
the "most obvious" conclusions that political experts are easily
able to draw upon and to agree upon by observing qualitative
changes more in depth (1980, 17).
Another reason for seeing a quantitative approach to content analysis as
inappropriate for this particular research is the insignificance of frequency of
word occurrence in the literature of the five groups. The objective is not to
determine the relative support of a subject by a given group, but instead to view
the groups' positions, in order to determine which ones might qualify as
epistemic communities. Greater frequency of expression of an opinion in this
literature is not necessarily indicative of a stronger position in a group's
message, and infrequent utterances must not be interpreted as reduction in
involvement or influence in the policy process. This is particularly true in the
case of the Vatican, which is led by one authoritative voice from which a single
statement is sufficient to define a position, and the developing country voices
who, in their eagerness to reduce attention to the population growth factor in
development or environmental problems, have been influential but rarely bring
the subject up in print.
90


Overview of the Research Plan
As explained in the introduction to this dissertation, the research plan
contains two stages of analysis. In Stage 1, the five identified groups will be
examined individually to learn which ones exhibit consistently held causal
beliefs. Evidence of shared beliefs is not sufficient to establish the presence of
epistemic communities, because Haas specifies that the epistemic communities
must also possess authoritative knowledge or expertise in the issue-area. A
decision on which of the groups can be labeled epistemic communities, then,
will need to be based on the results of this analysis in combination with a
decision about whether a group exhibits expertise in the population issue-area.
Although examining principled beliefs is not required to determine
whether the groups constitute epistemic communities,^ Stage 1 will include a
review of principled beliefs revealed by each group, because this category of
Haas' provides a view of the groups that goes beyond the highly visible interests
by which the groups were identified.
Stage 1 will be followed by a discussion of the results, including any
complications that developed in the process of implementing the analysis. Stage
2 of the analysis will consist simply of a rearrangement of the five groups'
positions, with the analytic categories adjusted in response to problems
encountered in Stage 1. One way to handle the anticipated problems is to see
them as opportunities for further analysis, and, using these, Stage 2 is expected
to provide a useful view of the groups' configuration in the population policy 28
28 This will be explained in the section below titled Identifying the Criteria.
91


arena. This second stage is not critical to judging whether groups qualify as
epistemic communities, but it is expected to provide insights for the shaping of a
recommended extension of Haas' theory.
Operationalizing the Analysis
Data Collection
Messages exhibiting causal beliefs concerning a set of subjects
commonly connected to population growth issues, and presented in the context
of population discussion, were identified and coded, and constitute the data in
the analysis. The units of analysis are the five groups identified as having been
influential in international population policy. The data sources are either
individuals or organizations representing these groups, serving as authors or
speakers. Data sources whose documents or speeches (hereafter to be referred
to simply as documents) are to be reviewed have been selected for their purist
representation of their respective groups' highly visible interests, a criterion to
be explained later in this chapter. 29
In these documents, sometimes a message will be provided in only a
single statement. Example:
If the earth's air is so foul, why do we continue to five longer and
longer? People can create pollution, but man's ingenuity
invariably finds ways to reduce or eliminate it. Just look at our
sewer systems .... There is no correlation between population and
pollution (Forbes 1992, 25). 29
29 See Data Sources section in this chapter, following Identifying the Groups.
92


More often, a single message is found repeated in a number of similar but not
identical statements within the same document, amounting to conveyance of the
message in a variety of explicit and implicit forms. The five statements below,
from a single document, constitute an example of a single message in numerous
forms:
Population growth in the Third World is being increasingly and
falsely identified as a primary cause of environmental destruction
(Shiva 1991b, 33).
Even documents not related to population issues erroneously
identify population growth as a cause for environmental
destruction (Shiva 1991b, 33).
There are four main reasons why population growth cannot be
identified as the primary cause of environmental destruction
(Shiva 1991b, 33).
[P]opulation growth is not a cause of the environmental crisis but
an aspect of it and both are related to the alienation of resources
and destruction of livelihoods, first by colonialism and then by
Northern-imposed models of maldevelopment (Shiva 1991b, 33).
The focus on population as the cause of environmental destruction
is erroneous at two levels (Shiva 1991b, 33).
In this case, the message found in all five statements, that population growth is
not a cause for environmental destruction, will be coded as a single message.
The sheer cumbersomeness of listing every individual statement containing an
explicit message, plus every variation of that message, including portions of
messages and hints of relevant meaning detected in the author's choice of
vocabulary and syntax, would amount to overwhelming and unproductive labor.
93